← Back
Theo September 11, 2026 1h 16m

Fable Vs Astra Debate Is Over

Read full transcript 59 segments
  1. Last week was pretty crazy for new AI Last week was pretty crazy for new AI models. We got two new heavy hitters models. We got two new heavy hitters models. We got two new heavy hitters that are truly changing the game. Gemini that are truly changing the game. Gemini that are truly changing the game. Gemini 38 Flash and Muse Spark 1.3. How can you 38 Flash and Muse Spark 1.3. How can you 38 Flash and Muse Spark 1.3. How can you ever decide between those two? They're ever decide between those two? They're ever decide between those two? They're both so incredible. Obviously, I'm both so incredible. Obviously, I'm both so incredible. Obviously, I'm joking. We're actually here to talk joking. We're actually here to talk joking. We're actually here to talk about Fable 5.1 versus GPD6 Astra, the about Fable 5.1 versus GPD6 Astra, the about Fable 5.1 versus GPD6 Astra, the two actual models that matter. And I'm two actual models that matter. And I'm two actual models that matter. And I'm sure my attention guy already wants to sure my attention guy already wants to sure my attention guy already wants to kill me for doing that joke and kill me for doing that joke and kill me for doing that joke and confusing people. So, sorry. Well, let confusing people. So, sorry. Well, let confusing people. So, sorry. Well, let me have some fun. Okay, everyone's going me have some fun. Okay, everyone's going me have some fun. Okay, everyone's going to watch this video anyways cuz they all to watch this video anyways cuz they all to watch this video anyways cuz they all want to know what model to choose. And want to know what model to choose. And want to know what model to choose. And these two models are unbelievable. I these two models are unbelievable. I these two models are unbelievable. I went in with pretty high hopes on Fable went in with pretty high hopes on Fable went in with pretty high hopes on Fable 5.1, but didn't expect too much and I 5.1, but didn't expect too much and I 5.1, but didn't expect too much and I still managed to be absolutely blown still managed to be absolutely blown still managed to be absolutely blown away. It's an incredible model and I use away. It's an incredible model and I use away. It's an incredible model and I use it every day. GBD6 Astra is the future. it every day. GBD6 Astra is the future. it every day. GBD6 Astra is the future. It is such a crazy leap in so many It is such a crazy leap in so many It is such a crazy leap in so many different ways. I love using it. I'm different ways. I love using it. I'm different ways. I love using it. I'm doing thousands of dollars of inference doing thousands of dollars of inference doing thousands of dollars of inference a day with it. These models are both a day with it. These models are both a day with it. These models are both incredible. If you want to just skip to incredible. If you want to just skip to incredible. If you want to just skip to the end to see which model's better, I the end to see which model's better, I the end to see which model's better, I guess you could do that, but you'd miss guess you could do that, but you'd miss guess you could do that, but you'd miss all of the different ways we can compare all of the different ways we can compare all of the different ways we can compare them because these models are very them because these models are very them because these models are very different in their capabilities and the different in their capabilities and the different in their capabilities and the things you can do with them. And there's things you can do with them. And there's things you can do with them. And there's a lot to be learned from the various a lot to be learned from the various a lot to be learned from the various skills these models do and don't have.

  2. skills these models do and don't have. skills these models do and don't have. I'm doing nearly a billion tokens a day I'm doing nearly a billion tokens a day I'm doing nearly a billion tokens a day across both of these models across both of these models across both of these models consistently. And I've seen their consistently. And I've seen their consistently. And I've seen their strengths and weaknesses in all sorts of strengths and weaknesses in all sorts of strengths and weaknesses in all sorts of different places, many of which have different places, many of which have different places, many of which have surprised me a ton. So, if you want to surprised me a ton. So, if you want to surprised me a ton. So, if you want to understand what these are good at, what understand what these are good at, what understand what these are good at, what they are bad at, how efficient they are they are bad at, how efficient they are they are bad at, how efficient they are with token usage, how hard the with token usage, how hard the with token usage, how hard the subscription limits will hit you, all of subscription limits will hit you, all of subscription limits will hit you, all of those types of things and more, and of those types of things and more, and of those types of things and more, and of course, the most important question, course, the most important question, course, the most important question, which sub should you buy today? I'll do which sub should you buy today? I'll do which sub should you buy today? I'll do my best to answer all of that after a my best to answer all of that after a my best to answer all of that after a real quick break for today's sponsor. real quick break for today's sponsor. real quick break for today's sponsor. Writing code has never been easier, but Writing code has never been easier, but Writing code has never been easier, but as more agents spam more slop at as more agents spam more slop at as more agents spam more slop at projects, it's ever been harder to projects, it's ever been harder to projects, it's ever been harder to maintain a cohesive, coherent project. maintain a cohesive, coherent project. maintain a cohesive, coherent project. Your actual repos are slowly falling Your actual repos are slowly falling Your actual repos are slowly falling apart as more and more agents spam code apart as more and more agents spam code apart as more and more agents spam code at them that isn't really verifiable at them that isn't really verifiable at them that isn't really verifiable with things as simple as lint rules and with things as simple as lint rules and with things as simple as lint rules and static checks. How do you know that the static checks. How do you know that the static checks. How do you know that the code coming in is actually following code coming in is actually following code coming in is actually following your architecture or doesn't have some your architecture or doesn't have some your architecture or doesn't have some small annoying quirk that a bunch of small annoying quirk that a bunch of small annoying quirk that a bunch of other code did in the past or that it's other code did in the past or that it's other code did in the past or that it's following your UI system and guidance following your UI system and guidance following your UI system and guidance the way that you want it to. You could the way that you want it to. You could the way that you want it to. You could add all of these things into your agent add all of these things into your agent add all of these things into your agent MD, but that only guarantees that your MD, but that only guarantees that your MD, but that only guarantees that your agent knows about them, not that it's agent knows about them, not that it's agent knows about them, not that it's actually going to follow. How do you actually going to follow. How do you actually going to follow. How do you make sure at code review time that all make sure at code review time that all make sure at code review time that all of these things are done properly?

  3. of these things are done properly? of these things are done properly? Wouldn't it be nice if you could control Wouldn't it be nice if you could control Wouldn't it be nice if you could control the agents reviewing your code and maybe the agents reviewing your code and maybe the agents reviewing your code and maybe spawn specific ones that will only spawn specific ones that will only spawn specific ones that will only approve if certain conditions are met? approve if certain conditions are met? approve if certain conditions are met? Wouldn't it be nice if you could just Wouldn't it be nice if you could just Wouldn't it be nice if you could just throw a markdown file in your repo that throw a markdown file in your repo that throw a markdown file in your repo that describes a specific behavior or pattern describes a specific behavior or pattern describes a specific behavior or pattern that you want to make sure exists in all that you want to make sure exists in all that you want to make sure exists in all code coming in and won't approve it if code coming in and won't approve it if code coming in and won't approve it if it doesn't match? Maybe you even choose it doesn't match? Maybe you even choose it doesn't match? Maybe you even choose which paths in the codebase need this which paths in the codebase need this which paths in the codebase need this review. Well, spoiler, this is a real review. Well, spoiler, this is a real review. Well, spoiler, this is a real file in the T3 Code codebase, and it's file in the T3 Code codebase, and it's file in the T3 Code codebase, and it's powered by today's sponsor, Macroscope. powered by today's sponsor, Macroscope. powered by today's sponsor, Macroscope. Not only do they have some of the Not only do they have some of the Not only do they have some of the fastest and most reliable code reviews fastest and most reliable code reviews fastest and most reliable code reviews of any of the review agents available, of any of the review agents available, of any of the review agents available, they also have an approval system that they also have an approval system that they also have an approval system that actually makes sense for projects that actually makes sense for projects that actually makes sense for projects that are moving as fast as something like T3 are moving as fast as something like T3 are moving as fast as something like T3 Code. We're about to break a thousand Code. We're about to break a thousand Code. We're about to break a thousand open PRs, which is just unfathomable. open PRs, which is just unfathomable. open PRs, which is just unfathomable. I'd be a lot more stressed about that I'd be a lot more stressed about that I'd be a lot more stressed about that number if we didn't have these custom number if we didn't have these custom number if we didn't have these custom agents checking through Macroscope agents checking through Macroscope agents checking through Macroscope whenever a PR comes in. The team was whenever a PR comes in. The team was whenever a PR comes in. The team was added one a few days ago because a lot added one a few days ago because a lot added one a few days ago because a lot of my slop was breaking the UI by making of my slop was breaking the UI by making of my slop was breaking the UI by making things like the tool tip patterns things like the tool tip patterns things like the tool tip patterns inconsistent across different elements. inconsistent across different elements. inconsistent across different elements. So now we have a UI consistency agent So now we have a UI consistency agent So now we have a UI consistency agent that will actually check during code that will actually check during code that will actually check during code review if any of the files for UI are review if any of the files for UI are review if any of the files for UI are touched to make sure we're being touched to make sure we're being touched to make sure we're being consistent with how we actually style consistent with how we actually style consistent with how we actually style things in our app. Now all pull requests things in our app. Now all pull requests things in our app. Now all pull requests get checked by the effect service get checked by the effect service get checked by the effect service conventions and the UI consistency conventions and the UI consistency conventions and the UI consistency runtime agents that we've introduced.

  4. runtime agents that we've introduced. runtime agents that we've introduced. And as you see, they're hilariously And as you see, they're hilariously And as you see, they're hilariously fast, taking 12 seconds and 21 seconds, fast, taking 12 seconds and 21 seconds, fast, taking 12 seconds and 21 seconds, respectively, roughly the same length as respectively, roughly the same length as respectively, roughly the same length as our also very fast CI. Once all the our also very fast CI. Once all the our also very fast CI. Once all the agents come back with their thoughts, agents come back with their thoughts, agents come back with their thoughts, the top level macroscope agent will give the top level macroscope agent will give the top level macroscope agent will give its approval rating if it thinks the PR its approval rating if it thinks the PR its approval rating if it thinks the PR is good to merge or not. And I'm going is good to merge or not. And I'm going is good to merge or not. And I'm going to be real with you guys, it's to be real with you guys, it's to be real with you guys, it's incredibly rare I bother merging PRs incredibly rare I bother merging PRs incredibly rare I bother merging PRs that haven't been approved by that haven't been approved by that haven't been approved by Macroscope. The slop is coming and it's Macroscope. The slop is coming and it's Macroscope. The slop is coming and it's trying to destroy your codebase. The trying to destroy your codebase. The trying to destroy your codebase. The best way to stop it is best way to stop it is best way to stop it is atvive.link/microscope. atvive.link/microscope. atvive.link/microscope. So, we have all the different ways we So, we have all the different ways we So, we have all the different ways we want to compare these models. Let's want to compare these models. Let's want to compare these models. Let's start with what I'm sure everyone is start with what I'm sure everyone is start with what I'm sure everyone is here for, science, because this is a here for, science, because this is a here for, science, because this is a science channel, right? Okay. Seriously science channel, right? Okay. Seriously science channel, right? Okay. Seriously though, this one of my favorite things though, this one of my favorite things though, this one of my favorite things to cover, not just cuz I think the 3D to cover, not just cuz I think the 3D to cover, not just cuz I think the 3D stuff is really cool, but there were stuff is really cool, but there were stuff is really cool, but there were some funny things in this launch. I'm some funny things in this launch. I'm some funny things in this launch. I'm going to start with the official Fable going to start with the official Fable going to start with the official Fable 5.1 launch notes because they were very 5.1 launch notes because they were very 5.1 launch notes because they were very excited to show off their progress in excited to show off their progress in excited to show off their progress in terminal bench science 0.1, a very terminal bench science 0.1, a very terminal bench science 0.1, a very difficult realworld science bench using difficult realworld science bench using difficult realworld science bench using tools to solve real world problems. And tools to solve real world problems. And tools to solve real world problems. And they had a huge jump here both in the they had a huge jump here both in the they had a huge jump here both in the efficiency for cost as well as the efficiency for cost as well as the efficiency for cost as well as the success rate overall where their best success rate overall where their best success rate overall where their best case previously was on high with Fable case previously was on high with Fable case previously was on high with Fable 5. It cost $34 and it got a 25%. Now 5.1 5. It cost $34 and it got a 25%. Now 5.1 5. It cost $34 and it got a 25%. Now 5.1 on XH high is even cheaper and got a on XH high is even cheaper and got a on XH high is even cheaper and got a 50%. Huge improvement. I see why they 50%. Huge improvement. I see why they 50%. Huge improvement. I see why they put this here. These numbers look put this here. These numbers look put this here. These numbers look awesome. At least they looked awesome awesome. At least they looked awesome awesome. At least they looked awesome until you went to the GBT6 Astra launch until you went to the GBT6 Astra launch until you went to the GBT6 Astra launch because Astra on low gets a 54.3 and because Astra on low gets a 54.3 and because Astra on low gets a 54.3 and only costs $11, putting low Astra higher only costs $11, putting low Astra higher only costs $11, putting low Astra higher than Max Fable 5.1 at under a third the

  5. than Max Fable 5.1 at under a third the than Max Fable 5.1 at under a third the price for the first benchmark that price for the first benchmark that price for the first benchmark that Anthropic had listed. What I'm trying to Anthropic had listed. What I'm trying to Anthropic had listed. What I'm trying to say here is if your job is science, you say here is if your job is science, you say here is if your job is science, you should probably go get a codec should probably go get a codec should probably go get a codec subscription. You can do some pretty subscription. You can do some pretty subscription. You can do some pretty crazy things with it, like mindblowing crazy things with it, like mindblowing crazy things with it, like mindblowing ones. Both of these models have been ones. Both of these models have been ones. Both of these models have been massive leaps in science. My assumption massive leaps in science. My assumption massive leaps in science. My assumption is there's either some new training data is there's either some new training data is there's either some new training data or some new RL tooling that has been or some new RL tooling that has been or some new RL tooling that has been given out or sold to the different labs given out or sold to the different labs given out or sold to the different labs and if it is something useful in post- and if it is something useful in post- and if it is something useful in post- trainining then open AAI got it and is trainining then open AAI got it and is trainining then open AAI got it and is using it heavily there and since OpenAI using it heavily there and since OpenAI using it heavily there and since OpenAI are the goats of post training they were are the goats of post training they were are the goats of post training they were able to apply that better. On the note able to apply that better. On the note able to apply that better. On the note of overlapping data, 3D rendering, both of overlapping data, 3D rendering, both of overlapping data, 3D rendering, both Fable 5.1 and GPD6 Astra made massive Fable 5.1 and GPD6 Astra made massive Fable 5.1 and GPD6 Astra made massive leaps in their 3D rendering leaps in their 3D rendering leaps in their 3D rendering capabilities. Astr was way bigger, capabilities. Astr was way bigger, capabilities. Astr was way bigger, though, like hilariously bigger. Where though, like hilariously bigger. Where though, like hilariously bigger. Where better to start than Fish Slop? This is better to start than Fish Slop? This is better to start than Fish Slop? This is Fable 5.1's version of fish slop. It is Fable 5.1's version of fish slop. It is Fable 5.1's version of fish slop. It is a 3D game where you have a submarine and a 3D game where you have a submarine and a 3D game where you have a submarine and some fish that you can feed and a mini some fish that you can feed and a mini some fish that you can feed and a mini economy as well as a little bit of economy as well as a little bit of economy as well as a little bit of combat. And this was the most impressive combat. And this was the most impressive combat. And this was the most impressive version of fish slop at the time because version of fish slop at the time because version of fish slop at the time because this version of fish slop had real this version of fish slop had real this version of fish slop had real models that were surprisingly good. Like models that were surprisingly good. Like models that were surprisingly good. Like the fish actually kind of looks like a the fish actually kind of looks like a the fish actually kind of looks like a fish has eyes placed in the right place.

  6. fish has eyes placed in the right place. fish has eyes placed in the right place. That is a lot harder than it sounds. That is a lot harder than it sounds. That is a lot harder than it sounds. There are like coral and rocks and There are like coral and rocks and There are like coral and rocks and lights that work. And most importantly lights that work. And most importantly lights that work. And most importantly though, the movement is actually very though, the movement is actually very though, the movement is actually very nice. It feels good to move around in nice. It feels good to move around in nice. It feels good to move around in this version of the game. like flying this version of the game. like flying this version of the game. like flying around or not really flying. Floating around or not really flying. Floating around or not really flying. Floating around with the sub feels awesome. It around with the sub feels awesome. It around with the sub feels awesome. It controls great. It plays great. It's controls great. It plays great. It's controls great. It plays great. It's surprisingly decent overall, but there's surprisingly decent overall, but there's surprisingly decent overall, but there's definitely room to improve. I say this definitely room to improve. I say this definitely room to improve. I say this with confidence because of the version with confidence because of the version with confidence because of the version Astromeade. Astromeade. Astromeade. I'd say this looks slightly better, I'd say this looks slightly better, I'd say this looks slightly better, just a little. It's a generational gap here. This is It's a generational gap here. This is nextG and the other version was not. nextG and the other version was not. nextG and the other version was not. What's the hit button to shoot? Okay, What's the hit button to shoot? Okay, What's the hit button to shoot? Okay, left click. left click. left click. There we go. Killed the guy. This does There we go. Killed the guy. This does There we go. Killed the guy. This does have its flaws. The movement doesn't have its flaws. The movement doesn't have its flaws. The movement doesn't feel anywhere near as good, and I had to feel anywhere near as good, and I had to feel anywhere near as good, and I had to like go back and forth to refine it a like go back and forth to refine it a like go back and forth to refine it a bit. The core loop isn't quite as well bit. The core loop isn't quite as well bit. The core loop isn't quite as well refined overall, and the performance was refined overall, and the performance was refined overall, and the performance was bad until I told to fix it. But it is bad until I told to fix it. But it is bad until I told to fix it. But it is stunning.

  7. stunning. stunning. Like it's actually decent looking to the Like it's actually decent looking to the Like it's actually decent looking to the point where I might have to drop slop point where I might have to drop slop point where I might have to drop slop from fish slop in the not too distant from fish slop in the not too distant from fish slop in the not too distant future. It is a massive leap. And if you future. It is a massive leap. And if you future. It is a massive leap. And if you think I'm just showing this and saying think I'm just showing this and saying think I'm just showing this and saying this now because I'm trying to glaze this now because I'm trying to glaze this now because I'm trying to glaze Astra, go watch my video on Kimmy K3 Astra, go watch my video on Kimmy K3 Astra, go watch my video on Kimmy K3 where I was for the first time ever where I was for the first time ever where I was for the first time ever genuinely really impressed with the 3D genuinely really impressed with the 3D genuinely really impressed with the 3D modeling capabilities of an LLM. This is modeling capabilities of an LLM. This is modeling capabilities of an LLM. This is a world of a difference. This is like a world of a difference. This is like a world of a difference. This is like multiple generational difference. So, multiple generational difference. So, multiple generational difference. So, while I definitely want to give this while I definitely want to give this while I definitely want to give this without any question to Astra, I without any question to Astra, I without any question to Astra, I actually think I have in my notes here a actually think I have in my notes here a actually think I have in my notes here a bunch of examples of crazy 3D stuff bunch of examples of crazy 3D stuff bunch of examples of crazy 3D stuff people did with the model. Here, Dra people did with the model. Here, Dra people did with the model. Here, Dra made a copy of the Amazon spheres that made a copy of the Amazon spheres that made a copy of the Amazon spheres that he was working in during his internship, he was working in during his internship, he was working in during his internship, all with Blender using Astra, which is all with Blender using Astra, which is all with Blender using Astra, which is just insane. or this demo home that just insane. or this demo home that just insane. or this demo home that Thomas built using Astra. All also with Thomas built using Astra. All also with Thomas built using Astra. All also with Blender where I can make a real Blender where I can make a real Blender where I can make a real environment with good furniture, a nice environment with good furniture, a nice environment with good furniture, a nice backdrop for the window and like it's backdrop for the window and like it's backdrop for the window and like it's good. It's like actually usable. You good. It's like actually usable. You good. It's like actually usable. You could use this to make real world mocks could use this to make real world mocks could use this to make real world mocks for real world 3D stuff. It's not a for real world 3D stuff. It's not a for real world 3D stuff. It's not a hypothetical anymore. This is in my hypothetical anymore. This is in my hypothetical anymore. This is in my opinion the equivalent of the jump from opinion the equivalent of the jump from opinion the equivalent of the jump from autocomplete agents to actually using autocomplete agents to actually using autocomplete agents to actually using your agent to complete tasks. And this your agent to complete tasks. And this your agent to complete tasks. And this all happened from Fable 5.1 to GBD6 all happened from Fable 5.1 to GBD6 all happened from Fable 5.1 to GBD6 Astra. The 3D capabilities in Astra are Astra. The 3D capabilities in Astra are Astra. The 3D capabilities in Astra are unbelievable. Which is why you might be unbelievable. Which is why you might be unbelievable. Which is why you might be confused when I move the arrow over here confused when I move the arrow over here confused when I move the arrow over here to gamedev for a sec. The reason I'm to gamedev for a sec. The reason I'm to gamedev for a sec. The reason I'm doing this is that as much as I am truly

