Scott & Mark Learn To...Build from Vibes to Craft

You know, I got high school graduation You know, I got high school graduation tomorrow and tomorrow and tomorrow and party the party at the house. We had to party the party at the house. We had to party the party at the house. We had to go to Costco today and get all the go to Costco today and get all the go to Costco today and get all the stuff. So, stuff. So, stuff. So, you can't send your claw to Costco right you can't send your claw to Costco right you can't send your claw to Costco right now. >> [laughter] >> [laughter] >> You just have a very You have a very >> You just have a very You have a very >> You just have a very You have a very satisfied look on your face. You have satisfied look on your face. You have satisfied look on your face. You have >> I could hire somebody from >> I could hire somebody from >> I could hire somebody from >> You're not going to hire somebody to go >> You're not going to hire somebody to go >> You're not going to hire somebody to go to Costco. to Costco. to Costco. >> yeah. >> yeah. >> yeah. >> No, come on. >> No, come on. >> No, come on. >> It's important to >> It's important to >> It's important to do your own shopping at Costco. do your own shopping at Costco. do your own shopping at Costco. Do you have a person for that? Do you have a person for that? Do you have a person for that? >> Uh well, >> Uh well, >> Uh well, Andrea, my wife will go or will Andrea, my wife will go or will Andrea, my wife will go or will >> go to Costco? >> go to Costco? >> go to Costco? >> Huh? >> Huh? >> Huh? >> Do you know what Costco is? It's like >> Do you know what Costco is? It's like >> Do you know what Costco is? It's like this big this big this big >> We are in >> We are in >> We are in Costco's We're in Costco's hometown. Costco's We're in Costco's hometown. Costco's We're in Costco's hometown. >> It's like a data center with food, Mark. >> It's like a data center with food, Mark. >> It's like a data center with food, Mark. >> Yeah. >> Yeah. >> Yeah. >> Yeah. [laughter] >> Yeah. [laughter] >> Yeah. [laughter] So, you would feel comfortable. So, you would feel comfortable. So, you would feel comfortable. There's like loud AC and then there's There's like loud AC and then there's There's like loud AC and then there's racks of stuff, except it's snacks, racks of stuff, except it's snacks, racks of stuff, except it's snacks, Mark. Mark. Mark. >> Cheap hot dogs. >> Cheap hot dogs. >> Cheap hot dogs. >> Cheap hot dogs. >> Cheap hot dogs. >> Cheap hot dogs. >> The hot dogs, yeah. That's a good reason >> The hot dogs, yeah. That's a good reason >> The hot dogs, yeah. That's a good reason to go to Costco. to go to Costco. to go to Costco. >> [music] >> [laughter] >> [laughter] >> How many sessions did you do?

>> How many sessions did you do? >> How many sessions did you do? >> I did the keynote. I worked on two >> I did the keynote. I worked on two >> I did the keynote. I worked on two segments within the keynote. I did the segments within the keynote. I did the segments within the keynote. I did the our our our you know, the the two hour thing which you know, the the two hour thing which you know, the the two hour thing which required planning, planning meetings required planning, planning meetings required planning, planning meetings that you didn't show up to. that you didn't show up to. that you didn't show up to. >> Yep. >> Yep. >> Yep. >> Um then I did uh the session with the >> Um then I did uh the session with the >> Um then I did uh the session with the with you, uh which required planning with you, uh which required planning with you, uh which required planning meetings that uh meetings that uh meetings that uh >> you >> you >> you >> told us to go to. >> told us to go to. >> told us to go to. >> Mhm. You didn't show up to. >> Mhm. You didn't show up to. >> Mhm. You didn't show up to. >> You said, "Make go and plan that." >> You said, "Make go and plan that." >> You said, "Make go and plan that." Um and then I did like a ton of press, Um and then I did like a ton of press, Um and then I did like a ton of press, which was This is the first build that I which was This is the first build that I which was This is the first build that I was like, they they let me talk to was like, they they let me talk to was like, they they let me talk to actual newspapers and magazines and actual newspapers and magazines and actual newspapers and magazines and stuff. stuff. stuff. >> I did two. I did the three I I our >> I did two. I did the three I I our >> I did two. I did the three I I our thing, our two things, and two two other thing, our two things, and two two other thing, our two things, and two two other sessions. sessions. sessions. >> Okay. Well, >> Okay. Well, >> Okay. Well, so that's a little bit of work, I guess. so that's a little bit of work, I guess. so that's a little bit of work, I guess. >> Yeah. >> Yeah. >> Yeah. >> Yeah. >> Yeah. >> Yeah. >> [laughter] >> [laughter] >> [laughter] >> More work than you. >> More work than you. >> More work than you. >> I don't know. I It was definitely It was >> I don't know. I It was definitely It was >> I don't know. I It was definitely It was hard. I was for the first like I had a hard. I was for the first like I had a hard. I was for the first like I had a person They assigned me a person. person They assigned me a person. person They assigned me a person. A lovely person named Heather who helped A lovely person named Heather who helped A lovely person named Heather who helped me like manage my calendar and Heather me like manage my calendar and Heather me like manage my calendar and Heather just keeps me like just keeps me like just keeps me like >> on track. >> on track. >> on track. >> conscious. Keeps me on track and like >> conscious. Keeps me on track and like >> conscious. Keeps me on track and like says go into that room and do this thing says go into that room and do this thing says go into that room and do this thing and da da da da da. So, that was kind of and da da da da da. So, that was kind of and da da da da da. So, that was kind of cool.