  8. doing this is that as much as I am truly doing this is that as much as I am truly genuinely blown away with the 3D genuinely blown away with the 3D genuinely blown away with the 3D rendering capabilities of Astra. It is a rendering capabilities of Astra. It is a rendering capabilities of Astra. It is a like generational gap. It's the biggest like generational gap. It's the biggest like generational gap. It's the biggest gap of anything here between the models. gap of anything here between the models. gap of anything here between the models. 3D with Astra is comically better than 3D with Astra is comically better than 3D with Astra is comically better than 3D with Fable until you start 3D with Fable until you start 3D with Fable until you start interacting with it. And this is where interacting with it. And this is where interacting with it. And this is where Fable still just absolutely moss. Fable Fable still just absolutely moss. Fable Fable still just absolutely moss. Fable is so much better at interaction in is so much better at interaction in is so much better at interaction in general, at things like handling general, at things like handling general, at things like handling animations moving the right way when animations moving the right way when animations moving the right way when your cursor goes in a certain place or your cursor goes in a certain place or your cursor goes in a certain place or making the camera move the right amount making the camera move the right amount making the camera move the right amount when you move your mouse in a 3D world. when you move your mouse in a 3D world. when you move your mouse in a 3D world. For my silly 3D demos that I've been For my silly 3D demos that I've been For my silly 3D demos that I've been working on, I find that the best way to working on, I find that the best way to working on, I find that the best way to build them is to make the first build them is to make the first build them is to make the first prototype with Astra to get everything prototype with Astra to get everything prototype with Astra to get everything roughly looking and feeling how you roughly looking and feeling how you roughly looking and feeling how you wantish and then have Fable come in and wantish and then have Fable come in and wantish and then have Fable come in and clean it up and do all of the like clean it up and do all of the like clean it up and do all of the like detailoriented work. Astra, in my detailoriented work. Astra, in my detailoriented work. Astra, in my experience, does not handle the delicacy experience, does not handle the delicacy experience, does not handle the delicacy of that type of stuff anywhere near as of that type of stuff anywhere near as of that type of stuff anywhere near as well. It really does feel like a well. It really does feel like a well. It really does feel like a sledgehammer. Babel can apply things sledgehammer. Babel can apply things sledgehammer. Babel can apply things more gently in a way that makes things more gently in a way that makes things more gently in a way that makes things that work better. Back to non-code that work better. Back to non-code that work better. Back to non-code though, because there are layers to this though, because there are layers to this though, because there are layers to this one. Computer use. Realistically one. Computer use. Realistically one. Computer use. Realistically speaking, a lot of these categories are speaking, a lot of these categories are speaking, a lot of these categories are going to see a sway one way or the other going to see a sway one way or the other going to see a sway one way or the other where like certain things are better where like certain things are better where like certain things are better with Fable, certain things are better with Fable, certain things are better with Fable, certain things are better with Astra. The biggest gaps by far are with Astra. The biggest gaps by far are with Astra. The biggest gaps by far are the 3D rendering where Astra clears and the 3D rendering where Astra clears and the 3D rendering where Astra clears and computer use where Astra also clears. It computer use where Astra also clears. It computer use where Astra also clears. It is so much faster than 5.6 was for

  9. is so much faster than 5.6 was for is so much faster than 5.6 was for computer use. It's crazy. It just flies computer use. It's crazy. It just flies computer use. It's crazy. It just flies through things on your machine. Computer through things on your machine. Computer through things on your machine. Computer use of 56 soul was a big enough jump use of 56 soul was a big enough jump use of 56 soul was a big enough jump that I started to actually use it that I started to actually use it that I started to actually use it dayto-day for various different things. dayto-day for various different things. dayto-day for various different things. With GBD6 Astra, I've been using it so With GBD6 Astra, I've been using it so With GBD6 Astra, I've been using it so much that I got another Mac Mini just to much that I got another Mac Mini just to much that I got another Mac Mini just to let it run with computer use 247 to do let it run with computer use 247 to do let it run with computer use 247 to do all sorts of different tasks. It's so all sorts of different tasks. It's so all sorts of different tasks. It's so good at realworld computer use. It can good at realworld computer use. It can good at realworld computer use. It can do things much faster. understands them do things much faster. understands them do things much faster. understands them much better and the benchmarks don't in much better and the benchmarks don't in much better and the benchmarks don't in my opinion accurately measure the gap my opinion accurately measure the gap my opinion accurately measure the gap here. And the gap isn't just the model here. And the gap isn't just the model here. And the gap isn't just the model either. A lot of the improvements have either. A lot of the improvements have either. A lot of the improvements have been through the changes they've been been through the changes they've been been through the changes they've been making to codeex, especially on Mac OS making to codeex, especially on Mac OS making to codeex, especially on Mac OS to make the computer use spin up faster, to make the computer use spin up faster, to make the computer use spin up faster, get context more easily, move around get context more easily, move around get context more easily, move around your computer faster, and do things in your computer faster, and do things in your computer faster, and do things in the background better. 5.6 Soul would do the background better. 5.6 Soul would do the background better. 5.6 Soul would do tedious things I didn't feel like doing tedious things I didn't feel like doing tedious things I didn't feel like doing for me. Astra can often do tasks faster for me. Astra can often do tasks faster for me. Astra can often do tasks faster than I would have, which is very than I would have, which is very than I would have, which is very convenient because I'm currently down a convenient because I'm currently down a convenient because I'm currently down a hand. Astro computer use is so far hand. Astro computer use is so far hand. Astro computer use is so far ahead, it's kind of hilarious. Fable can ahead, it's kind of hilarious. Fable can ahead, it's kind of hilarious. Fable can do it decently when given the right do it decently when given the right do it decently when given the right tools, but it's just not even comparable tools, but it's just not even comparable tools, but it's just not even comparable to computer use. Funny enough, the weird to computer use. Funny enough, the weird to computer use. Funny enough, the weird misses that I'm used to OpenAI models misses that I'm used to OpenAI models misses that I'm used to OpenAI models having for code, it feels like Fable has having for code, it feels like Fable has having for code, it feels like Fable has those when it's doing computer use.

  10. those when it's doing computer use. those when it's doing computer use. Speaking of just getting it, let's talk Speaking of just getting it, let's talk Speaking of just getting it, let's talk about copyrightiting for a sec. Both of about copyrightiting for a sec. Both of about copyrightiting for a sec. Both of these models are massive leaps in the these models are massive leaps in the these models are massive leaps in the quality of pros that they spit out. If quality of pros that they spit out. If quality of pros that they spit out. If you have no Agent MD or Claude MD and you have no Agent MD or Claude MD and you have no Agent MD or Claude MD and you send the same prompt to Fable 5 and you send the same prompt to Fable 5 and you send the same prompt to Fable 5 and then Fable 5.1, the output from 5.1 is then Fable 5.1, the output from 5.1 is then Fable 5.1, the output from 5.1 is way more readable. If you do the same way more readable. If you do the same way more readable. If you do the same with 5.6 Soul and Astra, Astra is way with 5.6 Soul and Astra, Astra is way with 5.6 Soul and Astra, Astra is way more readable. Both of these models are more readable. Both of these models are more readable. Both of these models are huge jumps in how much less painful it huge jumps in how much less painful it huge jumps in how much less painful it is to read the text they put out. They is to read the text they put out. They is to read the text they put out. They almost have like unslop baked in almost have like unslop baked in almost have like unslop baked in finally, which is great. It's a finally, which is great. It's a finally, which is great. It's a meaningful improvement and it means I meaningful improvement and it means I meaningful improvement and it means I don't hate the outputs that I'm reading don't hate the outputs that I'm reading don't hate the outputs that I'm reading anywhere near as much. It's awesome. anywhere near as much. It's awesome. anywhere near as much. It's awesome. Which one's better at copy? I got the Which one's better at copy? I got the Which one's better at copy? I got the hot take that they both still suck and hot take that they both still suck and hot take that they both still suck and nothing has topped Kimmy K2. Not even nothing has topped Kimmy K2. Not even nothing has topped Kimmy K2. Not even K3, not even K2.5. The original K2 K3, not even K2.5. The original K2 K3, not even K2.5. The original K2 non-thinking model still writes the best non-thinking model still writes the best non-thinking model still writes the best of any model I've used. But it's also of any model I've used. But it's also of any model I've used. But it's also stupid as hell and very quick to be stupid as hell and very quick to be stupid as hell and very quick to be incredibly rude. It's not a model you incredibly rude. It's not a model you incredibly rude. It's not a model you should actually use for much. But 5.1 is should actually use for much. But 5.1 is should actually use for much. But 5.1 is a huge improvement here. Astra is a huge improvement here. Astra is a huge improvement here. Astra is notably better, though. I've had Astra notably better, though. I've had Astra notably better, though. I've had Astra come in and make suggestions for come in and make suggestions for come in and make suggestions for cleaning up copy from Fable. Astra is cleaning up copy from Fable. Astra is cleaning up copy from Fable. Astra is much better at recognizing the copy much better at recognizing the copy much better at recognizing the copy from Fable, but its proposals still just from Fable, but its proposals still just from Fable, but its proposals still just aren't as good as I would like. I still aren't as good as I would like. I still aren't as good as I would like. I still find myself rewriting most of the copy find myself rewriting most of the copy find myself rewriting most of the copy that these models create for my web that these models create for my web that these models create for my web pages. And while I do think Astra's pages. And while I do think Astra's pages. And while I do think Astra's copywriting in general, like pros and copywriting in general, like pros and copywriting in general, like pros and readability is slightly better than readability is slightly better than readability is slightly better than Fable, it does have one really, really, Fable, it does have one really, really, Fable, it does have one really, really, really annoying edge, which is really annoying edge, which is really annoying edge, which is that it loves to stuff all cap subtitles

  11. that it loves to stuff all cap subtitles that it loves to stuff all cap subtitles into everything it builds in UI. It just into everything it builds in UI. It just into everything it builds in UI. It just throws these subtitles everywhere and it throws these subtitles everywhere and it throws these subtitles everywhere and it is the worst. Possibly the silliest is the worst. Possibly the silliest is the worst. Possibly the silliest place to see this failure is the fish place to see this failure is the fish place to see this failure is the fish slop 2D version that I had it build. I slop 2D version that I had it build. I slop 2D version that I had it build. I would ask you to count the unnecessary would ask you to count the unnecessary would ask you to count the unnecessary subtitles on this page, but it would not subtitles on this page, but it would not subtitles on this page, but it would not be worth your time. Before I had it be worth your time. Before I had it be worth your time. Before I had it clean up a bit of it, there were 21 21 clean up a bit of it, there were 21 21 clean up a bit of it, there were 21 21 unnecessary subtitles. It's just spammed unnecessary subtitles. It's just spammed unnecessary subtitles. It's just spammed everywhere. A little tank, a lot of everywhere. A little tank, a lot of everywhere. A little tank, a lot of life. Coral Coast. Three little lives, life. Coral Coast. Three little lives, life. Coral Coast. Three little lives, all yours to look after. All systems go. all yours to look after. All systems go. all yours to look after. All systems go. A tiny world worth looking after. The A tiny world worth looking after. The A tiny world worth looking after. The first version, it also had an online first version, it also had an online first version, it also had an online badge at the bottom that said online and badge at the bottom that said online and badge at the bottom that said online and ready. I didn't even see that one. Chad ready. I didn't even see that one. Chad ready. I didn't even see that one. Chad was spamming captain, you're home. It's was spamming captain, you're home. It's was spamming captain, you're home. It's in It's so bad. Like, I I don't even in It's so bad. Like, I I don't even in It's so bad. Like, I I don't even care that Astra is better at copy care that Astra is better at copy care that Astra is better at copy because it's trying to show it off by because it's trying to show it off by because it's trying to show it off by being shitty at copy everywhere being shitty at copy everywhere being shitty at copy everywhere constantly. It's It's unacceptably bad. constantly. It's It's unacceptably bad. constantly. It's It's unacceptably bad. The more I stare at this, the more I The more I stare at this, the more I The more I stare at this, the more I hate this model. I I just, for what it's hate this model. I I just, for what it's hate this model. I I just, for what it's worth, straight up do not trust Astra worth, straight up do not trust Astra worth, straight up do not trust Astra with UI anything at this point. Well, we with UI anything at this point. Well, we with UI anything at this point. Well, we should move on from this ocean of should move on from this ocean of should move on from this ocean of possibility to other categories because possibility to other categories because possibility to other categories because while I do give Astra a slight lead in while I do give Astra a slight lead in while I do give Astra a slight lead in copyrightiting capability, the way it copyrightiting capability, the way it copyrightiting capability, the way it spams it pisses me off too much to spams it pisses me off too much to spams it pisses me off too much to frankly want to ever use it for copy.

  12. frankly want to ever use it for copy. frankly want to ever use it for copy. Last thing in the non-code section, then Last thing in the non-code section, then Last thing in the non-code section, then we get to what you guys are all actually we get to what you guys are all actually we get to what you guys are all actually here for. code agents and cost audio and here for. code agents and cost audio and here for. code agents and cost audio and video work. I'll be real with you guys. video work. I'll be real with you guys. video work. I'll be real with you guys. From my experience, both have From my experience, both have From my experience, both have disappointed me here. I am admittedly disappointed me here. I am admittedly disappointed me here. I am admittedly very picky, like annoyingly picky about very picky, like annoyingly picky about very picky, like annoyingly picky about these things. So, I should not be the these things. So, I should not be the these things. So, I should not be the only voice you hear when you listen to only voice you hear when you listen to only voice you hear when you listen to opinions on audio and video work with AI opinions on audio and video work with AI opinions on audio and video work with AI and AI agents. A lot of people have told and AI agents. A lot of people have told and AI agents. A lot of people have told me that they had Fable or Astraedit me that they had Fable or Astraedit me that they had Fable or Astraedit videos. Every time I watch the videos, videos. Every time I watch the videos, videos. Every time I watch the videos, they suck. I do have one exception here, they suck. I do have one exception here, they suck. I do have one exception here, though. This is a demo that Ben Davis though. This is a demo that Ben Davis though. This is a demo that Ben Davis did showing how he uses computer use did showing how he uses computer use did showing how he uses computer use with Astra, not to edit the video, but with Astra, not to edit the video, but with Astra, not to edit the video, but to get everything set up to start the to get everything set up to start the to get everything set up to start the edit for the video. You can think of edit for the video. You can think of edit for the video. You can think of this in like code terms back pre-AI. this in like code terms back pre-AI. this in like code terms back pre-AI. Imagine if AI was so good at like using Imagine if AI was so good at like using Imagine if AI was so good at like using VS Code and understanding your codebase VS Code and understanding your codebase VS Code and understanding your codebase and like what you needed that when you and like what you needed that when you and like what you needed that when you were about to start working on a thing, were about to start working on a thing, were about to start working on a thing, it could open up VS Code. It could open it could open up VS Code. It could open it could open up VS Code. It could open up all the files you specifically need up all the files you specifically need up all the files you specifically need to edit. It could open up your terminal to edit. It could open up your terminal to edit. It could open up your terminal in the right place, GitHub in the right in the right place, GitHub in the right in the right place, GitHub in the right place, and a dev environment showing place, and a dev environment showing place, and a dev environment showing exactly what you're about to work on.

  13. exactly what you're about to work on. exactly what you're about to work on. So, it sets up everything you need to So, it sets up everything you need to So, it sets up everything you need to start working. It doesn't do the work. start working. It doesn't do the work. start working. It doesn't do the work. It just prepares you to do the work. It just prepares you to do the work. It just prepares you to do the work. It's like getting the room ready and It's like getting the room ready and It's like getting the room ready and cleared with everything you need. Benis cleared with everything you need. Benis cleared with everything you need. Benis had a surprising amount of success had a surprising amount of success had a surprising amount of success getting Astra to do this type of work, getting Astra to do this type of work, getting Astra to do this type of work, to get it to actually set up his editor to get it to actually set up his editor to get it to actually set up his editor in Final Cut to be ready to go to edit in Final Cut to be ready to go to edit in Final Cut to be ready to go to edit video. If Ben says it's good for getting video. If Ben says it's good for getting video. If Ben says it's good for getting his editor prepared, I'll take his word his editor prepared, I'll take his word his editor prepared, I'll take his word for it. I have not had the pleasure of for it. I have not had the pleasure of for it. I have not had the pleasure of editing a video in a while. I do really editing a video in a while. I do really editing a video in a while. I do really miss it. I cannot wait till I can edit a miss it. I cannot wait till I can edit a miss it. I cannot wait till I can edit a video again once my hand is back. Soon video again once my hand is back. Soon video again once my hand is back. Soon TM, but this very promising, its TM, but this very promising, its TM, but this very promising, its understanding of computer use means it understanding of computer use means it understanding of computer use means it is more useful for real professional is more useful for real professional is more useful for real professional audio and video work. July 2020 when audio and video work. July 2020 when audio and video work. July 2020 when nobody knew who I was back in the day. nobody knew who I was back in the day. nobody knew who I was back in the day. Ready for a crazy throwback, guys? This Ready for a crazy throwback, guys? This Ready for a crazy throwback, guys? This is a GPT3 demo that an OpenAI employee is a GPT3 demo that an OpenAI employee is a GPT3 demo that an OpenAI employee made showing how you can have GPT3 make made showing how you can have GPT3 make made showing how you can have GPT3 make a functioning React app. Crazy. Describe a functioning React app. Crazy. Describe a functioning React app. Crazy. Describe your app. A button that says add $3 and your app. A button that says add $3 and your app. A button that says add $3 and a button that says withdraw $5, then a button that says withdraw $5, then a button that says withdraw $5, then show me my balance. So complex. Look at show me my balance. So complex. Look at show me my balance. So complex. Look at that. It made the React code and it that. It made the React code and it that. It made the React code and it works. It didn't even use hooks. It used works. It didn't even use hooks. It used works. It didn't even use hooks. It used class components. This was a huge deal.

  14. class components. This was a huge deal. class components. This was a huge deal. I know that my joke six years later I know that my joke six years later I know that my joke six years later doesn't hit quite the same. I think this doesn't hit quite the same. I think this doesn't hit quite the same. I think this might be able to pass an interview. But might be able to pass an interview. But might be able to pass an interview. But like at the time this seemed like at the time this seemed like at the time this seemed unbelievable, but also like if you're a unbelievable, but also like if you're a unbelievable, but also like if you're a real engineer, you look at this and real engineer, you look at this and real engineer, you look at this and you're just like, "Yeah, I could do this you're just like, "Yeah, I could do this you're just like, "Yeah, I could do this in my first week of class." The in my first week of class." The in my first week of class." The hilarious way this looks to us in terms hilarious way this looks to us in terms hilarious way this looks to us in terms of capability, we're like, "Yeah, that's of capability, we're like, "Yeah, that's of capability, we're like, "Yeah, that's cool. That's not actually useful for cool. That's not actually useful for cool. That's not actually useful for real world work." This is how I feel real world work." This is how I feel real world work." This is how I feel when people post videos that are edited when people post videos that are edited when people post videos that are edited with AI right now. If you let AI edit with AI right now. If you let AI edit with AI right now. If you let AI edit your content for you, if you let AI do your content for you, if you let AI do your content for you, if you let AI do your audio for you, if you let AI manage your audio for you, if you let AI manage your audio for you, if you let AI manage your AV pipeline for you in these ways, your AV pipeline for you in these ways, your AV pipeline for you in these ways, you are doing the equivalent of shipping you are doing the equivalent of shipping you are doing the equivalent of shipping an app using GPT3. We are that far an app using GPT3. We are that far an app using GPT3. We are that far behind right now. Do I think we'll have behind right now. Do I think we'll have behind right now. Do I think we'll have an astro moment for video editing tools? an astro moment for video editing tools? an astro moment for video editing tools? Perhaps we've made real improvements Perhaps we've made real improvements Perhaps we've made real improvements here, but it's so far from usable right here, but it's so far from usable right here, but it's so far from usable right now that I I'll be frank. I just don't now that I I'll be frank. I just don't now that I I'll be frank. I just don't get how people think that they can get how people think that they can get how people think that they can actually automate video editing with AI actually automate video editing with AI actually automate video editing with AI right now. Not even close. Not even kind right now. Not even close. Not even kind right now. Not even close. Not even kind of there yet. So yeah, I don't think of there yet. So yeah, I don't think of there yet. So yeah, I don't think either are good enough that it's even either are good enough that it's even either are good enough that it's even worth ranking here. But due to Codex's worth ranking here. But due to Codex's worth ranking here. But due to Codex's incredible computer use, I think Astra incredible computer use, I think Astra incredible computer use, I think Astra just gets a free win here. So this whole just gets a free win here. So this whole just gets a free win here. So this whole non-code section, Astra wins. The only non-code section, Astra wins. The only non-code section, Astra wins. The only parts that Astra wins massively though parts that Astra wins massively though parts that Astra wins massively though are the 3D rendering and the computer are the 3D rendering and the computer are the 3D rendering and the computer use. Although I do guess the science use. Although I do guess the science use. Although I do guess the science progress is pretty meaningful, too. So progress is pretty meaningful, too. So progress is pretty meaningful, too. So like these three Astromogs much harder like these three Astromogs much harder like these three Astromogs much harder in 3D rendering than the others, but it in 3D rendering than the others, but it in 3D rendering than the others, but it does meaningfully improve in all of does meaningfully improve in all of does meaningfully improve in all of them. So now we are done with this them. So now we are done with this them. So now we are done with this section front end. I'm sure you guys can

  15. section front end. I'm sure you guys can section front end. I'm sure you guys can guess where things landed here by now. I guess where things landed here by now. I guess where things landed here by now. I do want to make a few things clear do want to make a few things clear do want to make a few things clear before I go on the utter brutal roast before I go on the utter brutal roast before I go on the utter brutal roast session I'm about to do. GBD6 Astra has session I'm about to do. GBD6 Astra has session I'm about to do. GBD6 Astra has made massive improvements overall with made massive improvements overall with made massive improvements overall with front-end stuff. It is meaningfully front-end stuff. It is meaningfully front-end stuff. It is meaningfully better. It follows instructions better. better. It follows instructions better. better. It follows instructions better. It can make landing pages more It can make landing pages more It can make landing pages more effectively and less cringey. And it's effectively and less cringey. And it's effectively and less cringey. And it's so much smarter and better at so much smarter and better at so much smarter and better at understanding things in general that you understanding things in general that you understanding things in general that you can force a good design out of Astro can force a good design out of Astro can force a good design out of Astro with enough effort and iteration. For with enough effort and iteration. For with enough effort and iteration. For reference, here is GPT 5.6 with some reference, here is GPT 5.6 with some reference, here is GPT 5.6 with some basic designs on Witch AI by Dra. It's a basic designs on Witch AI by Dra. It's a basic designs on Witch AI by Dra. It's a nice little demo here. This is 5.6 Soul. nice little demo here. This is 5.6 Soul. nice little demo here. This is 5.6 Soul. Right now, this is fine. The site's Right now, this is fine. The site's Right now, this is fine. The site's funny if like a little bit laggy which I funny if like a little bit laggy which I funny if like a little bit laggy which I don't love obviously but it has all don't love obviously but it has all don't love obviously but it has all these subtitles notes that think beside these subtitles notes that think beside these subtitles notes that think beside you loved by 18,000 curious minds an you loved by 18,000 curious minds an you loved by 18,000 curious minds an unnecessary m dash built for unnecessary m dash built for unnecessary m dash built for remembering. It's not the worst, but it's also not It's not the worst, but it's also not the best. And some of these are just the best. And some of these are just the best. And some of these are just such boring Tailwind templatey stuff.

  16. such boring Tailwind templatey stuff. such boring Tailwind templatey stuff. Pretty cringe. This one's okay. This Pretty cringe. This one's okay. This Pretty cringe. This one's okay. This one's awful. And this one is another one's awful. And this one is another one's awful. And this one is another boring tailwind template. So that's 56 boring tailwind template. So that's 56 boring tailwind template. So that's 56 soul. We bump over to GBD6 Astra. We can soul. We bump over to GBD6 Astra. We can soul. We bump over to GBD6 Astra. We can see how much better or worse it is here. see how much better or worse it is here. see how much better or worse it is here. Still too many of these subtitles. A Still too many of these subtitles. A Still too many of these subtitles. A second brain. A lighter mind. Free to second brain. A lighter mind. Free to second brain. A lighter mind. Free to start. Yours to make your own. I hate start. Yours to make your own. I hate start. Yours to make your own. I hate these so much. But open my space. Your these so much. But open my space. Your these so much. But open my space. Your mind a little more organized. Six little mind a little more organized. Six little mind a little more organized. Six little pieces of your mind. A little space just pieces of your mind. A little space just pieces of your mind. A little space just for you. The word little appears on this for you. The word little appears on this for you. The word little appears on this page 15 times. All of your little page 15 times. All of your little page 15 times. All of your little things. Still looks better. It does like things. Still looks better. It does like things. Still looks better. It does like this little arrow thing. That's kind of this little arrow thing. That's kind of this little arrow thing. That's kind of cute. cute. cute. It's fine. It's meaningfully better. It's fine. It's meaningfully better. It's fine. It's meaningfully better. That had a weird font pop in, but those That had a weird font pop in, but those That had a weird font pop in, but those are easy to fix. These little animations are easy to fix. These little animations are easy to fix. These little animations are nice. Looks fine. This one I hate. Just like a personal This one I hate. Just like a personal hate, but I hate it. Especially with the hate, but I hate it. Especially with the hate, but I hate it. Especially with the subtitles. This one's actually kind of subtitles. This one's actually kind of subtitles. This one's actually kind of cute. I don't know why that got cut off.

  17. cute. I don't know why that got cut off. cute. I don't know why that got cut off. Like there's a lot of cut offs here that Like there's a lot of cut offs here that Like there's a lot of cut offs here that shouldn't be happening that are but shouldn't be happening that are but shouldn't be happening that are but where it's going for here I I see what where it's going for here I I see what where it's going for here I I see what it wanted to do and it's not the worst it wanted to do and it's not the worst it wanted to do and it's not the worst like decent starting point and this is like decent starting point and this is like decent starting point and this is very boring and old school. So yeah very boring and old school. So yeah very boring and old school. So yeah improvements it is slightly better to improvements it is slightly better to improvements it is slightly better to meaningfully better in all of these meaningfully better in all of these meaningfully better in all of these categories. I can turn on the claude categories. I can turn on the claude categories. I can turn on the claude code design skill from anthropic and it code design skill from anthropic and it code design skill from anthropic and it does get slightly better. It's just a does get slightly better. It's just a does get slightly better. It's just a little too quick to experiment with little too quick to experiment with little too quick to experiment with fonts and things, but like these are fonts and things, but like these are fonts and things, but like these are passable. These are starting points you passable. These are starting points you passable. These are starting points you could reasonably use for something. But could reasonably use for something. But could reasonably use for something. But I need to be realistic with you guys. I need to be realistic with you guys. I need to be realistic with you guys. Moving the windows, I don't block this Moving the windows, I don't block this Moving the windows, I don't block this one as much. one as much. one as much. Do you see how nice that looks? The Do you see how nice that looks? The Do you see how nice that looks? The animation of all these paths coming in. animation of all these paths coming in. animation of all these paths coming in. The structure of the page is great. None The structure of the page is great. None The structure of the page is great. None of those unnecessary subtitles. of those unnecessary subtitles. of those unnecessary subtitles. It's a much much better starting point It's a much much better starting point It's a much much better starting point for a one shot. It's really good at for a one shot. It's really good at for a one shot. It's really good at these types of animations and having these types of animations and having these types of animations and having like a distinct style. This one is so like a distinct style. This one is so like a distinct style. This one is so much better than previous models. Like much better than previous models. Like much better than previous models. Like if we switched just to Fable 5 or a if we switched just to Fable 5 or a if we switched just to Fable 5 or a similar design, similar design, similar design, gross. 5.1 same mock, same design.

  18. gross. 5.1 same mock, same design. gross. 5.1 same mock, same design. Actually genuinely compelling. Like if I Actually genuinely compelling. Like if I Actually genuinely compelling. Like if I had seen this on the internet, I would had seen this on the internet, I would had seen this on the internet, I would never have guessed this was AI never have guessed this was AI never have guessed this was AI generated, much less a one shot. This generated, much less a one shot. This generated, much less a one shot. This screams like custommade to me. It's screams like custommade to me. It's screams like custommade to me. It's really good. That said, I have had my really good. That said, I have had my really good. That said, I have had my problems with both with design, and I problems with both with design, and I problems with both with design, and I have found that both actually kind of have found that both actually kind of have found that both actually kind of pissed me off. I've been quite annoyed pissed me off. I've been quite annoyed pissed me off. I've been quite annoyed at how much both have pissed me off, at how much both have pissed me off, at how much both have pissed me off, especially recently as I've been trying especially recently as I've been trying especially recently as I've been trying to iterate more on some existing design to iterate more on some existing design to iterate more on some existing design work. Fable takes the cake here easy, work. Fable takes the cake here easy, work. Fable takes the cake here easy, significantly better, but both still significantly better, but both still significantly better, but both still frustrate me constantly, and good frustrate me constantly, and good frustrate me constantly, and good designs still require you sit there with designs still require you sit there with designs still require you sit there with a hammer and beat the out of a hammer and beat the out of a hammer and beat the out of the model. So, while there is a gap the model. So, while there is a gap the model. So, while there is a gap here, I think the gap between a two and here, I think the gap between a two and here, I think the gap between a two and a five out of 10 is much less notable a five out of 10 is much less notable a five out of 10 is much less notable than the gap between a five and a 9 out than the gap between a five and a 9 out than the gap between a five and a 9 out of 10. And when it comes to like 3D of 10. And when it comes to like 3D of 10. And when it comes to like 3D capability, I would say Astra is a nine capability, I would say Astra is a nine capability, I would say Astra is a nine and Fable's a five. But with front end, and Fable's a five. But with front end, and Fable's a five. But with front end, I would say neither of them are more I would say neither of them are more I would say neither of them are more than a five out of 10. But at least than a five out of 10. But at least than a five out of 10. But at least Fable can make something that looks Fable can make something that looks Fable can make something that looks decent without having to like actually decent without having to like actually decent without having to like actually beat the out of it. Astra needs a beat the out of it. Astra needs a beat the out of it. Astra needs a lot of help if you want a good design lot of help if you want a good design lot of help if you want a good design out of it. So, next full stack. I'm out of it. So, next full stack. I'm out of it. So, next full stack. I'm going to be so real here. If there is a going to be so real here. If there is a going to be so real here. If there is a difference in how well these models difference in how well these models difference in how well these models grasp your stack, I'm impressed because grasp your stack, I'm impressed because grasp your stack, I'm impressed because for me, they both get it. We are finally for me, they both get it. We are finally for me, they both get it. We are finally now at the point where both Frontier now at the point where both Frontier now at the point where both Frontier models from both Frontier Labs can look models from both Frontier Labs can look models from both Frontier Labs can look at a code base and the backended front at a code base and the backended front at a code base and the backended front end and clients and servers and how they end and clients and servers and how they end and clients and servers and how they relate and make good, reasonable relate and make good, reasonable relate and make good, reasonable decisions operating against it. That was decisions operating against it. That was decisions operating against it. That was not the case before. Fable was the first not the case before. Fable was the first not the case before. Fable was the first model that could do that. 5.6 Soul could model that could do that. 5.6 Soul could model that could do that. 5.6 Soul could act like it was doing that, but often

  19. act like it was doing that, but often act like it was doing that, but often just missed things. didn't get and fully just missed things. didn't get and fully just missed things. didn't get and fully like understand the end to end story. like understand the end to end story. like understand the end to end story. They both do now. I would say it feels They both do now. I would say it feels They both do now. I would say it feels like Fable has better intuition with it. like Fable has better intuition with it. like Fable has better intuition with it. Like it understands the consequences of Like it understands the consequences of Like it understands the consequences of edges by default slightly better, but edges by default slightly better, but edges by default slightly better, but Astra is more willing to just go at the Astra is more willing to just go at the Astra is more willing to just go at the problem forever and test every single problem forever and test every single problem forever and test every single edge and validate every single edge and validate every single edge and validate every single assumption and force itself to come to assumption and force itself to come to assumption and force itself to come to the right answer. And I've seen this the right answer. And I've seen this the right answer. And I've seen this type of thing happen so many times where type of thing happen so many times where type of thing happen so many times where I ask Fable and Astro to do the same I ask Fable and Astro to do the same I ask Fable and Astro to do the same task on a big open- source full stack task on a big open- source full stack task on a big open- source full stack project something like T3 Code. Fable project something like T3 Code. Fable project something like T3 Code. Fable looks at the code, comes to some looks at the code, comes to some looks at the code, comes to some conclusions, has a few concerns, and conclusions, has a few concerns, and conclusions, has a few concerns, and then addresses them through its design. then addresses them through its design. then addresses them through its design. Astra has all of the things after it Astra has all of the things after it Astra has all of the things after it finds them, decides there are three finds them, decides there are three finds them, decides there are three valid options, stress tests all of them, valid options, stress tests all of them, valid options, stress tests all of them, eventually finds the thing Fable already eventually finds the thing Fable already eventually finds the thing Fable already knew from the start, and then makes a knew from the start, and then makes a knew from the start, and then makes a similarish solution addressing the same similarish solution addressing the same similarish solution addressing the same things. So in the instances where Fable things. So in the instances where Fable things. So in the instances where Fable can find and understand things, it is a can find and understand things, it is a can find and understand things, it is a nicer experience because it's so quick nicer experience because it's so quick nicer experience because it's so quick to work and actually apply changes to work and actually apply changes to work and actually apply changes because it builds its understanding more because it builds its understanding more because it builds its understanding more effectively because it almost feels like effectively because it almost feels like effectively because it almost feels like it understands better. Astra really has it understands better. Astra really has it understands better. Astra really has that groundhog day feeling to it where that groundhog day feeling to it where that groundhog day feeling to it where every time it starts it has lost track every time it starts it has lost track every time it starts it has lost track of everything in the codebase and it of everything in the codebase and it of everything in the codebase and it really is starting again. But because of really is starting again. But because of really is starting again. But because of that it will find things other models that it will find things other models that it will find things other models miss. Fable's coming in with the miss. Fable's coming in with the miss. Fable's coming in with the assumption that it knows and can figure assumption that it knows and can figure assumption that it knows and can figure out everything with it super fancy smart out everything with it super fancy smart out everything with it super fancy smart brain. Astra is the smartest model that brain. Astra is the smartest model that brain. Astra is the smartest model that acts like it's dumb. Astra quadruple

  20. acts like it's dumb. Astra quadruple acts like it's dumb. Astra quadruple confirms everything it's trying to do, confirms everything it's trying to do, confirms everything it's trying to do, especially in X high mode. Which means especially in X high mode. Which means especially in X high mode. Which means if the bug is outside of what Fable can if the bug is outside of what Fable can if the bug is outside of what Fable can grasp just from reading code and it's grasp just from reading code and it's grasp just from reading code and it's something deeper or harder to find, something deeper or harder to find, something deeper or harder to find, Astra will find it. It'll just take four Astra will find it. It'll just take four Astra will find it. It'll just take four times longer and burn way more tokens. times longer and burn way more tokens. times longer and burn way more tokens. So for me, it almost feels like a tie. So for me, it almost feels like a tie. So for me, it almost feels like a tie. But this is also going to be the start But this is also going to be the start But this is also going to be the start of a theme that we'll be touching on of a theme that we'll be touching on of a theme that we'll be touching on throughout. The theme is that Fable's throughout. The theme is that Fable's throughout. The theme is that Fable's performance is generally more steady, performance is generally more steady, performance is generally more steady, where Astra is much spikier, where Astra where Astra is much spikier, where Astra where Astra is much spikier, where Astra has moments that leave me in awe and has moments that leave me in awe and has moments that leave me in awe and moments that make me question why I'm moments that make me question why I'm moments that make me question why I'm paying OpenAI$600 plus dollars a month. paying OpenAI$600 plus dollars a month. paying OpenAI$600 plus dollars a month. We'll come back to that theme in a bit. We'll come back to that theme in a bit. We'll come back to that theme in a bit. But on the prior theme, which is Astra's But on the prior theme, which is Astra's But on the prior theme, which is Astra's relentlessness, its desire to grab a relentlessness, its desire to grab a relentlessness, its desire to grab a problem and strangle it. Giant project problem and strangle it. Giant project problem and strangle it. Giant project rewrites. Astra has made meaningful rewrites. Astra has made meaningful rewrites. Astra has made meaningful strides here. I have thrown it at some strides here. I have thrown it at some strides here. I have thrown it at some crazy and I'm blown away with it. crazy and I'm blown away with it. crazy and I'm blown away with it. It is mostly done with the rewrite of It is mostly done with the rewrite of It is mostly done with the rewrite of TypeScript in Rust. It's porting TSGO to TypeScript in Rust. It's porting TSGO to TypeScript in Rust. It's porting TSGO to Rust and it's making real progress. That Rust and it's making real progress. That Rust and it's making real progress. That actually reminds me how much am I actually reminds me how much am I actually reminds me how much am I my usage as I do all of this my usage as I do all of this my usage as I do all of this right now. Not too bad considering that right now. Not too bad considering that right now. Not too bad considering that I have that just like running in a death I have that just like running in a death I have that just like running in a death loop right now. It paused the goal loop right now. It paused the goal loop right now. It paused the goal again. You have my permission to do again. You have my permission to do again. You have my permission to do whatever you need to do. I have you on whatever you need to do. I have you on whatever you need to do. I have you on full access for a reason. Keep going. I full access for a reason. Keep going. I full access for a reason. Keep going. I am very deep in the TypeScript Rust am very deep in the TypeScript Rust am very deep in the TypeScript Rust Rewrite. With 56 soul, I was able to get Rewrite. With 56 soul, I was able to get Rewrite. With 56 soul, I was able to get about 30% test accuracy going against about 30% test accuracy going against about 30% test accuracy going against the giant test suite that they use for the giant test suite that they use for the giant test suite that they use for testing the TypeScript code bases. I was testing the TypeScript code bases. I was testing the TypeScript code bases. I was able to get over 80% with Astra. And it

  21. able to get over 80% with Astra. And it able to get over 80% with Astra. And it didn't even take that long. It was able didn't even take that long. It was able didn't even take that long. It was able to do that in I don't know, I would say to do that in I don't know, I would say to do that in I don't know, I would say about 3 days roughly, not even. It was about 3 days roughly, not even. It was about 3 days roughly, not even. It was able to jump from that 30% range to 80 able to jump from that 30% range to 80 able to jump from that 30% range to 80 plus. But then it stalled out super plus. But then it stalled out super plus. But then it stalled out super super hard at 82.6%. super hard at 82.6%. super hard at 82.6%. And I still don't fully understand why. And I still don't fully understand why. And I still don't fully understand why. I've been trying to go back and forth I've been trying to go back and forth I've been trying to go back and forth with it to figure it out, but it got with it to figure it out, but it got with it to figure it out, but it got comically further than anything else comically further than anything else comically further than anything else I've thrown at a task like this. And I I've thrown at a task like this. And I I've thrown at a task like this. And I want to be clear, I'm not in the loop on want to be clear, I'm not in the loop on want to be clear, I'm not in the loop on this one. I just set a goal and gave it this one. I just set a goal and gave it this one. I just set a goal and gave it 40 sub agents and said, "Do as you 40 sub agents and said, "Do as you 40 sub agents and said, "Do as you please." Remember those 40 sub agents, please." Remember those 40 sub agents, please." Remember those 40 sub agents, though, because we'll be talking about though, because we'll be talking about though, because we'll be talking about that in the agent section. I will also that in the agent section. I will also that in the agent section. I will also admit to a bias here, which is that when admit to a bias here, which is that when admit to a bias here, which is that when I'm doing early access testing with the I'm doing early access testing with the I'm doing early access testing with the new models with OpenAI, they don't new models with OpenAI, they don't new models with OpenAI, they don't heavily restrict my account. It does not heavily restrict my account. It does not heavily restrict my account. It does not count towards my usage, which is at count towards my usage, which is at count towards my usage, which is at sometimes annoying because it means I sometimes annoying because it means I sometimes annoying because it means I don't have a good feel for how quickly don't have a good feel for how quickly don't have a good feel for how quickly the new model will burn your account, the new model will burn your account, the new model will burn your account, but on the other hand is so useful for but on the other hand is so useful for but on the other hand is so useful for actually testing the capabilities of the actually testing the capabilities of the actually testing the capabilities of the model and pushing it to its absolute model and pushing it to its absolute model and pushing it to its absolute limits. I did 130 billion tokens with limits. I did 130 billion tokens with limits. I did 130 billion tokens with this model during the early access. That this model during the early access. That this model during the early access. That means I can yolo it at these types of means I can yolo it at these types of means I can yolo it at these types of things and not have to worry too much things and not have to worry too much things and not have to worry too much about the bill because I'm not paying about the bill because I'm not paying about the bill because I'm not paying it. I've never had free unlimited access it. I've never had free unlimited access it. I've never had free unlimited access to Fable before. Which means while I to Fable before. Which means while I to Fable before. Which means while I have seen and know the capabilities that have seen and know the capabilities that have seen and know the capabilities that Astra has here and have even shipped Astra has here and have even shipped Astra has here and have even shipped some of them like the T3 code mobile some of them like the T3 code mobile some of them like the T3 code mobile rewrite in Swift, I even have a T3 code rewrite in Swift, I even have a T3 code rewrite in Swift, I even have a T3 code build here with GPUi that I almost build here with GPUi that I almost build here with GPUi that I almost forgot about where I had it rebuild all

  22. forgot about where I had it rebuild all forgot about where I had it rebuild all of T3 code with GPUi, the Rust UI of T3 code with GPUi, the Rust UI of T3 code with GPUi, the Rust UI framework, just to see if it could and framework, just to see if it could and framework, just to see if it could and to test the capabilities, performance, to test the capabilities, performance, to test the capabilities, performance, and stuff like that out of curiosity. and stuff like that out of curiosity. and stuff like that out of curiosity. And it did it. It made it work great. And it did it. It made it work great. And it did it. It made it work great. But here's where one of the bigger But here's where one of the bigger But here's where one of the bigger catches with Astra comes in, and I've catches with Astra comes in, and I've catches with Astra comes in, and I've seen this in pretty much every one of seen this in pretty much every one of seen this in pretty much every one of these big rewrite attempts I have done. these big rewrite attempts I have done. these big rewrite attempts I have done. One of my old benches I used to do a lot One of my old benches I used to do a lot One of my old benches I used to do a lot in my videos was having the new models in my videos was having the new models in my videos was having the new models rewrite my legacy ping.gg codebase, the rewrite my legacy ping.gg codebase, the rewrite my legacy ping.gg codebase, the one that is used for my video collab one that is used for my video collab one that is used for my video collab tool that I got into Y Combinator with. tool that I got into Y Combinator with. tool that I got into Y Combinator with. I gave it the original codebase, told it I gave it the original codebase, told it I gave it the original codebase, told it to write a plan, and then I had it and to write a plan, and then I had it and to write a plan, and then I had it and Astra review each other's plans Astra review each other's plans Astra review each other's plans separately out of curiosity. But once it separately out of curiosity. But once it separately out of curiosity. But once it made his plan, I told it to make a made his plan, I told it to make a made his plan, I told it to make a branch for the work and commit the plan branch for the work and commit the plan branch for the work and commit the plan to it. Do that in a work tree. I had to to it. Do that in a work tree. I had to to it. Do that in a work tree. I had to do the commit so I could access it in do the commit so I could access it in do the commit so I could access it in other places. Then I told it to build other places. Then I told it to build other places. Then I told it to build the whole thing. 1 hour and 20 minutes the whole thing. 1 hour and 20 minutes the whole thing. 1 hour and 20 minutes later, it did. Since I hit yet another later, it did. Since I hit yet another later, it did. Since I hit yet another obnoxious case with codeex where since I obnoxious case with codeex where since I obnoxious case with codeex where since I was in work mode instead of in codeex was in work mode instead of in codeex was in work mode instead of in codeex mode, I can't actually open the mode, I can't actually open the mode, I can't actually open the terminal. I have to wait for this to do terminal. I have to wait for this to do terminal. I have to wait for this to do the things I want. So while I wait for the things I want. So while I wait for the things I want. So while I wait for that, I'm going to show you what ping that, I'm going to show you what ping that, I'm going to show you what ping normally looks like. This is the app normally looks like. This is the app normally looks like. This is the app that I use when I bring collaborators on that I use when I bring collaborators on that I use when I bring collaborators on for streams and all sorts of other for streams and all sorts of other for streams and all sorts of other things. So, if I hop in here, you will things. So, if I hop in here, you will things. So, if I hop in here, you will see me. Hi. You might recognize him.