cool. cool. But, I just don't like I'm tired. How But, I just don't like I'm tired. How But, I just don't like I'm tired. How did your other talk go? I I was I was did your other talk go? I I was I was did your other talk go? I I was I was busy busy busy working on working on working on >> talks. >> talks. >> talks. >> our talk. >> our talk. >> our talk. >> talks. >> talks. >> talks. >> three talks. You did three talks? >> three talks. You did three talks? >> three talks. You did three talks? >> Including yours, yeah. >> Including yours, yeah. >> Including yours, yeah. >> You did two sessions. >> You did two sessions. >> You did two sessions. Our session and you did your innovation Our session and you did your innovation Our session and you did your innovation session. session. session. >> No, I did two other sessions. >> No, I did two other sessions. >> No, I did two other sessions. >> Seriously? Oh my god, what was the other >> Seriously? Oh my god, what was the other >> Seriously? Oh my god, what was the other one? one? one? What was the third one? What was the third one? What was the third one? >> I was making stuff up? >> I was making stuff up? >> I was making stuff up? >> I thought you were making stuff. I >> I thought you were making stuff. I >> I thought you were making stuff. I thought you hallucinated that. thought you hallucinated that. thought you hallucinated that. >> No, I did one which was a chat between >> No, I did one which was a chat between >> No, I did one which was a chat between me and Ion Stoica. me and Ion Stoica. me and Ion Stoica. >> The Yeah, was he Was he a good guy? >> The Yeah, was he Was he a good guy? >> The Yeah, was he Was he a good guy? >> Yeah, he's great. I mean, I'd known him >> Yeah, he's great. I mean, I'd known him >> Yeah, he's great. I mean, I'd known him from before. We from before. We from before. We met a few times and I you know, I'd met a few times and I you know, I'd met a few times and I you know, I'd visit Berkeley and visit Berkeley and visit Berkeley and >> Okay. Explain to the people why he's >> Okay. Explain to the people why he's >> Okay. Explain to the people why he's significant person. significant person. significant person. >> So, if you've heard of Mesos, if you've >> So, if you've heard of Mesos, if you've >> So, if you've heard of Mesos, if you've heard of Ray, if you've heard of VLLM, heard of Ray, if you've heard of VLLM, heard of Ray, if you've heard of VLLM, if you've heard of if you've heard of if you've heard of Spark, and if you if you've heard of Spark, and if you if you've heard of Spark, and if you if you've heard of Databricks, he's behind all of those Databricks, he's behind all of those Databricks, he's behind all of those things. things. things. >> Mhm. >> Mhm. >> Mhm. >> He runs He's a professor in Berkeley. >> He runs He's a professor in Berkeley. >> He runs He's a professor in Berkeley. Has been doing systems research forever. Has been doing systems research forever. Has been doing systems research forever. >> Mhm. >> Mhm. >> Mhm. >> And all of those projects came out of >> And all of those projects came out of >> And all of those projects came out of his lab and he's was the co-founder and his lab and he's was the co-founder and his lab and he's was the co-founder and original CEO of Databricks. Wow.