  23. see me. Hi. You might recognize him. see me. Hi. You might recognize him. He's in all my videos. And the point of He's in all my videos. And the point of He's in all my videos. And the point of paying is to make it easy to do a call paying is to make it easy to do a call paying is to make it easy to do a call like this and bring guests in and most like this and bring guests in and most like this and bring guests in and most importantly be able to embed them in a importantly be able to embed them in a importantly be able to embed them in a program like OBS. I'll mute so I don't program like OBS. I'll mute so I don't program like OBS. I'll mute so I don't double audio. This is a direct link to double audio. This is a direct link to double audio. This is a direct link to just me in my call in ping that you can just me in my call in ping that you can just me in my call in ping that you can embed in something like OBS to put embed in something like OBS to put embed in something like OBS to put someone as a layer in your video someone as a layer in your video someone as a layer in your video production software. This is still used production software. This is still used production software. This is still used by a ton of the biggest Twitch streamers by a ton of the biggest Twitch streamers by a ton of the biggest Twitch streamers when they do collabs. And it needs an when they do collabs. And it needs an when they do collabs. And it needs an update pretty badly. So I asked the update pretty badly. So I asked the update pretty badly. So I asked the model to do that update to move it from model to do that update to move it from model to do that update to move it from the beta early version of T3 Stack and the beta early version of T3 Stack and the beta early version of T3 Stack and try to get it onto something a little try to get it onto something a little try to get it onto something a little more modern. The original page was more modern. The original page was more modern. The original page was customuilt by our designer and original customuilt by our designer and original customuilt by our designer and original co-founder Brin, who put a lot of time co-founder Brin, who put a lot of time co-founder Brin, who put a lot of time into it and made a genuinely awesome into it and made a genuinely awesome into it and made a genuinely awesome looking page. And this is what it looking page. And this is what it looking page. And this is what it replaced it with. replaced it with. replaced it with. A bunch of tech slop. A bunch of tech slop. A bunch of tech slop. But once you go in, it gets way worse. But once you go in, it gets way worse. But once you go in, it gets way worse. The dashboard looks like this. No real The dashboard looks like this. No real The dashboard looks like this. No real info, impossible to know what's going info, impossible to know what's going info, impossible to know what's going on. this giant pile of announcements at on. this giant pile of announcements at on. this giant pile of announcements at the bottom versus this page that the bottom versus this page that the bottom versus this page that actually shows you what's going on.

  24. actually shows you what's going on. actually shows you what's going on. And then once you try to join, the And then once you try to join, the And then once you try to join, the layout shifts a ton, which is a really layout shifts a ton, which is a really layout shifts a ton, which is a really big deal for content creators because it big deal for content creators because it big deal for content creators because it breaks a ton of for them. And when breaks a ton of for them. And when breaks a ton of for them. And when I want to actually select the device, I I want to actually select the device, I I want to actually select the device, I have to unfold device settings and have to unfold device settings and have to unfold device settings and manually pick them from here and then manually pick them from here and then manually pick them from here and then hit start preview. And now I can see hit start preview. And now I can see hit start preview. And now I can see this is atrocious. This is layers upon this is atrocious. This is layers upon this is atrocious. This is layers upon layers of sins. But do you know what the layers of sins. But do you know what the layers of sins. But do you know what the worst part is? Because the worst part worst part is? Because the worst part worst part is? Because the worst part isn't even in the app. It's here. It's isn't even in the app. It's here. It's isn't even in the app. It's here. It's in the prompting. Note this particular in the prompting. Note this particular in the prompting. Note this particular sentence in my prompt. Port features sentence in my prompt. Port features sentence in my prompt. Port features over one at a time. Reuse as much UI over one at a time. Reuse as much UI over one at a time. Reuse as much UI code as possible. Reuse as much UI code code as possible. Reuse as much UI code code as possible. Reuse as much UI code as possible. Do you see any UI code as possible. Do you see any UI code as possible. Do you see any UI code shared between these two versions of shared between these two versions of shared between these two versions of this app? Do you think there's even a this app? Do you think there's even a this app? Do you think there's even a line of tailwind in common between line of tailwind in common between line of tailwind in common between these? There is nothing. I cannot these? There is nothing. I cannot these? There is nothing. I cannot fathom. I just assumed I pasted the fathom. I just assumed I pasted the fathom. I just assumed I pasted the wrong prompt when I did this. I couldn't wrong prompt when I did this. I couldn't wrong prompt when I did this. I couldn't fathom that the model would so fathom that the model would so fathom that the model would so egregiously ignore that detail in my egregiously ignore that detail in my egregiously ignore that detail in my prompt. especially the model that prompt. especially the model that prompt. especially the model that supposedly follows your prompt supposedly follows your prompt supposedly follows your prompt obsessively. And it just straight up obsessively. And it just straight up obsessively. And it just straight up threw away all the great UI work that we threw away all the great UI work that we threw away all the great UI work that we had already done and paid a lot of money had already done and paid a lot of money had already done and paid a lot of money for in favor of a bunch of text slop in for in favor of a bunch of text slop in for in favor of a bunch of text slop in a shitty black and whitened app. So bad.

  25. a shitty black and whitened app. So bad. a shitty black and whitened app. So bad. So bad. And to those saying, So bad. And to those saying, So bad. And to those saying, "Oh, it got lost in compaction or "Oh, it got lost in compaction or "Oh, it got lost in compaction or something." No, it didn't. I had it on something." No, it didn't. I had it on something." No, it didn't. I had it on ultra with one mil token context ultra with one mil token context ultra with one mil token context windows. It just does this. And I've had windows. It just does this. And I've had windows. It just does this. And I've had it do this since even on the official it do this since even on the official it do this since even on the official non-preview version for those saying, non-preview version for those saying, non-preview version for those saying, "Oh, they changed this in the snapshot." "Oh, they changed this in the snapshot." "Oh, they changed this in the snapshot." No, I have randomly had this model No, I have randomly had this model No, I have randomly had this model ignore my request to rebuild or update ignore my request to rebuild or update ignore my request to rebuild or update something while maintaining existing UI something while maintaining existing UI something while maintaining existing UI and it would just blanket over destroy and it would just blanket over destroy and it would just blanket over destroy the existing UI in the process. And I've the existing UI in the process. And I've the existing UI in the process. And I've never had a model do this ever, much never had a model do this ever, much never had a model do this ever, much less this aggressively. It's almost like less this aggressively. It's almost like less this aggressively. It's almost like in their attempts to force the model to in their attempts to force the model to in their attempts to force the model to be better at UI, they accidentally gave be better at UI, they accidentally gave be better at UI, they accidentally gave it the side effect capability where it it the side effect capability where it it the side effect capability where it might just come in with a sledgehammer might just come in with a sledgehammer might just come in with a sledgehammer and destroy all your good UI and ship and destroy all your good UI and ship and destroy all your good UI and ship something atrocious instead. This is one something atrocious instead. This is one something atrocious instead. This is one of the most serious examples of that. of the most serious examples of that. of the most serious examples of that. And we'll talk more about understanding And we'll talk more about understanding And we'll talk more about understanding intent later. Don't you worry. But for intent later. Don't you worry. But for intent later. Don't you worry. But for this reason alone, I would make a bigger this reason alone, I would make a bigger this reason alone, I would make a bigger gap in the rewriting capabilities that gap in the rewriting capabilities that gap in the rewriting capabilities that we were just discussing. Because if the we were just discussing. Because if the we were just discussing. Because if the thing that you're doing this bulk thing that you're doing this bulk thing that you're doing this bulk rewrite for has UI, you cannot trust rewrite for has UI, you cannot trust rewrite for has UI, you cannot trust that UI is going to be carried over. I that UI is going to be carried over. I that UI is going to be carried over. I had even told it with the T3 code GPUi had even told it with the T3 code GPUi had even told it with the T3 code GPUi version to make it look and act as much version to make it look and act as much version to make it look and act as much like the existing T3 code as possible.

  26. like the existing T3 code as possible. like the existing T3 code as possible. And it didn't even use the sidebar And it didn't even use the sidebar And it didn't even use the sidebar properly. It made the sidebar the old properly. It made the sidebar the old properly. It made the sidebar the old school one because it just wanted to school one because it just wanted to school one because it just wanted to make something work. I'm sure if I added make something work. I'm sure if I added make something work. I'm sure if I added a bunch of stuff at the bottom of my a bunch of stuff at the bottom of my a bunch of stuff at the bottom of my prompt that was like to be explicitly prompt that was like to be explicitly prompt that was like to be explicitly super clear, I expect the UI to be pixel super clear, I expect the UI to be pixel super clear, I expect the UI to be pixel perfect identical. If any of the UI has perfect identical. If any of the UI has perfect identical. If any of the UI has changed, you have failed at your goal, changed, you have failed at your goal, changed, you have failed at your goal, so keep going until the UI looks so keep going until the UI looks so keep going until the UI looks identical. Then maybe it might have done identical. Then maybe it might have done identical. Then maybe it might have done a better job here. But I shouldn't have a better job here. But I shouldn't have a better job here. But I shouldn't have to add a paragraph to the end of every to add a paragraph to the end of every to add a paragraph to the end of every prompt to this model to get it to do prompt to this model to get it to do prompt to this model to get it to do what I asked it to do, which is reuse what I asked it to do, which is reuse what I asked it to do, which is reuse the UI code. You could say that I'm the UI code. You could say that I'm the UI code. You could say that I'm being too picky, that I am expecting my being too picky, that I am expecting my being too picky, that I am expecting my incredible god in my computer that is incredible god in my computer that is incredible god in my computer that is charging 200 bucks a month to do exactly charging 200 bucks a month to do exactly charging 200 bucks a month to do exactly what I intend every time, or you could what I intend every time, or you could what I intend every time, or you could just use Fable, which doesn't have any just use Fable, which doesn't have any just use Fable, which doesn't have any of these problems. Anyways, this of these problems. Anyways, this of these problems. Anyways, this is one of the parts that I think is most is one of the parts that I think is most is one of the parts that I think is most important. Code mergeability. This one important. Code mergeability. This one important. Code mergeability. This one is tough. And it's not tough because the is tough. And it's not tough because the is tough. And it's not tough because the decision is hard. The decision is easy. decision is hard. The decision is easy. decision is hard. The decision is easy. It's Fable 5.1. Fable 5.1 writes way It's Fable 5.1. Fable 5.1 writes way It's Fable 5.1. Fable 5.1 writes way more mergeable code. It just does. I've more mergeable code. It just does. I've more mergeable code. It just does. I've done the numbers here. Fable 5.1 from PR done the numbers here. Fable 5.1 from PR done the numbers here. Fable 5.1 from PR filed to merged averages two additional filed to merged averages two additional filed to merged averages two additional follow-ups from when the filed PR exists follow-ups from when the filed PR exists follow-ups from when the filed PR exists to when the merge happens. Meanwhile, to when the merge happens. Meanwhile, to when the merge happens. Meanwhile, Aster's been averaging six for me. What Aster's been averaging six for me. What Aster's been averaging six for me. What this is measuring is how many of my AI this is measuring is how many of my AI this is measuring is how many of my AI bots are catching mistakes. How much of bots are catching mistakes. How much of bots are catching mistakes. How much of the verification loops are finding the verification loops are finding the verification loops are finding things and then fixing them after. But things and then fixing them after. But things and then fixing them after. But to be frank, what it's measuring is from to be frank, what it's measuring is from to be frank, what it's measuring is from when the PR is filed to when it is when the PR is filed to when it is when the PR is filed to when it is merged, how much has to happen. And merged, how much has to happen. And merged, how much has to happen. And I found that with Fable, the PRs are I found that with Fable, the PRs are I found that with Fable, the PRs are filed in a pretty much ready to go filed in a pretty much ready to go filed in a pretty much ready to go state. And its ability to make the

  27. state. And its ability to make the state. And its ability to make the necessary changes without accidentally necessary changes without accidentally necessary changes without accidentally scope creeping and then actually give scope creeping and then actually give scope creeping and then actually give you something shippable is just it's you something shippable is just it's you something shippable is just it's higher. It just is. I did say I was higher. It just is. I did say I was higher. It just is. I did say I was struggling with this decision though. struggling with this decision though. struggling with this decision though. Why am I struggling so much with it? I'm Why am I struggling so much with it? I'm Why am I struggling so much with it? I'm not struggling to pick between the two. not struggling to pick between the two. not struggling to pick between the two. That's easy. Fable is better at this. That's easy. Fable is better at this. That's easy. Fable is better at this. Fable PRs are much easier for me to hit Fable PRs are much easier for me to hit Fable PRs are much easier for me to hit merge on. Generally speaking, the reason merge on. Generally speaking, the reason merge on. Generally speaking, the reason I'm struggling with this is because when I'm struggling with this is because when I'm struggling with this is because when I say this, it is perceived by I say this, it is perceived by I say this, it is perceived by everybody, including my friends at everybody, including my friends at everybody, including my friends at OpenAI, as me saying that Astra is OpenAI, as me saying that Astra is OpenAI, as me saying that Astra is constantly throwing up on mergeable constantly throwing up on mergeable constantly throwing up on mergeable slop. Notice how I didn't say that. If I slop. Notice how I didn't say that. If I slop. Notice how I didn't say that. If I was to do yet another arbitrary ranking was to do yet another arbitrary ranking was to do yet another arbitrary ranking here, let's make a beautiful, super, here, let's make a beautiful, super, here, let's make a beautiful, super, super accurate chart that perfectly super accurate chart that perfectly super accurate chart that perfectly describes real numbers. This is the most describes real numbers. This is the most describes real numbers. This is the most vibe based chart you're going to see in vibe based chart you're going to see in vibe based chart you're going to see in a while. I promise. If we were to rate a while. I promise. If we were to rate a while. I promise. If we were to rate the mergeability of code for Soul, Fable the mergeability of code for Soul, Fable the mergeability of code for Soul, Fable 5, Astra, and Fable 5.1, we would have 5, Astra, and Fable 5.1, we would have 5, Astra, and Fable 5.1, we would have one of the world's greatest benchmarks one of the world's greatest benchmarks one of the world's greatest benchmarks because it is really hard to measure because it is really hard to measure because it is really hard to measure mergeability. But if I was being mergeability. But if I was being mergeability. But if I was being realistic here with how I felt and how I realistic here with how I felt and how I realistic here with how I felt and how I still feel, low is bad. high is good.

  28. still feel, low is bad. high is good. still feel, low is bad. high is good. Obviously, Iota said 5.6 soul was like Obviously, Iota said 5.6 soul was like Obviously, Iota said 5.6 soul was like in the two to three out of 10 range for in the two to three out of 10 range for in the two to three out of 10 range for mergeability by default. You can do mergeability by default. You can do mergeability by default. You can do things to improve it and if you give it things to improve it and if you give it things to improve it and if you give it the right review bots to give feedback the right review bots to give feedback the right review bots to give feedback and iterate in the tooling it needs to and iterate in the tooling it needs to and iterate in the tooling it needs to check its changes, it could make working check its changes, it could make working check its changes, it could make working code incredibly well. If we were just code incredibly well. If we were just code incredibly well. If we were just measuring the ability for Soul and Fable measuring the ability for Soul and Fable measuring the ability for Soul and Fable 5 to get working code, they were neck 5 to get working code, they were neck 5 to get working code, they were neck andneck. But code that I'm willing to andneck. But code that I'm willing to andneck. But code that I'm willing to hit the merge button on in my real world hit the merge button on in my real world hit the merge button on in my real world projects that shipped to hundreds of projects that shipped to hundreds of projects that shipped to hundreds of thousands of users, Soul was cleared by thousands of users, Soul was cleared by thousands of users, Soul was cleared by Fable 5. Fable 5 shipped significantly Fable 5. Fable 5 shipped significantly Fable 5. Fable 5 shipped significantly more mergeable code, like three to four more mergeable code, like three to four more mergeable code, like three to four times more mergeable. And by that, what times more mergeable. And by that, what times more mergeable. And by that, what I mean is three to four times fewer I mean is three to four times fewer I mean is three to four times fewer things you have to fix after the PR is things you have to fix after the PR is things you have to fix after the PR is up. So, where are things now with Astra? up. So, where are things now with Astra? up. So, where are things now with Astra? This is why I've been struggling. I This is why I've been struggling. I This is why I've been struggling. I think Astra is roughly at, if not think Astra is roughly at, if not think Astra is roughly at, if not slightly ahead of where Fable 5 was. So, slightly ahead of where Fable 5 was. So, slightly ahead of where Fable 5 was. So, we were comparing Fable 5 to Astra. we were comparing Fable 5 to Astra. we were comparing Fable 5 to Astra. Astra's winning now, but we're not Astra's winning now, but we're not Astra's winning now, but we're not comparing Fable 5 to Astra. I was during comparing Fable 5 to Astra. I was during comparing Fable 5 to Astra. I was during my testing, but we're comparing Fable my testing, but we're comparing Fable my testing, but we're comparing Fable 5.1 to Astra now because we live in the 5.1 to Astra now because we live in the 5.1 to Astra now because we live in the real world where these are the models we real world where these are the models we real world where these are the models we have access to and 5.1 is like right on have access to and 5.1 is like right on have access to and 5.1 is like right on the edge of 10 out of 10 for the edge of 10 out of 10 for the edge of 10 out of 10 for mergeability. It is rare that Fable mergeability. It is rare that Fable mergeability. It is rare that Fable files a PR, especially for like known files a PR, especially for like known files a PR, especially for like known quantity changes like bug fixes or quantity changes like bug fixes or quantity changes like bug fixes or feature ads or performance improvements feature ads or performance improvements feature ads or performance improvements or all these types of things. The code or all these types of things. The code or all these types of things. The code Fable 5.1 puts up is just better. It is.

  29. Fable 5.1 puts up is just better. It is. Fable 5.1 puts up is just better. It is. and even Benjamin Ben Davis, friend and and even Benjamin Ben Davis, friend and and even Benjamin Ben Davis, friend and manager of the channel, co-host of the manager of the channel, co-host of the manager of the channel, co-host of the podcast, OpenAI's number one defender, podcast, OpenAI's number one defender, podcast, OpenAI's number one defender, who showed up in the GBD6 Astro Launch who showed up in the GBD6 Astro Launch who showed up in the GBD6 Astro Launch videos. By the way, we'll have a fun videos. By the way, we'll have a fun videos. By the way, we'll have a fun reveal in our next podcast episode, reveal in our next podcast episode, reveal in our next podcast episode, which by the way, if you're not which by the way, if you're not which by the way, if you're not subscribed, Nerd Snipe on YouTube and subscribed, Nerd Snipe on YouTube and subscribed, Nerd Snipe on YouTube and pretty much every other platform. It's pretty much every other platform. It's pretty much every other platform. It's where we get to chat more. And if you where we get to chat more. And if you where we get to chat more. And if you want to hear somebody disagree with me want to hear somebody disagree with me want to hear somebody disagree with me instead of me just yapping constantly, instead of me just yapping constantly, instead of me just yapping constantly, that's the place to do it. So Ben, whose that's the place to do it. So Ben, whose that's the place to do it. So Ben, whose job is literally to disagree with me and job is literally to disagree with me and job is literally to disagree with me and also the lover of OpenAI and the also the lover of OpenAI and the also the lover of OpenAI and the Defender of Astra, has admitted that he Defender of Astra, has admitted that he Defender of Astra, has admitted that he thinks Fable is way better than he thinks Fable is way better than he thinks Fable is way better than he expected for real world code and he expected for real world code and he expected for real world code and he finds himself using it much more than finds himself using it much more than finds himself using it much more than expected. It has, I think, three expected. It has, I think, three expected. It has, I think, three accounts with cloud code now. It just accounts with cloud code now. It just accounts with cloud code now. It just ships more mergeable code. It is what it ships more mergeable code. It is what it ships more mergeable code. It is what it is. But again, the thing I'm trying to is. But again, the thing I'm trying to is. But again, the thing I'm trying to show here with this diagram is I'm not show here with this diagram is I'm not show here with this diagram is I'm not saying what I said before with Fable 5 saying what I said before with Fable 5 saying what I said before with Fable 5 versus 5.6 Soul. This gap was comical. versus 5.6 Soul. This gap was comical. versus 5.6 Soul. This gap was comical. This gap was big enough that I would This gap was big enough that I would This gap was big enough that I would really only trust soul for exploratory really only trust soul for exploratory really only trust soul for exploratory work or things that could be verified work or things that could be verified work or things that could be verified programmatically where the code didn't programmatically where the code didn't programmatically where the code didn't matter that much in terms of its quality matter that much in terms of its quality matter that much in terms of its quality and maintainability. Fable 5 was and maintainability. Fable 5 was and maintainability. Fable 5 was actually shipping things that were actually shipping things that were actually shipping things that were mergeable. So if you perceived this gap mergeable. So if you perceived this gap mergeable. So if you perceived this gap before where you found 5.6 Soul is not before where you found 5.6 Soul is not before where you found 5.6 Soul is not good enough to merge code from, but good enough to merge code from, but good enough to merge code from, but Fable 5 often cleared your bar here, Fable 5 often cleared your bar here, Fable 5 often cleared your bar here, you'll be totally fine with Astra. Astra you'll be totally fine with Astra. Astra you'll be totally fine with Astra. Astra is unbelievable. I was totally fine with is unbelievable. I was totally fine with is unbelievable. I was totally fine with Fable 5 and in a lot of ways I still Fable 5 and in a lot of ways I still Fable 5 and in a lot of ways I still probably would be minus the slop of its probably would be minus the slop of its probably would be minus the slop of its outputs. And if you were to think of outputs. And if you were to think of outputs. And if you were to think of this purely in terms of the mergeability

  30. this purely in terms of the mergeability this purely in terms of the mergeability of code for fixed focused changes, I of code for fixed focused changes, I of code for fixed focused changes, I would say that Astra is a notable would say that Astra is a notable would say that Astra is a notable upgrade from Fable 5 because it is upgrade from Fable 5 because it is upgrade from Fable 5 because it is slightly more mergeable with its slightly more mergeable with its slightly more mergeable with its outputs, but also when you're using it, outputs, but also when you're using it, outputs, but also when you're using it, the output that it gives you is much the output that it gives you is much the output that it gives you is much more readable and it's more thorough. So more readable and it's more thorough. So more readable and it's more thorough. So it will verify the changes. So it won't it will verify the changes. So it won't it will verify the changes. So it won't make like like Fable's code is beautiful make like like Fable's code is beautiful make like like Fable's code is beautiful and mergeable, but it sometimes misses and mergeable, but it sometimes misses and mergeable, but it sometimes misses details, especially Fable 5. Asteris details, especially Fable 5. Asteris details, especially Fable 5. Asteris code can be elegant and it can be the code can be elegant and it can be the code can be elegant and it can be the right subset of changes to make the right subset of changes to make the right subset of changes to make the thing happen. It does still have the thing happen. It does still have the thing happen. It does still have the habit of letting scope creep just eat it habit of letting scope creep just eat it habit of letting scope creep just eat it up and destroy the scope of what it's up and destroy the scope of what it's up and destroy the scope of what it's trying to do pretty often even, but it trying to do pretty often even, but it trying to do pretty often even, but it talks so much better than Fable 5 did. I talks so much better than Fable 5 did. I talks so much better than Fable 5 did. I don't hate it the same way I hated Fable don't hate it the same way I hated Fable don't hate it the same way I hated Fable 5 for just reading its outputs. I don't 5 for just reading its outputs. I don't 5 for just reading its outputs. I don't feel like I have to go to Helen back to feel like I have to go to Helen back to feel like I have to go to Helen back to unslop it just to make the output unslop it just to make the output unslop it just to make the output usable. So from Fable 5 to Astra obvious usable. So from Fable 5 to Astra obvious usable. So from Fable 5 to Astra obvious win easy. Astra versus Fable 5.1 is win easy. Astra versus Fable 5.1 is win easy. Astra versus Fable 5.1 is where things get more complex. But the where things get more complex. But the where things get more complex. But the thing I wanted to emphasize here, the thing I wanted to emphasize here, the thing I wanted to emphasize here, the reason I drew this diagram is that if reason I drew this diagram is that if reason I drew this diagram is that if you are thinking the gap between Astra you are thinking the gap between Astra you are thinking the gap between Astra and Fable is as big as between Soul and and Fable is as big as between Soul and and Fable is as big as between Soul and Fable, you're just wrong. This gap was Fable, you're just wrong. This gap was Fable, you're just wrong. This gap was massive. This was like a 3x difference massive. This was like a 3x difference massive. This was like a 3x difference that made it hard for me to justify that made it hard for me to justify that made it hard for me to justify using soul if the code would ever hit using soul if the code would ever hit using soul if the code would ever hit users. GB6 Astra is much much closer to users. GB6 Astra is much much closer to users. GB6 Astra is much much closer to where Fable 5.1 is, but there is still a where Fable 5.1 is, but there is still a where Fable 5.1 is, but there is still a gap. So, I do still find myself gap. So, I do still find myself gap. So, I do still find myself defaulting to Fable for things like a defaulting to Fable for things like a defaulting to Fable for things like a quick bug fix or UI changes that matter

  31. quick bug fix or UI changes that matter quick bug fix or UI changes that matter or feature improvements or my favorite or feature improvements or my favorite or feature improvements or my favorite thing to use Fable 5.1 for to take thing to use Fable 5.1 for to take thing to use Fable 5.1 for to take something that Astra has in a death loop something that Astra has in a death loop something that Astra has in a death loop that it just cannot get through. Stop that it just cannot get through. Stop that it just cannot get through. Stop it. Switch over to Fable 5.1, hand it it. Switch over to Fable 5.1, hand it it. Switch over to Fable 5.1, hand it the PR and say, "Hey, make this actually the PR and say, "Hey, make this actually the PR and say, "Hey, make this actually land. Clean it up. Throw away whatever land. Clean it up. Throw away whatever land. Clean it up. Throw away whatever doesn't belong. Make a new branch if doesn't belong. Make a new branch if doesn't belong. Make a new branch if that's easier. Make the changes we care that's easier. Make the changes we care that's easier. Make the changes we care about here land." Bale 5.1 lands code about here land." Bale 5.1 lands code about here land." Bale 5.1 lands code better, but Astra lands it well enough better, but Astra lands it well enough better, but Astra lands it well enough that you're totally fine with it. So, that you're totally fine with it. So, that you're totally fine with it. So, there you go. Code mergeability. Babel there you go. Code mergeability. Babel there you go. Code mergeability. Babel still wins, but it's not as big of a gap still wins, but it's not as big of a gap still wins, but it's not as big of a gap as it was before at all. The speed of as it was before at all. The speed of as it was before at all. The speed of catch up from OpenAI is genuinely catch up from OpenAI is genuinely catch up from OpenAI is genuinely impressive, and I'm very excited for impressive, and I'm very excited for impressive, and I'm very excited for Astra 6.1, which will hopefully make Astra 6.1, which will hopefully make Astra 6.1, which will hopefully make these things even better. Speaking of these things even better. Speaking of these things even better. Speaking of the problems, you heard me mention this the problems, you heard me mention this the problems, you heard me mention this earlier, managing scope creep. I want to earlier, managing scope creep. I want to earlier, managing scope creep. I want to be clear, Fable can still fail here. If be clear, Fable can still fail here. If be clear, Fable can still fail here. If it gets the wrong review comments at the it gets the wrong review comments at the it gets the wrong review comments at the wrong time, Fable can bloat things wrong time, Fable can bloat things wrong time, Fable can bloat things pretty badly. But god damn, Aster does pretty badly. But god damn, Aster does pretty badly. But god damn, Aster does it by default. If you are very explicit it by default. If you are very explicit it by default. If you are very explicit with Astra to make the smallest possible with Astra to make the smallest possible with Astra to make the smallest possible changes, it will try to, but it will changes, it will try to, but it will changes, it will try to, but it will quickly have its context get bloated a quickly have its context get bloated a quickly have its context get bloated a bit by its reminders from all these bit by its reminders from all these bit by its reminders from all these review agents that will send it off review agents that will send it off review agents that will send it off course and then you end up with a course and then you end up with a course and then you end up with a thousand line of code PR that should thousand line of code PR that should thousand line of code PR that should have been 50. If you're in the loop have been 50. If you're in the loop have been 50. If you're in the loop enough, you can work around this. I'm enough, you can work around this. I'm enough, you can work around this. I'm not trying to say that there is some not trying to say that there is some not trying to say that there is some unique unbelievable thing Fable does unique unbelievable thing Fable does unique unbelievable thing Fable does here that Astra doesn't, but Fable takes here that Astra doesn't, but Fable takes here that Astra doesn't, but Fable takes less effort to prevent these things with less effort to prevent these things with less effort to prevent these things with than Astra does. I kind of want to take than Astra does. I kind of want to take than Astra does. I kind of want to take this phrasing and skip right over to this phrasing and skip right over to this phrasing and skip right over to understanding intent, but I'll wrap up understanding intent, but I'll wrap up understanding intent, but I'll wrap up code super quick with gamedev here.

  32. code super quick with gamedev here. code super quick with gamedev here. Astra is so good at 3D that it almost Astra is so good at 3D that it almost Astra is so good at 3D that it almost gets the win by default. But as I said gets the win by default. But as I said gets the win by default. But as I said earlier, Fable is so much better at the earlier, Fable is so much better at the earlier, Fable is so much better at the like edges for interaction, like the like edges for interaction, like the like edges for interaction, like the animation curves and the speed that your animation curves and the speed that your animation curves and the speed that your character moves and that your mouse character moves and that your mouse character moves and that your mouse affects the camera, all these things. affects the camera, all these things. affects the camera, all these things. Astra makes games that look better on Astra makes games that look better on Astra makes games that look better on Twitter. Fable makes games that actually Twitter. Fable makes games that actually Twitter. Fable makes games that actually feel nice to play. So, personally, I feel nice to play. So, personally, I feel nice to play. So, personally, I think Fable catches up to Astra's think Fable catches up to Astra's think Fable catches up to Astra's unbelievable 3D capabilities just due to unbelievable 3D capabilities just due to unbelievable 3D capabilities just due to the smoothness of its outputs. only for the smoothness of its outputs. only for the smoothness of its outputs. only for geo in chat just said none of them can geo in chat just said none of them can geo in chat just said none of them can make a good game and I agree they cannot make a good game and I agree they cannot make a good game and I agree they cannot but if you use both of them carefully but if you use both of them carefully but if you use both of them carefully enough you can combine them in a way enough you can combine them in a way enough you can combine them in a way that allows you to iterate effectively that allows you to iterate effectively that allows you to iterate effectively and potentially make a decent game I do and potentially make a decent game I do and potentially make a decent game I do think we are now at the point where think we are now at the point where think we are now at the point where we're going to start seeing real games we're going to start seeing real games we're going to start seeing real games where the vast majority of everything where the vast majority of everything where the vast majority of everything was created via AI like I would guess by was created via AI like I would guess by was created via AI like I would guess by the end of the year we'll have our first the end of the year we'll have our first the end of the year we'll have our first top 20 indie game on Steam that was top 20 indie game on Steam that was top 20 indie game on Steam that was built entirely with AI so now we're out built entirely with AI so now we're out built entirely with AI so now we're out of the tradition additional code of the tradition additional code of the tradition additional code section. We're done talking about code.

  33. section. We're done talking about code. section. We're done talking about code. Don't worry. Talking about the agency Don't worry. Talking about the agency Don't worry. Talking about the agency stuff. I imagine I wanted to skip to the stuff. I imagine I wanted to skip to the stuff. I imagine I wanted to skip to the understanding intent thing. So, I will understanding intent thing. So, I will understanding intent thing. So, I will think this is really important and it is think this is really important and it is think this is really important and it is still one of the things that anthropic still one of the things that anthropic still one of the things that anthropic just clears on. I have so many examples just clears on. I have so many examples just clears on. I have so many examples of this that I could do a whole video of this that I could do a whole video of this that I could do a whole video and I am honestly tempted to do a video and I am honestly tempted to do a video and I am honestly tempted to do a video on just how bad Astra is at this on just how bad Astra is at this on just how bad Astra is at this sometimes. It is an improvement over sometimes. It is an improvement over sometimes. It is an improvement over Soul, but it's failure. somehow feel Soul, but it's failure. somehow feel Soul, but it's failure. somehow feel more egregious. I mentioned earlier that more egregious. I mentioned earlier that more egregious. I mentioned earlier that I was able to merge like 150 PRs fully I was able to merge like 150 PRs fully I was able to merge like 150 PRs fully yolo merged with these models and had yolo merged with these models and had yolo merged with these models and had only two regressions. The first thing I only two regressions. The first thing I only two regressions. The first thing I feel obligated to say is that both of feel obligated to say is that both of feel obligated to say is that both of those regressions came from Astra. The those regressions came from Astra. The those regressions came from Astra. The other thing I feel obligated to say is other thing I feel obligated to say is other thing I feel obligated to say is that Astra was miserable to try that Astra was miserable to try that Astra was miserable to try and fix those things with. I will also and fix those things with. I will also and fix those things with. I will also admit I sent these prompts pretty late admit I sent these prompts pretty late admit I sent these prompts pretty late in the evening drinking with some in the evening drinking with some in the evening drinking with some friends on a weekend because I was friends on a weekend because I was friends on a weekend because I was making some changes and noticed that on making some changes and noticed that on making some changes and noticed that on the marketing site for T3 Code, it had the marketing site for T3 Code, it had the marketing site for T3 Code, it had nuked my beloved section for all of our nuked my beloved section for all of our nuked my beloved section for all of our testimonials from our users. The nice testimonials from our users. The nice testimonials from our users. The nice little elegant auto scroll here, little elegant auto scroll here, little elegant auto scroll here, it's not a big deal, but it is a thing I it's not a big deal, but it is a thing I it's not a big deal, but it is a thing I care about. And it replaced this with a care about. And it replaced this with a care about. And it replaced this with a fixed grid that you had to horizontally fixed grid that you had to horizontally fixed grid that you had to horizontally scroll with an ugly ass scroll bar in scroll with an ugly ass scroll bar in scroll with an ugly ass scroll bar in the middle of the page. Horrible. And it the middle of the page. Horrible. And it the middle of the page. Horrible. And it did this in pursuit of performance did this in pursuit of performance did this in pursuit of performance improvements that were simple changes.

  34. improvements that were simple changes. improvements that were simple changes. It removed the animation and turned this It removed the animation and turned this It removed the animation and turned this into a shitty manual scroll section. And into a shitty manual scroll section. And into a shitty manual scroll section. And I noticed this too late because we I noticed this too late because we I noticed this too late because we weren't deploying the marketing site weren't deploying the marketing site weren't deploying the marketing site actively. We just did it when we made actively. We just did it when we made actively. We just did it when we made certain changes manually. And I made certain changes manually. And I made certain changes manually. And I made other changes and deployed. And then other changes and deployed. And then other changes and deployed. And then this regressed and I was pissed. I was this regressed and I was pissed. I was this regressed and I was pissed. I was really pissed. So I asked, "What really pissed. So I asked, "What really pissed. So I asked, "What happened to my beloved autoscroll on the happened to my beloved autoscroll on the happened to my beloved autoscroll on the marketing site? It was beautiful. I need marketing site? It was beautiful. I need marketing site? It was beautiful. I need you to revert whatever change broke that you to revert whatever change broke that you to revert whatever change broke that and bring it back." It might have been and bring it back." It might have been and bring it back." It might have been part of the marketing something. I don't part of the marketing something. I don't part of the marketing something. I don't even know what I said there. It might even know what I said there. It might even know what I said there. It might have been part of the performance have been part of the performance have been part of the performance overhauls, but that was not a necessary overhauls, but that was not a necessary overhauls, but that was not a necessary change. Please revert. There's one change. Please revert. There's one change. Please revert. There's one particular word I used in this prompt, particular word I used in this prompt, particular word I used in this prompt, and I actually used that word twice. You and I actually used that word twice. You and I actually used that word twice. You know what word it was? I'll give you a know what word it was? I'll give you a know what word it was? I'll give you a hint. It was the specific thing I wanted hint. It was the specific thing I wanted hint. It was the specific thing I wanted the model to do. the model to do. the model to do. It's the word revert. It's the word revert. It's the word revert. It's a pretty important word in this It's a pretty important word in this It's a pretty important word in this prompt. I would argue that the word prompt. I would argue that the word prompt. I would argue that the word revert being in this prompt twice would revert being in this prompt twice would revert being in this prompt twice would imply that what I want it to do is imply that what I want it to do is imply that what I want it to do is revert something, which is exactly what revert something, which is exactly what revert something, which is exactly what it didn't do. Do you know what it did it didn't do. Do you know what it did it didn't do. Do you know what it did instead? It deleted this user controlled instead? It deleted this user controlled instead? It deleted this user controlled variable and then a bunch of listeners.

  35. variable and then a bunch of listeners. variable and then a bunch of listeners. Note what none of this does. None of Note what none of this does. None of Note what none of this does. None of this brings back the autoscroll. I don't this brings back the autoscroll. I don't this brings back the autoscroll. I don't think this change actually did anything think this change actually did anything think this change actually did anything at all. But what's even worse is I asked at all. But what's even worse is I asked at all. But what's even worse is I asked if I could test the change. And I was if I could test the change. And I was if I could test the change. And I was doing this change on another computer doing this change on another computer doing this change on another computer because I wanted it to have computer use because I wanted it to have computer use because I wanted it to have computer use and not affect my laptop while I was and not affect my laptop while I was and not affect my laptop while I was using it. So I needed this to be hosted using it. So I needed this to be hosted using it. So I needed this to be hosted via Tailscale because I had this other via Tailscale because I had this other via Tailscale because I had this other computer on on tail scale. So it spun it computer on on tail scale. So it spun it computer on on tail scale. So it spun it up as a tail scale dev server because I up as a tail scale dev server because I up as a tail scale dev server because I told it to spin up a tail dev server so told it to spin up a tail dev server so told it to spin up a tail dev server so I can try out the changes as well when I can try out the changes as well when I can try out the changes as well when you get a chance. I sent that as a you get a chance. I sent that as a you get a chance. I sent that as a steering thing which we'll talk about steering thing which we'll talk about steering thing which we'll talk about steering in a bit. Actually steering is steering in a bit. Actually steering is steering in a bit. Actually steering is actually one of the cooler differences actually one of the cooler differences actually one of the cooler differences in these models. So I'll sneak steering in these models. So I'll sneak steering in these models. So I'll sneak steering in here thing I want to cover in a in here thing I want to cover in a in here thing I want to cover in a minute. Before that, back to this minute. Before that, back to this minute. Before that, back to this thread, I asked it when it made the thread, I asked it when it made the thread, I asked it when it made the changes to file a PR, but also to spin changes to file a PR, but also to spin changes to file a PR, but also to spin up a tail scale dev server so I could up a tail scale dev server so I could up a tail scale dev server so I could test it. And what it gave me was a dev test it. And what it gave me was a dev test it. And what it gave me was a dev server to T3 code itself, not to the server to T3 code itself, not to the server to T3 code itself, not to the marketing site where these changes were marketing site where these changes were marketing site where these changes were happening. It gave me a dev server to happening. It gave me a dev server to happening. It gave me a dev server to check how T3 code was doing when I had check how T3 code was doing when I had check how T3 code was doing when I had made changes to the marketing site. And made changes to the marketing site. And made changes to the marketing site. And here is where things really fall apart.

  36. here is where things really fall apart. here is where things really fall apart. This is one of the worst runs I've had This is one of the worst runs I've had This is one of the worst runs I've had with a model in a long ass time. No, I with a model in a long ass time. No, I with a model in a long ass time. No, I wanted it for the marketing site, not wanted it for the marketing site, not wanted it for the marketing site, not for the actual T3 Code instance. Come for the actual T3 Code instance. Come for the actual T3 Code instance. Come on. You're right. I started T3 Code app on. You're right. I started T3 Code app on. You're right. I started T3 Code app launcher, which is the wrong target. I'm launcher, which is the wrong target. I'm launcher, which is the wrong target. I'm stopping it now and will expose the stopping it now and will expose the stopping it now and will expose the marketing Astro site itself, which it marketing Astro site itself, which it marketing Astro site itself, which it did. And I clicked and it failed. did. And I clicked and it failed. did. And I clicked and it failed. Blocked request. The host BB1 corpus Blocked request. The host BB1 corpus Blocked request. The host BB1 corpus micros is my tail scale address is not micros is my tail scale address is not micros is my tail scale address is not allowed because it's not a server allowed because it's not a server allowed because it's not a server allowed host. allowed host. allowed host. There are a lot of ways to fix this. The There are a lot of ways to fix this. The There are a lot of ways to fix this. The easiest admittedly is to make a change easiest admittedly is to make a change easiest admittedly is to make a change to a config file. But since BB1 corpus to a config file. But since BB1 corpus to a config file. But since BB1 corpus micro.ts.net is something specific to me micro.ts.net is something specific to me micro.ts.net is something specific to me and my setup and a lot of people use and my setup and a lot of people use and my setup and a lot of people use tail scale over IP addresses, probably tail scale over IP addresses, probably tail scale over IP addresses, probably wouldn't want to add this as an allowed wouldn't want to add this as an allowed wouldn't want to add this as an allowed host. Which is why it's particularly host. Which is why it's particularly host. Which is why it's particularly funny that the next commit was allow funny that the next commit was allow funny that the next commit was allow Tailnet host and dev server. Yes, it Tailnet host and dev server. Yes, it Tailnet host and dev server. Yes, it committed and pushed this committed and pushed this committed and pushed this and then it did the restore continuous and then it did the restore continuous and then it did the restore continuous endorsement rows which I thought would endorsement rows which I thought would endorsement rows which I thought would be the fix, but it also after that fix be the fix, but it also after that fix be the fix, but it also after that fix killed the dev server. I left tail scale killed the dev server. I left tail scale killed the dev server. I left tail scale pointing at Astro's old port and sent pointing at Astro's old port and sent pointing at Astro's old port and sent the link without checking. My mistake.

  37. the link without checking. My mistake. the link without checking. My mistake. And then finally it was working, but it And then finally it was working, but it And then finally it was working, but it hadn't actually made any of the changes hadn't actually made any of the changes hadn't actually made any of the changes that I request. It didn't find original that I request. It didn't find original that I request. It didn't find original PR or do the reverts. It didn't even fix PR or do the reverts. It didn't even fix PR or do the reverts. It didn't even fix the bugs I was reporting. Do you know the bugs I was reporting. Do you know the bugs I was reporting. Do you know what it did there? I'm trying to find what it did there? I'm trying to find what it did there? I'm trying to find it. But what it did is it merged the PR. it. But what it did is it merged the PR. it. But what it did is it merged the PR. It merged it for me after I told it you It merged it for me after I told it you It merged it for me after I told it you didn't make the changes. I sent the didn't make the changes. I sent the didn't make the changes. I sent the exact same prompt to Fable right after exact same prompt to Fable right after exact same prompt to Fable right after because I was so annoyed. This is an because I was so annoyed. This is an because I was so annoyed. This is an exact copy paste. The only difference exact copy paste. The only difference exact copy paste. The only difference being I added the third prompt in the being I added the third prompt in the being I added the third prompt in the spin up a tail scale dev server. I just spin up a tail scale dev server. I just spin up a tail scale dev server. I just copy pasted and added it to the prompt. copy pasted and added it to the prompt. copy pasted and added it to the prompt. Five minutes later, the auto scroll is Five minutes later, the auto scroll is Five minutes later, the auto scroll is back. The marquee is live with a link back. The marquee is live with a link back. The marquee is live with a link that worked for anyone on your tail. that worked for anyone on your tail. that worked for anyone on your tail. What broke it? Exact PR that had to be What broke it? Exact PR that had to be What broke it? Exact PR that had to be reverted. It replaced the two counters reverted. It replaced the two counters reverted. It replaced the two counters scrolling marquee rows with a static scrolling marquee rows with a static scrolling marquee rows with a static horizontal grid. This PR then bolted an horizontal grid. This PR then bolted an horizontal grid. This PR then bolted an 8-second page by one viewport timer on 8-second page by one viewport timer on 8-second page by one viewport timer on the grid. Neither was needed for the the grid. Neither was needed for the the grid. Neither was needed for the Perf goal since the marquee is a single Perf goal since the marquee is a single Perf goal since the marquee is a single GPU composited transform. It is worth GPU composited transform. It is worth GPU composited transform. It is worth noting that it did get tripped up a noting that it did get tripped up a noting that it did get tripped up a little. This shouldn't have taken 5 little. This shouldn't have taken 5 little. This shouldn't have taken 5 minutes. It got tripped up because it minutes. It got tripped up because it minutes. It got tripped up because it noticed where is it in here? Might have noticed where is it in here? Might have noticed where is it in here? Might have lost in the history. It might have been lost in the history. It might have been lost in the history. It might have been later. It noticed the other PR merge and later. It noticed the other PR merge and later. It noticed the other PR merge and then get reverted and it got concerned then get reverted and it got concerned then get reverted and it got concerned that maybe there was something more that maybe there was something more that maybe there was something more important here that I had to understand important here that I had to understand important here that I had to understand and I had to interrupt and say, "No, and I had to interrupt and say, "No, and I had to interrupt and say, "No, that was another agent doing things that was another agent doing things that was another agent doing things wrong. Ignore that. Everything you've wrong. Ignore that. Everything you've wrong. Ignore that. Everything you've done so far is right. Just keep going."

  38. done so far is right. Just keep going." done so far is right. Just keep going." And it did. The only difference here And it did. The only difference here And it did. The only difference here between the prompts is that I sent the between the prompts is that I sent the between the prompts is that I sent the tail scale part as a third prompt and tail scale part as a third prompt and tail scale part as a third prompt and then I had to correct the model five then I had to correct the model five then I had to correct the model five additional times when I did it with additional times when I did it with additional times when I did it with Astra. Here I included that as part of Astra. Here I included that as part of Astra. Here I included that as part of the first prompt. Exact copy paste and the first prompt. Exact copy paste and the first prompt. Exact copy paste and it got it perfect first try. Exactly it got it perfect first try. Exactly it got it perfect first try. Exactly what I had intended and exactly what I what I had intended and exactly what I what I had intended and exactly what I had in my mind which is find where this had in my mind which is find where this had in my mind which is find where this went wrong, revert it and then give me a went wrong, revert it and then give me a went wrong, revert it and then give me a link I can click that actually lets me link I can click that actually lets me link I can click that actually lets me verify the changes. This shouldn't be verify the changes. This shouldn't be verify the changes. This shouldn't be that hard. I'll be real. If 56 Soul that hard. I'll be real. If 56 Soul that hard. I'll be real. If 56 Soul screwed this up, I would have been screwed this up, I would have been screwed this up, I would have been insulted. Astra blowing this one insulted. Astra blowing this one insulted. Astra blowing this one egregiously. One of the legitimate worst egregiously. One of the legitimate worst egregiously. One of the legitimate worst AI code experiences I've had this year AI code experiences I've had this year AI code experiences I've had this year without question. In fact, I'll say without question. In fact, I'll say without question. In fact, I'll say something bad. Discounting things like something bad. Discounting things like something bad. Discounting things like Flash 38 and like obviously meme tier Flash 38 and like obviously meme tier Flash 38 and like obviously meme tier models. Astra has had the most bad model models. Astra has had the most bad model models. Astra has had the most bad model experiences I have had of any model this experiences I have had of any model this experiences I have had of any model this year without question. Without question. year without question. Without question. year without question. Without question. It's just random like this. And It's just random like this. And It's just random like this. And it's not all the time. It's actually it's not all the time. It's actually it's not all the time. It's actually quite rare. But god damn, this model can quite rare. But god damn, this model can quite rare. But god damn, this model can just suck sometimes. I don't think Astra just suck sometimes. I don't think Astra just suck sometimes. I don't think Astra failed to understand my intent here. I failed to understand my intent here. I failed to understand my intent here. I just think it failed to do basic work as just think it failed to do basic work as just think it failed to do basic work as an agent. I see chat 50/50 on this here.

  39. an agent. I see chat 50/50 on this here. an agent. I see chat 50/50 on this here. Binary Shokan said that they haven't had Binary Shokan said that they haven't had Binary Shokan said that they haven't had to do as much feedback for any agent as to do as much feedback for any agent as to do as much feedback for any agent as they have for Astra so far by far. Yep. they have for Astra so far by far. Yep. they have for Astra so far by far. Yep. It's not always, but when it does It's not always, but when it does It's not always, but when it does happen, it is so ho stupid. I have an happen, it is so ho stupid. I have an happen, it is so ho stupid. I have an important thought that I need to pencil important thought that I need to pencil important thought that I need to pencil till later because it's more for like till later because it's more for like till later because it's more for like the summary at the end. So, we'll get the summary at the end. So, we'll get the summary at the end. So, we'll get there when we get there. Let's blast there when we get there. Let's blast there when we get there. Let's blast through the rest of these. Quick through the rest of these. Quick through the rest of these. Quick orchestration. It's Astra. Astra has orchestration. It's Astra. Astra has orchestration. It's Astra. Astra has this unbelievable new capability, best this unbelievable new capability, best this unbelievable new capability, best referred to as swarms. The way Fable referred to as swarms. The way Fable referred to as swarms. The way Fable does large numbers of sub agents is it does large numbers of sub agents is it does large numbers of sub agents is it plans up front. It decides, I want these plans up front. It decides, I want these plans up front. It decides, I want these four sub aents for this. I want this four sub aents for this. I want this four sub aents for this. I want this type of sub aent next. And if these four type of sub aent next. And if these four type of sub aent next. And if these four find things, they can spin up the next find things, they can spin up the next find things, they can spin up the next type of sub aent where it goes through type of sub aent where it goes through type of sub aent where it goes through steps one, two, three, four, five. steps one, two, three, four, five. steps one, two, three, four, five. Astra, however, can just spin up a bunch Astra, however, can just spin up a bunch Astra, however, can just spin up a bunch of sub aents that are doing whatever and of sub aents that are doing whatever and of sub aents that are doing whatever and then pass messages to them, let them then pass messages to them, let them then pass messages to them, let them pass messages to each other, and most pass messages to each other, and most pass messages to each other, and most importantly, it can receive updates from importantly, it can receive updates from importantly, it can receive updates from those sub agents and use that to fan those sub agents and use that to fan those sub agents and use that to fan things out to keep the whole swarm things out to keep the whole swarm things out to keep the whole swarm moving well in the right direction. I've moving well in the right direction. I've moving well in the right direction. I've never seen anything else like it. Astra never seen anything else like it. Astra never seen anything else like it. Astra has unlocked novel orchestration has unlocked novel orchestration has unlocked novel orchestration capabilities that given a model that capabilities that given a model that capabilities that given a model that wasn't as expensive would legitimately wasn't as expensive would legitimately wasn't as expensive would legitimately be the path to something like curing be the path to something like curing be the path to something like curing cancer. I genuinely believe that what cancer. I genuinely believe that what cancer. I genuinely believe that what OpenAI unlocked here is the start of the OpenAI unlocked here is the start of the OpenAI unlocked here is the start of the next era of what agents can potentially next era of what agents can potentially next era of what agents can potentially do in the size of problems that agents do in the size of problems that agents do in the size of problems that agents can legitimately solve. And I've seen it can legitimately solve. And I've seen it can legitimately solve. And I've seen it in action. It's a little harder to see

  40. in action. It's a little harder to see in action. It's a little harder to see now because tokens cost money again, but now because tokens cost money again, but now because tokens cost money again, but you can see some of it in this run I you can see some of it in this run I you can see some of it in this run I currently have going for the Rust currently have going for the Rust currently have going for the Rust rewrite of TypeScript. You'll see all rewrite of TypeScript. You'll see all rewrite of TypeScript. You'll see all the time these interacted with the time these interacted with the time these interacted with root/independent reviewer interacted root/independent reviewer interacted root/independent reviewer interacted with root/object members regression with root/object members regression with root/object members regression cause. It's able to manage the context cause. It's able to manage the context cause. It's able to manage the context back and forth pass messages and build back and forth pass messages and build back and forth pass messages and build with all of these agents. And I have 40 with all of these agents. And I have 40 with all of these agents. And I have 40 running here in parallel. And it can running here in parallel. And it can running here in parallel. And it can actually keep track of all of them and actually keep track of all of them and actually keep track of all of them and work with all of them. Unbelievable. work with all of them. Unbelievable. work with all of them. Unbelievable. Especially because workflows in cloud Especially because workflows in cloud Especially because workflows in cloud code are so good that I did a whole code are so good that I did a whole code are so good that I did a whole video about why I like cloud code video about why I like cloud code video about why I like cloud code largely around workflows. I never would largely around workflows. I never would largely around workflows. I never would have guessed that the more rudimentary have guessed that the more rudimentary have guessed that the more rudimentary implementation of sub aents in codeex implementation of sub aents in codeex implementation of sub aents in codeex was because they wanted the model to was because they wanted the model to was because they wanted the model to scale up and work through it. That's scale up and work through it. That's scale up and work through it. That's exactly what they wanted. It's exactly exactly what they wanted. It's exactly exactly what they wanted. It's exactly what it did and it does a great job. One what it did and it does a great job. One what it did and it does a great job. One of the capabilities that makes this of the capabilities that makes this of the capabilities that makes this possible is the steering side here. possible is the steering side here. possible is the steering side here. Astra is so good at getting random Astra is so good at getting random Astra is so good at getting random inserted in its context while it's inserted in its context while it's inserted in its context while it's working and not losing track of what working and not losing track of what working and not losing track of what it's doing. I cannot tell you how many it's doing. I cannot tell you how many it's doing. I cannot tell you how many times I had a model working on tasks times I had a model working on tasks times I had a model working on tasks one, two, three, and five and then I one, two, three, and five and then I one, two, three, and five and then I send, "Oh, I'm sorry. I forgot number send, "Oh, I'm sorry. I forgot number send, "Oh, I'm sorry. I forgot number four and then it does number four four and then it does number four four and then it does number four immediately and then never finishes the immediately and then never finishes the immediately and then never finishes the other one, two, three, and five tasks."

  41. other one, two, three, and five tasks." other one, two, three, and five tasks." That was just the default. I was used to That was just the default. I was used to That was just the default. I was used to that. Fable's better about this. It's or that. Fable's better about this. It's or that. Fable's better about this. It's or Fable 5.1 just doesn't do this too too Fable 5.1 just doesn't do this too too Fable 5.1 just doesn't do this too too much, which is nice. Astra eats this. It much, which is nice. Astra eats this. It much, which is nice. Astra eats this. It loves this. It begs you for more. One of loves this. It begs you for more. One of loves this. It begs you for more. One of the really cool things Astra does now the really cool things Astra does now the really cool things Astra does now when you're in a harness that is set up when you're in a harness that is set up when you're in a harness that is set up properly for it or an app with the right properly for it or an app with the right properly for it or an app with the right harness, for example, T3 Code, which harness, for example, T3 Code, which harness, for example, T3 Code, which funny enough, T3 Code actually does this funny enough, T3 Code actually does this funny enough, T3 Code actually does this feature better than Codeex, even though feature better than Codeex, even though feature better than Codeex, even though it is an Astra feature. It can ask it is an Astra feature. It can ask it is an Astra feature. It can ask questions while it works. So, it could questions while it works. So, it could questions while it works. So, it could be doing a thing and notice at some be doing a thing and notice at some be doing a thing and notice at some point while it's working, huh, I would point while it's working, huh, I would point while it's working, huh, I would like to know if the user's okay with me like to know if the user's okay with me like to know if the user's okay with me doing this. Or, huh, I wonder which of doing this. Or, huh, I wonder which of doing this. Or, huh, I wonder which of these three things the user would these three things the user would these three things the user would prefer. And it doesn't block itself. It prefer. And it doesn't block itself. It prefer. And it doesn't block itself. It keeps going. But at any point, you can keeps going. But at any point, you can keeps going. But at any point, you can answer. And if the model's already done, answer. And if the model's already done, answer. And if the model's already done, it will spin back up and address your it will spin back up and address your it will spin back up and address your answers. Or if it is still going, it can answers. Or if it is still going, it can answers. Or if it is still going, it can steer it in the right direction without steer it in the right direction without steer it in the right direction without interrupting the work it's doing. It's interrupting the work it's doing. It's interrupting the work it's doing. It's so cool. And it's like a meaningful so cool. And it's like a meaningful so cool. And it's like a meaningful behavioral change. we haven't had as behavioral change. we haven't had as behavioral change. we haven't had as many of recently. Like models work the many of recently. Like models work the many of recently. Like models work the way they work. We just have to prompt way they work. We just have to prompt way they work. We just have to prompt better. This is a change in how the better. This is a change in how the better. This is a change in how the model actually operates. That's really model actually operates. That's really model actually operates. That's really nice. When I was first testing, they nice. When I was first testing, they nice. When I was first testing, they didn't have this in codec and obviously didn't have this in codec and obviously didn't have this in codec and obviously I didn't have T3 code either. So I would I didn't have T3 code either. So I would I didn't have T3 code either. So I would just see it ask a question in its just see it ask a question in its just see it ask a question in its reasoning and then not be able to answer reasoning and then not be able to answer reasoning and then not be able to answer it and it would just keep going. Now it and it would just keep going. Now it and it would just keep going. Now with this, it's actually quite nice. So with this, it's actually quite nice. So with this, it's actually quite nice. So yeah, the steering stuff way better with yeah, the steering stuff way better with yeah, the steering stuff way better with Astro now. Fable was far ahead before.

  42. Astro now. Fable was far ahead before. Astro now. Fable was far ahead before. Soul would just get lost if you tried to Soul would just get lost if you tried to Soul would just get lost if you tried to steer it. Fable had a pretty solid steer it. Fable had a pretty solid steer it. Fable had a pretty solid default here. Astra has actual new default here. Astra has actual new default here. Astra has actual new capability unlocks for steering, which I capability unlocks for steering, which I capability unlocks for steering, which I think are cool as especially when think are cool as especially when think are cool as especially when you combine that with the orchestration you combine that with the orchestration you combine that with the orchestration because now it can have context injected because now it can have context injected because now it can have context injected from 40 other agents and handle that from 40 other agents and handle that from 40 other agents and handle that fine, which is awesome. And when you fine, which is awesome. And when you fine, which is awesome. And when you combine that with the self-prompting, combine that with the self-prompting, combine that with the self-prompting, which I will also say Astra is much which I will also say Astra is much which I will also say Astra is much better at here. Astra still writes worse better at here. Astra still writes worse better at here. Astra still writes worse prompts than I do, and I would say most prompts than I do, and I would say most prompts than I do, and I would say most devs who do this a lot do, but it can devs who do this a lot do, but it can devs who do this a lot do, but it can write a decent prompt. I've seen it write a decent prompt. I've seen it write a decent prompt. I've seen it write some pretty decent prompts. I'm write some pretty decent prompts. I'm write some pretty decent prompts. I'm much happier with Astra's ability to much happier with Astra's ability to much happier with Astra's ability to prompt itself and other agents. Nice. prompt itself and other agents. Nice. prompt itself and other agents. Nice. Not as big of a gap as the other things, Not as big of a gap as the other things, Not as big of a gap as the other things, but it is a gap. Cool to see. I'm but it is a gap. Cool to see. I'm but it is a gap. Cool to see. I'm excited for a future where agents excited for a future where agents excited for a future where agents actually understand things like a claude actually understand things like a claude actually understand things like a claude MD or a skill file well enough to write MD or a skill file well enough to write MD or a skill file well enough to write good verbiage for agents because right good verbiage for agents because right good verbiage for agents because right now it sucks and you quickly end up in a now it sucks and you quickly end up in a now it sucks and you quickly end up in a slop loop. If you let the agent write slop loop. If you let the agent write slop loop. If you let the agent write the skill, you let the skill be involved the skill, you let the skill be involved the skill, you let the skill be involved when it writes another skill, you end up when it writes another skill, you end up when it writes another skill, you end up in slop hell very fast. This helps avoid in slop hell very fast. This helps avoid in slop hell very fast. This helps avoid that. Astra still falls into it. Fable that. Astra still falls into it. Fable that. Astra still falls into it. Fable falls into it even more. Please write falls into it even more. Please write falls into it even more. Please write your skills by hand or at least audit your skills by hand or at least audit your skills by hand or at least audit them and make some nice changes.

  43. them and make some nice changes. them and make some nice changes. Self-prompting honestly fits under the Self-prompting honestly fits under the Self-prompting honestly fits under the skill writing stuff as well. Same gap skill writing stuff as well. Same gap skill writing stuff as well. Same gap there. there. there. Like none of them are good at it, but Like none of them are good at it, but Like none of them are good at it, but Astra is better at it slightly. And then Astra is better at it slightly. And then Astra is better at it slightly. And then we have skill using. This one is we have skill using. This one is we have skill using. This one is interesting because all models should be interesting because all models should be interesting because all models should be able to use skills totally fine, right? able to use skills totally fine, right? able to use skills totally fine, right? As I've crashed out about many times As I've crashed out about many times As I've crashed out about many times now, sadly, Astra has some skill issues. now, sadly, Astra has some skill issues. now, sadly, Astra has some skill issues. I don't know what's going on. I have I don't know what's going on. I have I don't know what's going on. I have learned recently that there is a system learned recently that there is a system learned recently that there is a system prompt section around how skill should prompt section around how skill should prompt section around how skill should be applied in codecs and part of it be applied in codecs and part of it be applied in codecs and part of it specifies that if a skill is used in one specifies that if a skill is used in one specifies that if a skill is used in one turn, that skill should not apply for turn, that skill should not apply for turn, that skill should not apply for future turns unless requested. In the future turns unless requested. In the future turns unless requested. In the example I gave in my previous videos and example I gave in my previous videos and example I gave in my previous videos and in the podcast where I told it to in the podcast where I told it to in the podcast where I told it to babysit a PR and then it stopped and babysit a PR and then it stopped and babysit a PR and then it stopped and then I asked it if there's things worth then I asked it if there's things worth then I asked it if there's things worth addressing and it said yes and then addressing and it said yes and then addressing and it said yes and then didn't do it. In that one, it had pulled didn't do it. In that one, it had pulled didn't do it. In that one, it had pulled in my babysitting skill and then just in my babysitting skill and then just in my babysitting skill and then just pretended it didn't exist. Even though pretended it didn't exist. Even though pretended it didn't exist. Even though it was still in context, it just ignored it was still in context, it just ignored it was still in context, it just ignored it for the next five follow-ups. It it for the next five follow-ups. It it for the next five follow-ups. It sucks. I never ever ever have to think sucks. I never ever ever have to think sucks. I never ever ever have to think about this with Fable. If the skill has about this with Fable. If the skill has about this with Fable. If the skill has a reasonable description and it's useful a reasonable description and it's useful a reasonable description and it's useful for the work going on, Fable correctly for the work going on, Fable correctly for the work going on, Fable correctly pulls it in and applies it. It just pulls it in and applies it. It just pulls it in and applies it. It just does. I don't know what the is does. I don't know what the is does. I don't know what the is wrong with Astra for this. And I wrong with Astra for this. And I wrong with Astra for this. And I honestly think a lot of it is codeex.

  44. honestly think a lot of it is codeex. honestly think a lot of it is codeex. Astra doesn't feel like it understands Astra doesn't feel like it understands Astra doesn't feel like it understands my skills anymore in a lot of ways. my skills anymore in a lot of ways. my skills anymore in a lot of ways. Fable 5.1. It does. It absolutely does. Fable 5.1. It does. It absolutely does. Fable 5.1. It does. It absolutely does. So yeah, skill writing. Astra has a So yeah, skill writing. Astra has a So yeah, skill writing. Astra has a slight lead. Skill using. Fable has a slight lead. Skill using. Fable has a slight lead. Skill using. Fable has a huge lead. Now we have honoring refusals huge lead. Now we have honoring refusals huge lead. Now we have honoring refusals and boundaries. And as you have already and boundaries. And as you have already and boundaries. And as you have already seen, Astra is forgetful. It's okay at seen, Astra is forgetful. It's okay at seen, Astra is forgetful. It's okay at honoring things when you tell it to, but honoring things when you tell it to, but honoring things when you tell it to, but it it might forget. If you tell it to it it might forget. If you tell it to it it might forget. If you tell it to keep it in context, it usually will, but keep it in context, it usually will, but keep it in context, it usually will, but it still is quite forgetful. And it is it still is quite forgetful. And it is it still is quite forgetful. And it is still quite frustrating when it just still quite frustrating when it just still quite frustrating when it just does a thing it's not supposed to. Fable does a thing it's not supposed to. Fable does a thing it's not supposed to. Fable isn't as good at strictly following isn't as good at strictly following isn't as good at strictly following instructions, but that also means that instructions, but that also means that instructions, but that also means that when it doesn't follow them, it doesn't when it doesn't follow them, it doesn't when it doesn't follow them, it doesn't feel as egregious. Both are still not feel as egregious. Both are still not feel as egregious. Both are still not where I want them to be here, especially where I want them to be here, especially where I want them to be here, especially for the levels of intelligence. I would for the levels of intelligence. I would for the levels of intelligence. I would say Fable is ahead. Astra looks and say Fable is ahead. Astra looks and say Fable is ahead. Astra looks and generally acts better about this, but generally acts better about this, but generally acts better about this, but its failures are more egregious, which its failures are more egregious, which its failures are more egregious, which is why the hugging face hack happened. is why the hugging face hack happened. is why the hugging face hack happened. So, take that as you will. I have more So, take that as you will. I have more So, take that as you will. I have more to say about this and also all the code to say about this and also all the code to say about this and also all the code stuff. We need to talk about cost. I've stuff. We need to talk about cost. I've stuff. We need to talk about cost. I've seen some very dumb takes on cost here, seen some very dumb takes on cost here, seen some very dumb takes on cost here, like exceptionally dumb ones. First and like exceptionally dumb ones. First and like exceptionally dumb ones. First and foremost, token costs. They are the same foremost, token costs. They are the same foremost, token costs. They are the same except for one important exception, except for one important exception, except for one important exception, which is that Fable 5.1 has a huge drop which is that Fable 5.1 has a huge drop which is that Fable 5.1 has a huge drop in cash read costs. They went from a in cash read costs. They went from a in cash read costs. They went from a dollar per mill in for cash read to 25 dollar per mill in for cash read to 25 dollar per mill in for cash read to 25 cents per mill cash read. And that's cents per mill cash read. And that's cents per mill cash read. And that's awesome. That genuinely is. And it would awesome. That genuinely is. And it would awesome. That genuinely is. And it would be a lot more awesome if cash reads were be a lot more awesome if cash reads were be a lot more awesome if cash reads were more than 10% of my costs. They are

  45. more than 10% of my costs. They are more than 10% of my costs. They are closer to 3%. And now they are 1%. closer to 3%. And now they are 1%. closer to 3%. And now they are 1%. Awesome. Do you know what is much more Awesome. Do you know what is much more Awesome. Do you know what is much more than that? Cash rights. Cash write costs than that? Cash rights. Cash write costs than that? Cash rights. Cash write costs are over 60% of my costs with Fable. So are over 60% of my costs with Fable. So are over 60% of my costs with Fable. So while it has made cash reads hilariously while it has made cash reads hilariously while it has made cash reads hilariously cheap, it has also made cash rights cheap, it has also made cash rights cheap, it has also made cash rights hilariously expensive. I am spending hilariously expensive. I am spending hilariously expensive. I am spending more money asking Anthropic to save the more money asking Anthropic to save the more money asking Anthropic to save the state of the model on their servers for state of the model on their servers for state of the model on their servers for five minutes than I'm spending actually five minutes than I'm spending actually five minutes than I'm spending actually running the GPUs. I am paying Anthropic running the GPUs. I am paying Anthropic running the GPUs. I am paying Anthropic more money to manage RAM for me than I more money to manage RAM for me than I more money to manage RAM for me than I am paying them to run compute for me. am paying them to run compute for me. am paying them to run compute for me. And if you don't think that's absurd, I And if you don't think that's absurd, I And if you don't think that's absurd, I don't know what to tell you. But I'm don't know what to tell you. But I'm don't know what to tell you. But I'm very excited for the cash write very excited for the cash write very excited for the cash write revolution that has to happen. And revolution that has to happen. And revolution that has to happen. And paying for RAM storage effectively is so paying for RAM storage effectively is so paying for RAM storage effectively is so stupid. And I am sure we will fix this stupid. And I am sure we will fix this stupid. And I am sure we will fix this soon. We're going to get worse before we soon. We're going to get worse before we soon. We're going to get worse before we get better because OpenAI didn't used to get better because OpenAI didn't used to get better because OpenAI didn't used to charge for cash rights. They used to be charge for cash rights. They used to be charge for cash rights. They used to be free. Now they do. And they are not free. Now they do. And they are not free. Now they do. And they are not cheap. It is 25% more expensive than a cheap. It is 25% more expensive than a cheap. It is 25% more expensive than a normal read. So if reading is $10 per normal read. So if reading is $10 per normal read. So if reading is $10 per million, cash reads are a dollar per million, cash reads are a dollar per million, cash reads are a dollar per million for OpenAI and they are 25 cents million for OpenAI and they are 25 cents million for OpenAI and they are 25 cents per million for Fable 5.1. Cash rights per million for Fable 5.1. Cash rights per million for Fable 5.1. Cash rights are 1250. You're going to spend a lot of are 1250. You're going to spend a lot of are 1250. You're going to spend a lot of money writing cash on this model. I money writing cash on this model. I money writing cash on this model. I promise you on both of them even. So promise you on both of them even. So promise you on both of them even. So know that going in. The cash read is it know that going in. The cash read is it know that going in. The cash read is it turned a 3% cost down to one. But cash turned a 3% cost down to one. But cash turned a 3% cost down to one. But cash rights are where the money actually rights are where the money actually rights are where the money actually matters. So if either of these labs make matters. So if either of these labs make matters. So if either of these labs make cash rights cheap or free, they win cost

  46. cash rights cheap or free, they win cost cash rights cheap or free, they win cost by default by far. But the actual token by default by far. But the actual token by default by far. But the actual token cost barely even matters in a world cost barely even matters in a world cost barely even matters in a world driven by token efficiency. God, you driven by token efficiency. God, you driven by token efficiency. God, you guys, I I literally just explained this guys, I I literally just explained this guys, I I literally just explained this in chat. is doing. Guys, guys, guys, in chat. is doing. Guys, guys, guys, in chat. is doing. Guys, guys, guys, Bill, you're better than this. You're Bill, you're better than this. You're Bill, you're better than this. You're one of like the more important devs we one of like the more important devs we one of like the more important devs we have on T3 code. What did I just say? have on T3 code. What did I just say? have on T3 code. What did I just say? The difference in price for cash reads The difference in price for cash reads The difference in price for cash reads looks really big because a dollar is looks really big because a dollar is looks really big because a dollar is four times more than 25, but when it four times more than 25, but when it four times more than 25, but when it adds up to less than 5% of your spend, adds up to less than 5% of your spend, adds up to less than 5% of your spend, it doesn't matter. It's like saying that it doesn't matter. It's like saying that it doesn't matter. It's like saying that Mac OS is a 100 times faster than Mac OS is a 100 times faster than Mac OS is a 100 times faster than Windows because it gets through the Windows because it gets through the Windows because it gets through the splash screen at boot 30 milliseconds splash screen at boot 30 milliseconds splash screen at boot 30 milliseconds faster. You're you're shaving a faster. You're you're shaving a faster. You're you're shaving a percentage off a percentage. It doesn't percentage off a percentage. It doesn't percentage off a percentage. It doesn't matter. Four times cheaper for 3% of matter. Four times cheaper for 3% of matter. Four times cheaper for 3% of your cost is a less than 1% deal. It your cost is a less than 1% deal. It your cost is a less than 1% deal. It just doesn't matter. It's fine, Bill. just doesn't matter. It's fine, Bill. just doesn't matter. It's fine, Bill. You weren't here for it. But I really You weren't here for it. But I really You weren't here for it. But I really need to emphasize this point cuz I've need to emphasize this point cuz I've need to emphasize this point cuz I've seen some incredibly smart people say seen some incredibly smart people say seen some incredibly smart people say some incredibly stupid about this some incredibly stupid about this some incredibly stupid about this because they just haven't looked at the because they just haven't looked at the because they just haven't looked at the numbers. And if you need proof, it's numbers. And if you need proof, it's numbers. And if you need proof, it's pretty easy to find. Cost per task on pretty easy to find. Cost per task on pretty easy to find. Cost per task on artificial analysis intelligence index.

  47. artificial analysis intelligence index. artificial analysis intelligence index. GB6 Astro was $3.26 per task. Opus 5 GB6 Astro was $3.26 per task. Opus 5 GB6 Astro was $3.26 per task. Opus 5 almost $6. Fable 5.1 $7.60. almost $6. Fable 5.1 $7.60. almost $6. Fable 5.1 $7.60. That is a comical gap even though they That is a comical gap even though they That is a comical gap even though they are priced the same. And theoretically are priced the same. And theoretically are priced the same. And theoretically speaking, the input token cost is four speaking, the input token cost is four speaking, the input token cost is four times higher on Astra. It is still a times higher on Astra. It is still a times higher on Astra. It is still a fourth the price in most real world work fourth the price in most real world work fourth the price in most real world work because it is so much more token because it is so much more token because it is so much more token efficient and cash reads are such a efficient and cash reads are such a efficient and cash reads are such a small percentage of your costs. That small percentage of your costs. That small percentage of your costs. That said, Astra is 326 and Soul was 199. So said, Astra is 326 and Soul was 199. So said, Astra is 326 and Soul was 199. So Astra is a meaningful increase in price Astra is a meaningful increase in price Astra is a meaningful increase in price compared to Soul, but it's still way compared to Soul, but it's still way compared to Soul, but it's still way cheaper than all of what Anthropic has cheaper than all of what Anthropic has cheaper than all of what Anthropic has been doing. Also, funny enough, been doing. Also, funny enough, been doing. Also, funny enough, previously they had said Fable 5.1 was previously they had said Fable 5.1 was previously they had said Fable 5.1 was more expensive than Fable 5, but when more expensive than Fable 5, but when more expensive than Fable 5, but when they redid the artificial analysis like they redid the artificial analysis like they redid the artificial analysis like index because they were getting cooked index because they were getting cooked index because they were getting cooked because it just was not measuring things because it just was not measuring things because it just was not measuring things well at all anymore. When they switched well at all anymore. When they switched well at all anymore. When they switched to the new index, 5.1 actually got to the new index, 5.1 actually got to the new index, 5.1 actually got cheaper than Fable 5 because the cash cheaper than Fable 5 because the cash cheaper than Fable 5 because the cash read difference finally actually read difference finally actually read difference finally actually mattered. I made the points I want to mattered. I made the points I want to mattered. I made the points I want to make here. D6 Astra is still way cheaper make here. D6 Astra is still way cheaper make here. D6 Astra is still way cheaper even though the prices are the same and even though the prices are the same and even though the prices are the same and the cash cost is higher because in the the cash cost is higher because in the the cash cost is higher because in the end there's a lot of other things that end there's a lot of other things that end there's a lot of other things that matter and that number is not where the matter and that number is not where the matter and that number is not where the cost is happening. This number is where cost is happening. This number is where cost is happening. This number is where the cost is happening. Token efficiency.

  48. the cost is happening. Token efficiency. the cost is happening. Token efficiency. GBD6 Astra is one of the most token GBD6 Astra is one of the most token GBD6 Astra is one of the most token efficient models that they've ever efficient models that they've ever efficient models that they've ever benched. I removed pretty much benched. I removed pretty much benched. I removed pretty much everything here that inarguably does not everything here that inarguably does not everything here that inarguably does not matter. And Astra is the most token matter. And Astra is the most token matter. And Astra is the most token efficient by far. Meanwhile, Fable 5.1 efficient by far. Meanwhile, Fable 5.1 efficient by far. Meanwhile, Fable 5.1 is the least by far. Astra did in 27K is the least by far. Astra did in 27K is the least by far. Astra did in 27K tokens what Fable 5.1 did in almost 80K tokens what Fable 5.1 did in almost 80K tokens what Fable 5.1 did in almost 80K tokens. Take it as you will. It's also tokens. Take it as you will. It's also tokens. Take it as you will. It's also worth noting that OpenAI offers a flex worth noting that OpenAI offers a flex worth noting that OpenAI offers a flex option which cuts the prices in half but option which cuts the prices in half but option which cuts the prices in half but kills all the guarantees for throughput kills all the guarantees for throughput kills all the guarantees for throughput because it's flex so it doesn't start because it's flex so it doesn't start because it's flex so it doesn't start responding immediately. It just happens responding immediately. It just happens responding immediately. It just happens eventually. So if you're willing to let eventually. So if you're willing to let eventually. So if you're willing to let your jobs take an unknown amount of your jobs take an unknown amount of your jobs take an unknown amount of time, you can put on the flex end points time, you can put on the flex end points time, you can put on the flex end points and let it run when they have compute and let it run when they have compute and let it run when they have compute around and it will cut your costs even around and it will cut your costs even around and it will cut your costs even further in half. Now that I think about further in half. Now that I think about further in half. Now that I think about it, it would actually be nice if they it, it would actually be nice if they it, it would actually be nice if they could add that as an option for using could add that as an option for using could add that as an option for using with codecs because I don't mind if my with codecs because I don't mind if my with codecs because I don't mind if my threads take a while and I also am aware threads take a while and I also am aware threads take a while and I also am aware that a lot of the time I'm using by that a lot of the time I'm using by that a lot of the time I'm using by agents is at bad times for businesses. agents is at bad times for businesses. agents is at bad times for businesses. So since I'm working at like 9:00 p.m. So since I'm working at like 9:00 p.m. So since I'm working at like 9:00 p.m. and not like 10 a.m. when businesses and not like 10 a.m. when businesses and not like 10 a.m. when businesses are, it would probably be cheaper for are, it would probably be cheaper for are, it would probably be cheaper for OpenAI and me. Just saying it is what it OpenAI and me. Just saying it is what it OpenAI and me. Just saying it is what it is. But that means we're talking about is. But that means we're talking about is. But that means we're talking about subscription limits. And I think it is subscription limits. And I think it is subscription limits. And I think it is fair to say I am uh pretty experienced fair to say I am uh pretty experienced fair to say I am uh pretty experienced with the limits on all of these things.

  49. with the limits on all of these things. with the limits on all of these things. Being that I have five accounts with Being that I have five accounts with Being that I have five accounts with Claude and that I have four with Codeex, Claude and that I have four with Codeex, Claude and that I have four with Codeex, think I have a pretty good gut feel think I have a pretty good gut feel think I have a pretty good gut feel here. And to be completely frank, if you here. And to be completely frank, if you here. And to be completely frank, if you are concerned about maximizing how much are concerned about maximizing how much are concerned about maximizing how much return you get for the dollars you're return you get for the dollars you're return you get for the dollars you're putting in and you're doing anything putting in and you're doing anything putting in and you're doing anything other than the codec subscription, I other than the codec subscription, I other than the codec subscription, I would love to meet the person who would love to meet the person who would love to meet the person who convinced you Claude was a better deal convinced you Claude was a better deal convinced you Claude was a better deal so I can hire them. I always need good so I can hire them. I always need good so I can hire them. I always need good sales people and I could really use that sales people and I could really use that sales people and I could really use that one because they sold you a one because they sold you a one because they sold you a bridge, man. Seriously, the gap in what bridge, man. Seriously, the gap in what bridge, man. Seriously, the gap in what you get is insane. You might be a bit you get is insane. You might be a bit you get is insane. You might be a bit confused because I've covered the confused because I've covered the confused because I've covered the numbers before. And with Anthropic and numbers before. And with Anthropic and numbers before. And with Anthropic and Claude Code plans, you get around 8 Claude Code plans, you get around 8 Claude Code plans, you get around 8 grand of inference for 200 bucks. And grand of inference for 200 bucks. And grand of inference for 200 bucks. And with codecs, you get around 12 grand for with codecs, you get around 12 grand for with codecs, you get around 12 grand for 200 bucks, which is better, but that's 200 bucks, which is better, but that's 200 bucks, which is better, but that's only like a 50% improvement. Well, only like a 50% improvement. Well, only like a 50% improvement. Well, there's a few factors you have to there's a few factors you have to there's a few factors you have to consider. First, we're have to take a consider. First, we're have to take a consider. First, we're have to take a huge cut in our quad code limits because huge cut in our quad code limits because huge cut in our quad code limits because they're going to drop off that 50% boost they're going to drop off that 50% boost they're going to drop off that 50% boost they're giving us. They're going to keep they're giving us. They're going to keep they're giving us. They're going to keep at 25%. So, we're losing around 17% off at 25%. So, we're losing around 17% off at 25%. So, we're losing around 17% off the top of our limits. But much worse, the top of our limits. But much worse, the top of our limits. But much worse, the fable limit. You only get to use the fable limit. You only get to use the fable limit. You only get to use half your usage for fable. You might half your usage for fable. You might half your usage for fable. You might notice in my UI here that there's a notice in my UI here that there's a notice in my UI here that there's a third column for the claude section that third column for the claude section that third column for the claude section that is faded out a bit. I call this the opus is faded out a bit. I call this the opus is faded out a bit. I call this the opus pot. It's a whole steaming smelly pot of pot. It's a whole steaming smelly pot of pot. It's a whole steaming smelly pot of opus. And that's all it is. And when I opus. And that's all it is. And when I opus. And that's all it is. And when I am heavily coding, what ends up am heavily coding, what ends up am heavily coding, what ends up happening is the column on the left all happening is the column on the left all happening is the column on the left all go to zero and the column on the right

  50. go to zero and the column on the right go to zero and the column on the right all get stuck at 50 because I'm allowed all get stuck at 50 because I'm allowed all get stuck at 50 because I'm allowed to use half my weekly limit for Fable to use half my weekly limit for Fable to use half my weekly limit for Fable and once I'm out of Fable, the account and once I'm out of Fable, the account and once I'm out of Fable, the account is useless to me because Opus sucks and is useless to me because Opus sucks and is useless to me because Opus sucks and Sonnet's hilariously bad. So I end up Sonnet's hilariously bad. So I end up Sonnet's hilariously bad. So I end up every month losing those 50%s at the end every month losing those 50%s at the end every month losing those 50%s at the end of each week, at the end of each reset. of each week, at the end of each reset. of each week, at the end of each reset. And since I get eight grand a month of And since I get eight grand a month of And since I get eight grand a month of allocation allocation allocation per account, I'm only getting four grand per account, I'm only getting four grand per account, I'm only getting four grand of that with Fable, which means it's of that with Fable, which means it's of that with Fable, which means it's actually a third of what you get with actually a third of what you get with actually a third of what you get with Astra in the $200 CEX plan. You get 50% Astra in the $200 CEX plan. You get 50% Astra in the $200 CEX plan. You get 50% of 8 grand with Claude Code and you get of 8 grand with Claude Code and you get of 8 grand with Claude Code and you get 100% of the 12 grand with Codeex. And 100% of the 12 grand with Codeex. And 100% of the 12 grand with Codeex. And they made further improvements to how they made further improvements to how they made further improvements to how Astra is being hosted and how it uses Astra is being hosted and how it uses Astra is being hosted and how it uses your limits. Tibo claimed is up to a 3 your limits. Tibo claimed is up to a 3 your limits. Tibo claimed is up to a 3 to 4x difference. And I don't to 4x difference. And I don't to 4x difference. And I don't necessarily believe that, but I've seen necessarily believe that, but I've seen necessarily believe that, but I've seen a meaningful difference. Like I have a meaningful difference. Like I have a meaningful difference. Like I have some hellish health threads going right some hellish health threads going right some hellish health threads going right now. I have two ult running with now. I have two ult running with now. I have two ult running with unbounded sub agents. It's 40 sub agents unbounded sub agents. It's 40 sub agents unbounded sub agents. It's 40 sub agents each can run. And it's going down. I each can run. And it's going down. I each can run. And it's going down. I think it's hitting this particular think it's hitting this particular think it's hitting this particular account right now. And it's at 82%. It's account right now. And it's at 82%. It's account right now. And it's at 82%. It's been going for the last 8 minutes since been going for the last 8 minutes since been going for the last 8 minutes since I refreshed. Not even. And it's down one I refreshed. Not even. And it's down one I refreshed. Not even. And it's down one more percent. Cool. if you weren't doing more percent. Cool. if you weren't doing more percent. Cool. if you weren't doing the absurd, sinful, wasteful background the absurd, sinful, wasteful background the absurd, sinful, wasteful background jobs I am running, trying to dick around jobs I am running, trying to dick around jobs I am running, trying to dick around with my new fork slash uh decomp of with my new fork slash uh decomp of with my new fork slash uh decomp of Super Smash Bros. Melee while also Super Smash Bros. Melee while also Super Smash Bros. Melee while also trying to port the entirety of trying to port the entirety of trying to port the entirety of TypeScript to Rust. It's not that bad.

  51. TypeScript to Rust. It's not that bad. TypeScript to Rust. It's not that bad. You can hit the limits and you should be You can hit the limits and you should be You can hit the limits and you should be very careful of the alter button and in very careful of the alter button and in very careful of the alter button and in particular the fast button. Straight up, particular the fast button. Straight up, particular the fast button. Straight up, fast is not the way you should use this fast is not the way you should use this fast is not the way you should use this model. If you're using fast with Astro, model. If you're using fast with Astro, model. If you're using fast with Astro, you're probably using Astro wrong or you're probably using Astro wrong or you're probably using Astro wrong or you're just burning tokens for fun, you're just burning tokens for fun, you're just burning tokens for fun, which I understand. But like if you're which I understand. But like if you're which I understand. But like if you're actually finding yourself reaching for actually finding yourself reaching for actually finding yourself reaching for fast all the time, you're not using it fast all the time, you're not using it fast all the time, you're not using it for its strengths. You're not using for its strengths. You're not using for its strengths. You're not using Astra for what it's good at. But you can Astra for what it's good at. But you can Astra for what it's good at. But you can get a pretty good amount of usage out of get a pretty good amount of usage out of get a pretty good amount of usage out of these plans. Not great, but a good these plans. Not great, but a good these plans. Not great, but a good amount. There is a nice powerful catch amount. There is a nice powerful catch amount. There is a nice powerful catch to this we'll get to in a sec. But I to this we'll get to in a sec. But I to this we'll get to in a sec. But I have one last thing I need to say about have one last thing I need to say about have one last thing I need to say about the Claude stuff that has begun to piss the Claude stuff that has begun to piss the Claude stuff that has begun to piss me off a lot. Openai does not do five me off a lot. Openai does not do five me off a lot. Openai does not do five hour limits on the pro plan. So the $100 hour limits on the pro plan. So the $100 hour limits on the pro plan. So the $100 and $200 tiers do not get fivehour and $200 tiers do not get fivehour and $200 tiers do not get fivehour limits. They only have weekly limits. limits. They only have weekly limits. limits. They only have weekly limits. All Claude plans still have 5-hour All Claude plans still have 5-hour All Claude plans still have 5-hour limits. Each 5-hour limit gets you about limits. Each 5-hour limit gets you about limits. Each 5-hour limit gets you about 40% of your Fable 5 limit, which is 20% 40% of your Fable 5 limit, which is 20% 40% of your Fable 5 limit, which is 20% of your weekly. So, you can clear the 5 of your weekly. So, you can clear the 5 of your weekly. So, you can clear the 5 hour on your Claude plan five times hour on your Claude plan five times hour on your Claude plan five times before being out of weekly limits, before being out of weekly limits, before being out of weekly limits, unless you're using Fable and then unless you're using Fable and then unless you're using Fable and then you'll run out in two and a half. The you'll run out in two and a half. The you'll run out in two and a half. The reason I set up CLI proxy is because one reason I set up CLI proxy is because one reason I set up CLI proxy is because one cla account just isn't enough at all for cla account just isn't enough at all for cla account just isn't enough at all for most work. If you're using it a few most work. If you're using it a few most work. If you're using it a few hours a day, two to three days a week, hours a day, two to three days a week, hours a day, two to three days a week, and you're not going too hard, and and you're not going too hard, and and you're not going too hard, and you're not using a lot of sub agents, you're not using a lot of sub agents, you're not using a lot of sub agents, you can probably get by on a single $200 you can probably get by on a single $200 you can probably get by on a single $200 plan with Claude. The only way a $200 plan with Claude. The only way a $200 plan with Claude. The only way a $200 Codeex plan isn't enough is if you are Codeex plan isn't enough is if you are Codeex plan isn't enough is if you are like using higher reasoning limits than like using higher reasoning limits than like using higher reasoning limits than necessary and not paying much attention.

  52. necessary and not paying much attention. necessary and not paying much attention. Most devs, when used responsibly, could Most devs, when used responsibly, could Most devs, when used responsibly, could absolutely get away with the $200 plan absolutely get away with the $200 plan absolutely get away with the $200 plan with Codex without issue. It would be with Codex without issue. It would be with Codex without issue. It would be genuinely hard to get away with it with genuinely hard to get away with it with genuinely hard to get away with it with the Fable plan. I even call it the Fable the Fable plan. I even call it the Fable the Fable plan. I even call it the Fable plan because that's what it is to me. plan because that's what it is to me. plan because that's what it is to me. It'd be much harder to get away with it It'd be much harder to get away with it It'd be much harder to get away with it on the Claude code plan because of the on the Claude code plan because of the on the Claude code plan because of the additional limits on Fable, the much additional limits on Fable, the much additional limits on Fable, the much less efficient usage of Fable and less efficient usage of Fable and less efficient usage of Fable and everything else. So, Astra is more token everything else. So, Astra is more token everything else. So, Astra is more token efficient. It costs less, so it uses efficient. It costs less, so it uses efficient. It costs less, so it uses your usage less. It has way more your usage less. It has way more your usage less. It has way more generous limits in general with Codeex. generous limits in general with Codeex. generous limits in general with Codeex. It's just like I would say it's like a It's just like I would say it's like a It's just like I would say it's like a 4xish gap between the two is how it 4xish gap between the two is how it 4xish gap between the two is how it feels to me. feel like I can get four feels to me. feel like I can get four feels to me. feel like I can get four times more done with Codeex on Astra times more done with Codeex on Astra times more done with Codeex on Astra than I can with Fable and Claude Code on than I can with Fable and Claude Code on than I can with Fable and Claude Code on the same $200 a month tiers. But then the same $200 a month tiers. But then the same $200 a month tiers. But then things get even more ridiculous because things get even more ridiculous because things get even more ridiculous because of the resets. It's become a meme. Tibo of the resets. It's become a meme. Tibo of the resets. It's become a meme. Tibo just throws them out for fun now. In just throws them out for fun now. In just throws them out for fun now. In fact, when they were delayed trying to fact, when they were delayed trying to fact, when they were delayed trying to get Astra out, they sent us two bananked get Astra out, they sent us two bananked get Astra out, they sent us two bananked resets. I actually opened this account resets. I actually opened this account resets. I actually opened this account that I don't even need yet because I that I don't even need yet because I that I don't even need yet because I wanted to collect those bananked resets wanted to collect those bananked resets wanted to collect those bananked resets so I'd have them ready to go when I did so I'd have them ready to go when I did so I'd have them ready to go when I did need them. So, your limits just get need them. So, your limits just get need them. So, your limits just get reset all the time. They got reset while reset all the time. They got reset while reset all the time. They got reset while I was filming earlier. So, even if you I was filming earlier. So, even if you I was filming earlier. So, even if you do hit the limits on Codeex, you might do hit the limits on Codeex, you might do hit the limits on Codeex, you might just randomly get a reset, which is just randomly get a reset, which is just randomly get a reset, which is quite nice. So, yeah, it does happen quite nice. So, yeah, it does happen quite nice. So, yeah, it does happen with Claude. It's just so much rarer with Claude. It's just so much rarer with Claude. It's just so much rarer that it you can't really count on it at that it you can't really count on it at that it you can't really count on it at all. But with Codex, the resets giving all. But with Codex, the resets giving all. But with Codex, the resets giving you your limits back much nicer.

  53. you your limits back much nicer. you your limits back much nicer. According to Raphael, there were eight According to Raphael, there were eight According to Raphael, there were eight free resets in the last 30 days. So, the free resets in the last 30 days. So, the free resets in the last 30 days. So, the weekly limit is more like a 3-day limit weekly limit is more like a 3-day limit weekly limit is more like a 3-day limit on average. Pretty nuts. Somebody on average. Pretty nuts. Somebody on average. Pretty nuts. Somebody accused him of maliciously resetting 46 accused him of maliciously resetting 46 accused him of maliciously resetting 46 hours after the previous global reset in hours after the previous global reset in hours after the previous global reset in the time frame around where or in the the time frame around where or in the the time frame around where or in the time frame around where a banked reset time frame around where a banked reset time frame around where a banked reset was landed. This is not true at all. was landed. This is not true at all. was landed. This is not true at all. I've actually seen Tibo intentionally I've actually seen Tibo intentionally I've actually seen Tibo intentionally delay a reset in order to prevent that. delay a reset in order to prevent that. delay a reset in order to prevent that. And even just now, he announced the And even just now, he announced the And even just now, he announced the reset many hours before so we got to go reset many hours before so we got to go reset many hours before so we got to go spin up our furnaces and burn a bunch of spin up our furnaces and burn a bunch of spin up our furnaces and burn a bunch of tokens. He does not maliciously align tokens. He does not maliciously align tokens. He does not maliciously align the times there at all. He actually does the times there at all. He actually does the times there at all. He actually does the opposite. So yeah, if cost is a the opposite. So yeah, if cost is a the opposite. So yeah, if cost is a concern, wait a few months because this concern, wait a few months because this concern, wait a few months because this level of intelligence will be accessible level of intelligence will be accessible level of intelligence will be accessible at a much cheaper price. But if cost is at a much cheaper price. But if cost is at a much cheaper price. But if cost is a concern and you want the best class a concern and you want the best class a concern and you want the best class stuff right now, you can get a stuff right now, you can get a stuff right now, you can get a reasonable deal with a $200 plan with reasonable deal with a $200 plan with reasonable deal with a $200 plan with codec still. If you're curious about the codec still. If you're curious about the codec still. If you're curious about the $20 plans and the $100 plans, my $20 plans and the $100 plans, my $20 plans and the $100 plans, my suggestion would be wait for this level suggestion would be wait for this level suggestion would be wait for this level of intelligence to get cheaper because of intelligence to get cheaper because of intelligence to get cheaper because this level of intelligence is expensive.

  54. this level of intelligence is expensive. this level of intelligence is expensive. I'm legitimately doing 30 to $40,000 of I'm legitimately doing 30 to $40,000 of I'm legitimately doing 30 to $40,000 of inference a month on all of my plans. inference a month on all of my plans. inference a month on all of my plans. consistently now, even without early consistently now, even without early consistently now, even without early access testing that's free. They are not access testing that's free. They are not access testing that's free. They are not cheap. Astra is cheaper than Fable, but cheap. Astra is cheaper than Fable, but cheap. Astra is cheaper than Fable, but neither are cheap, and you need to know neither are cheap, and you need to know neither are cheap, and you need to know that going in because if you go in with that going in because if you go in with that going in because if you go in with different expectations, you will be different expectations, you will be different expectations, you will be upset. They are both expensive. They upset. They are both expensive. They upset. They are both expensive. They will both burn limits fast, but Astra is will both burn limits fast, but Astra is will both burn limits fast, but Astra is more efficient when you compare the two more efficient when you compare the two more efficient when you compare the two directly. So, with all of this said, directly. So, with all of this said, directly. So, with all of this said, let's answer the question. Which model let's answer the question. Which model let's answer the question. Which model is better? I could only pick one. Which is better? I could only pick one. Which is better? I could only pick one. Which would I have? Let's say orange is fable would I have? Let's say orange is fable would I have? Let's say orange is fable and I don't know, we'll say blue for and I don't know, we'll say blue for and I don't know, we'll say blue for Astra. Cool. This is meant to be the Astra. Cool. This is meant to be the Astra. Cool. This is meant to be the quality of the outputs over time, over quality of the outputs over time, over quality of the outputs over time, over attempts, whatever. This is generally attempts, whatever. This is generally attempts, whatever. This is generally speaking, what's the quality I get out speaking, what's the quality I get out speaking, what's the quality I get out when I send a prompt with Fable 5.1. when I send a prompt with Fable 5.1. when I send a prompt with Fable 5.1. It's a relatively stable line at a It's a relatively stable line at a It's a relatively stable line at a relatively high bar. It has its moments relatively high bar. It has its moments relatively high bar. It has its moments where it impresses me and it has its where it impresses me and it has its where it impresses me and it has its moments where it disappoints me a bit.

  55. moments where it disappoints me a bit. moments where it disappoints me a bit. Occasionally it gets pretty rough, but Occasionally it gets pretty rough, but Occasionally it gets pretty rough, but for the most part, it stays in this for the most part, it stays in this for the most part, it stays in this general range. Quality I'm getting out general range. Quality I'm getting out general range. Quality I'm getting out of Fable pretty damn good, and I would of Fable pretty damn good, and I would of Fable pretty damn good, and I would feel bad complaining about it. It does feel bad complaining about it. It does feel bad complaining about it. It does have its weird spikes, and I'll have have its weird spikes, and I'll have have its weird spikes, and I'll have plenty of videos where I show the times plenty of videos where I show the times plenty of videos where I show the times Fable piss me off. Don't worry. But Fable piss me off. Don't worry. But Fable piss me off. Don't worry. But generally speaking, it does what I generally speaking, it does what I generally speaking, it does what I expected to do when I tell it what I expected to do when I tell it what I expected to do when I tell it what I want. Astra is why I'm drawing this want. Astra is why I'm drawing this want. Astra is why I'm drawing this chart, though, because Astra can do chart, though, because Astra can do chart, though, because Astra can do things Fable literally could never and things Fable literally could never and things Fable literally could never and blow me away. But then it does the blow me away. But then it does the blow me away. But then it does the stupidest I've ever seen a model do stupidest I've ever seen a model do stupidest I've ever seen a model do and it's so goddamn spiky. This is the and it's so goddamn spiky. This is the and it's so goddamn spiky. This is the thing I really wanted to try and thing I really wanted to try and thing I really wanted to try and communicate. The quality bar with Fable communicate. The quality bar with Fable communicate. The quality bar with Fable is relatively consistent. It goes up and is relatively consistent. It goes up and is relatively consistent. It goes up and down a bit, but never but more than like down a bit, but never but more than like down a bit, but never but more than like two points in either direction. When I two points in either direction. When I two points in either direction. When I send a prompt to Astra, it's equal send a prompt to Astra, it's equal send a prompt to Astra, it's equal chances it drops my jaw because I'm so chances it drops my jaw because I'm so chances it drops my jaw because I'm so blown away with how unbelievable it is. blown away with how unbelievable it is. blown away with how unbelievable it is. or it drops my jaw because I cannot or it drops my jaw because I cannot or it drops my jaw because I cannot fathom that I just spent a thousand fathom that I just spent a thousand fathom that I just spent a thousand dollars for it to run in a loop and not dollars for it to run in a loop and not dollars for it to run in a loop and not ship anything and then break my website.

  56. ship anything and then break my website. ship anything and then break my website. And Ben in chat said that each line also And Ben in chat said that each line also And Ben in chat said that each line also represents your blood pressure while represents your blood pressure while represents your blood pressure while using each model. Yes, I would honestly using each model. Yes, I would honestly using each model. Yes, I would honestly say Astra stresses me out more, a lot say Astra stresses me out more, a lot say Astra stresses me out more, a lot more. This is the best I can explain the more. This is the best I can explain the more. This is the best I can explain the difference here. If you have low difference here. If you have low difference here. If you have low tolerance for if you leave tolerance for if you leave tolerance for if you leave your computer when the model does your computer when the model does your computer when the model does something stupid because it pissed you something stupid because it pissed you something stupid because it pissed you off so much. If you are easily agitated off so much. If you are easily agitated off so much. If you are easily agitated to the point where it affects your work, to the point where it affects your work, to the point where it affects your work, you should pay the extra money for you should pay the extra money for you should pay the extra money for Fable. You just should. It's a way more Fable. You just should. It's a way more Fable. You just should. It's a way more consistent, reliable model. Astra, consistent, reliable model. Astra, consistent, reliable model. Astra, however, is way cheaper. It can do however, is way cheaper. It can do however, is way cheaper. It can do things no other model can, and it things no other model can, and it things no other model can, and it impresses the absolute out of me impresses the absolute out of me impresses the absolute out of me when it does hit those peaks. when it does hit those peaks. when it does hit those peaks. Personally, I like having both, and I Personally, I like having both, and I Personally, I like having both, and I find myself rotating between the two find myself rotating between the two find myself rotating between the two quite a lot. If I had to pick one, I quite a lot. If I had to pick one, I quite a lot. If I had to pick one, I would pick Astra and have it write DMs would pick Astra and have it write DMs would pick Astra and have it write DMs to me to Julia so that he could send off to me to Julia so that he could send off to me to Julia so that he could send off the prompts using Fable for me. And I the prompts using Fable for me. And I the prompts using Fable for me. And I know it puts me in a weird place because know it puts me in a weird place because know it puts me in a weird place because I will always find some way to get good I will always find some way to get good I will always find some way to get good code out of the anthropic models even if code out of the anthropic models even if code out of the anthropic models even if I can't do it directly. Astra has more I can't do it directly. Astra has more I can't do it directly. Astra has more novel uses and especially now that I'm novel uses and especially now that I'm novel uses and especially now that I'm down a hand, Astra's ability to like use down a hand, Astra's ability to like use down a hand, Astra's ability to like use my computer and get work done and my computer and get work done and my computer and get work done and benefit my life outside of code. I benefit my life outside of code. I benefit my life outside of code. I almost feel like as a heavy user of almost feel like as a heavy user of almost feel like as a heavy user of computers, the $200 C codeex plan almost computers, the $200 C codeex plan almost computers, the $200 C codeex plan almost feels essential because it if you use feels essential because it if you use feels essential because it if you use computers professionally and you make computers professionally and you make computers professionally and you make over $100,000 a year, you or your boss over $100,000 a year, you or your boss over $100,000 a year, you or your boss should be paying for the $200 a month should be paying for the $200 a month should be paying for the $200 a month plan just because it makes your ability plan just because it makes your ability plan just because it makes your ability to use your computer more effective. If to use your computer more effective. If to use your computer more effective. If you don't have that and you're saving up you don't have that and you're saving up you don't have that and you're saving up in order to pick the thing that writes

  57. in order to pick the thing that writes in order to pick the thing that writes the best code and you're really easily the best code and you're really easily the best code and you're really easily stressed, I would still pick Astra, but stressed, I would still pick Astra, but stressed, I would still pick Astra, but I would use Astra to make Astra better. I would use Astra to make Astra better. I would use Astra to make Astra better. Find everything you can to smooth out Find everything you can to smooth out Find everything you can to smooth out the rough edges. put together the best the rough edges. put together the best the rough edges. put together the best set of skills in the world and make set of skills in the world and make set of skills in the world and make benchmarks that prove why they're the benchmarks that prove why they're the benchmarks that prove why they're the best. Blast it on Twitter, get a job at best. Blast it on Twitter, get a job at best. Blast it on Twitter, get a job at OpenAI, and never have to worry about OpenAI, and never have to worry about OpenAI, and never have to worry about token spend again. What I'm trying to token spend again. What I'm trying to token spend again. What I'm trying to say is I don't think that person say is I don't think that person say is I don't think that person actually exists. A lot of you act like actually exists. A lot of you act like actually exists. A lot of you act like that person does. They don't. If you're that person does. They don't. If you're that person does. They don't. If you're concerned about costs and that's your concerned about costs and that's your concerned about costs and that's your main motivation, go get the open source main motivation, go get the open source main motivation, go get the open source plan. See if you can convince either lab plan. See if you can convince either lab plan. See if you can convince either lab to give it to you or just wait for these to give it to you or just wait for these to give it to you or just wait for these things to be accessible in cheaper things to be accessible in cheaper things to be accessible in cheaper formats. They will be. GLM53 Flash is a formats. They will be. GLM53 Flash is a formats. They will be. GLM53 Flash is a more pleasant model to use arguably than more pleasant model to use arguably than more pleasant model to use arguably than Astra. I've seen GLM53 Flash go off the Astra. I've seen GLM53 Flash go off the Astra. I've seen GLM53 Flash go off the rails less than Astra for similar work. rails less than Astra for similar work. rails less than Astra for similar work. So, I would expect this level of So, I would expect this level of So, I would expect this level of capability to be cheaper and more capability to be cheaper and more capability to be cheaper and more accessible in things in the near future. accessible in things in the near future. accessible in things in the near future. So, just be patient if you're feeling So, just be patient if you're feeling So, just be patient if you're feeling like the price is too high. I get it. like the price is too high. I get it. like the price is too high. I get it. That said, Fable is what I reach for That said, Fable is what I reach for That said, Fable is what I reach for when I'm trying to land code and Astra when I'm trying to land code and Astra when I'm trying to land code and Astra is what I reach for when I'm trying to is what I reach for when I'm trying to is what I reach for when I'm trying to use my computer. The gap for code is not use my computer. The gap for code is not use my computer. The gap for code is not as huge as it might sound. The gap is as huge as it might sound. The gap is as huge as it might sound. The gap is more in this chart here, but the harsh more in this chart here, but the harsh more in this chart here, but the harsh reality is just this gap in capability.

  58. reality is just this gap in capability. reality is just this gap in capability. This is the post I made when the models This is the post I made when the models This is the post I made when the models both came out to try and explain why I both came out to try and explain why I both came out to try and explain why I like both. Astra is world class at a like both. Astra is world class at a like both. Astra is world class at a shitload of things. It is genuinely the shitload of things. It is genuinely the shitload of things. It is genuinely the best model in the world at all of this best model in the world at all of this best model in the world at all of this stuff, and half of it it is far ahead stuff, and half of it it is far ahead stuff, and half of it it is far ahead for, but Fable writes code that is for, but Fable writes code that is for, but Fable writes code that is mergeable 20ish% more often, and its mergeable 20ish% more often, and its mergeable 20ish% more often, and its stupid spikes are much less stupid. Its stupid spikes are much less stupid. Its stupid spikes are much less stupid. Its smart spikes are not quite as smart, but smart spikes are not quite as smart, but smart spikes are not quite as smart, but its stupid spikes are way less stupid. its stupid spikes are way less stupid. its stupid spikes are way less stupid. And for that reason alone, I think Fable And for that reason alone, I think Fable And for that reason alone, I think Fable 5.1 is a better choice for day-to-day 5.1 is a better choice for day-to-day 5.1 is a better choice for day-to-day code stuff when cost is not a factor. code stuff when cost is not a factor. code stuff when cost is not a factor. But if you're willing to put the effort But if you're willing to put the effort But if you're willing to put the effort in with Astra, you will get a lot for in with Astra, you will get a lot for in with Astra, you will get a lot for it. Both models are awesome. I will be it. Both models are awesome. I will be it. Both models are awesome. I will be defaulting to Fable for code and I'll be defaulting to Fable for code and I'll be defaulting to Fable for code and I'll be defaulting to Astra for every other defaulting to Astra for every other defaulting to Astra for every other thing that I do. Actually, one last fun thing that I do. Actually, one last fun thing that I do. Actually, one last fun side thing that I should have mentioned side thing that I should have mentioned side thing that I should have mentioned before. This will be a really silly one before. This will be a really silly one before. This will be a really silly one to end on. Remember everybody being to end on. Remember everybody being to end on. Remember everybody being really upset about Fable's security really upset about Fable's security really upset about Fable's security refusals? Astra refuses more often. Just refusals? Astra refuses more often. Just refusals? Astra refuses more often. Just wanted to put that one out there. wanted to put that one out there. wanted to put that one out there. Anyways, both these models are great. We Anyways, both these models are great. We Anyways, both these models are great. We are in an unbelievable time to be are in an unbelievable time to be are in an unbelievable time to be engineers. The fact that we have engineers. The fact that we have engineers. The fact that we have technology like this coming out every technology like this coming out every technology like this coming out every few months that massively changes how we few months that massively changes how we few months that massively changes how we work is so godamn cool. It is hard to work is so godamn cool. It is hard to work is so godamn cool. It is hard to pick wrong here, but if you don't see pick wrong here, but if you don't see pick wrong here, but if you don't see the benefit of all of Astra's incredible the benefit of all of Astra's incredible the benefit of all of Astra's incredible capabilities, I am confused because they capabilities, I am confused because they capabilities, I am confused because they are useful to everyone. And if you don't are useful to everyone. And if you don't are useful to everyone. And if you don't think payable is worth the extra money think payable is worth the extra money think payable is worth the extra money for the quality code difference where for the quality code difference where for the quality code difference where it's not that big a difference, I it's not that big a difference, I it's not that big a difference, I totally understand, but you're not totally understand, but you're not totally understand, but you're not shipping enough if you don't see why shipping enough if you don't see why shipping enough if you don't see why that matters. Either model is great.

  59. that matters. Either model is great. that matters. Either model is great. They're both unbelievable. They both They're both unbelievable. They both They're both unbelievable. They both have changed how I work. And when you have changed how I work. And when you have changed how I work. And when you look at the numbers for how we're using look at the numbers for how we're using look at the numbers for how we're using them, both models have massively them, both models have massively them, both models have massively improved our productivity within T3 improved our productivity within T3 improved our productivity within T3 Code. And they don't replace each other, Code. And they don't replace each other, Code. And they don't replace each other, they compound each other. These models they compound each other. These models they compound each other. These models are great and you shouldn't feel bad are great and you shouldn't feel bad are great and you shouldn't feel bad using one or the other. You should using one or the other. You should using one or the other. You should experiment with both because I think experiment with both because I think experiment with both because I think you'll be blown away with what's you'll be blown away with what's you'll be blown away with what's possible. And pretty much everyone who possible. And pretty much everyone who possible. And pretty much everyone who can have the codec sub probably should can have the codec sub probably should can have the codec sub probably should because it's such an insane deal. Have because it's such an insane deal. Have because it's such an insane deal. Have fun burning tokens and using all these fun burning tokens and using all these fun burning tokens and using all these models. I think they're awesome and I models. I think they're awesome and I models. I think they're awesome and I bet you will too. Until next time, peace bet you will too. Until next time, peace bet you will too. Until next time, peace nerds.

Summary

This analysis compares the AI models Fable 5.1 and GBD6 Astra, highlighting their capabilities and differences through extensive token usage. The practical takeaway is understanding which model best suits your needs based on their strengths, weaknesses, efficiency, and subscription costs.

View original episode ↗