original CEO of Databricks. Wow. original CEO of Databricks. Wow. >> And um >> And um >> And um gosh, he's been doing this basically gosh, he's been doing this basically gosh, he's been doing this basically forever. I mean, he did like OCR back in forever. I mean, he did like OCR back in forever. I mean, he did like OCR back in the day. the day. the day. So, like when before So, like when before So, like when before before AI had branding. before AI had branding. before AI had branding. That's why they call him the godfather That's why they call him the godfather That's why they call him the godfather of of AI, right? of of AI, right? of of AI, right? >> Yeah. He I mean, he's his his impact has >> Yeah. He I mean, he's his his impact has >> Yeah. He I mean, he's his his impact has been just pretty monumental. been just pretty monumental. been just pretty monumental. >> Yeah. I mean, he was doing like >> Yeah. I mean, he was doing like >> Yeah. I mean, he was doing like transformers and neural networks in the transformers and neural networks in the transformers and neural networks in the '80s. '80s. '80s. >> Oh, and the LLM LLM LLM leaderboard, >> Oh, and the LLM LLM LLM leaderboard, >> Oh, and the LLM LLM LLM leaderboard, that's that's his, too. that's that's his, too. that's that's his, too. >> [snorts] >> [snorts] >> [snorts] >> So, did he see all this coming? >> So, did he see all this coming? >> So, did he see all this coming? Like, did he see the hockey stick as it Like, did he see the hockey stick as it Like, did he see the hockey stick as it went went went >> Uh I don't know. We didn't talk about >> Uh I don't know. We didn't talk about >> Uh I don't know. We didn't talk about that. We more we talked about that. We more we talked about that. We more we talked about how AI impacts systems architecture. how AI impacts systems architecture. how AI impacts systems architecture. >> Okay. >> Okay. >> Okay. >> We did get into the end. I had to ask >> We did get into the end. I had to ask >> We did get into the end. I had to ask him since I had him there him since I had him there him since I had him there what about the future of software what about the future of software what about the future of software engineering engineering engineering >> Mhm. >> Mhm. >> Mhm. >> and if we'd get to the point where which >> and if we'd get to the point where which >> and if we'd get to the point where which as you and I have discussed and we as you and I have discussed and we as you and I have discussed and we talked in our session talked in our session talked in our session I'll go tell, you know, this talk this I'll go tell, you know, this talk this I'll go tell, you know, this talk this topic will come up when I'm talking to topic will come up when I'm talking to topic will come up when I'm talking to somebody and they'll say, "Hey, um you somebody and they'll say, "Hey, um you somebody and they'll say, "Hey, um you know, at some point you're not going to know, at some point you're not going to know, at some point you're not going to need programmers need programmers need programmers because because because AI will just be able to do it all." And AI will just be able to do it all." And AI will just be able to do it all." And I and I just do not believe that to be I and I just do not believe that to be I and I just do not believe that to be true. I don't believe that true. I don't believe that true. I don't believe that transformers will get us there. I don't transformers will get us there. I don't transformers will get us there. I don't believe that agentic frameworks will get believe that agentic frameworks will get believe that agentic frameworks will get will get us there. And we've talked will get us there. And we've talked will get us there. And we've talked about this before on this podcast and about this before on this podcast and about this before on this podcast and other sessions about the reasons why, other sessions about the reasons why, other sessions about the reasons why, but I thought I'd great opportunity to but I thought I'd great opportunity to but I thought I'd great opportunity to ask Ian.

ask Ian. ask Ian. >> Mhm. >> Mhm. >> Mhm. >> And he immediately was like, "No, it's >> And he immediately was like, "No, it's >> And he immediately was like, "No, it's not going to happen." not going to happen." not going to happen." >> No. >> No. >> No. >> will need people that know what they're >> will need people that know what they're >> will need people that know what they're doing. doing. doing. And the the reason was exactly what And the the reason was exactly what And the the reason was exactly what we've talked about, which is we've talked about, which is we've talked about, which is you can't specify everything. And you can't specify everything. And you can't specify everything. And wherever there's a gap in the wherever there's a gap in the wherever there's a gap in the specification specification specification AI will just go make stuff up. AI will just go make stuff up. AI will just go make stuff up. >> 100%. >> 100%. >> 100%. >> stuff >> stuff >> stuff could violate your implicit assumptions could violate your implicit assumptions could violate your implicit assumptions might have uh gaps in might have uh gaps in might have uh gaps in their completeness their completeness their completeness and not be and not be and not be what you would hope and expect. And this what you would hope and expect. And this what you would hope and expect. And this is why we talk about sculpting, and I is why we talk about sculpting, and I is why we talk about sculpting, and I think software engineering, you know, is think software engineering, you know, is think software engineering, you know, is sculpting. sculpting. sculpting. Uh you need feedback. You need to use Uh you need feedback. You need to use Uh you need feedback. You need to use it. You need to say, "That button's in it. You need to say, "That button's in it. You need to say, "That button's in the wrong place." And the wrong place." And the wrong place." And >> Right. >> Right. >> Right. >> And it I need to get this information. >> And it I need to get this information. >> And it I need to get this information. And oh, it'd be nice if I could save And oh, it'd be nice if I could save And oh, it'd be nice if I could save these five steps with uh different these five steps with uh different these five steps with uh different gesture. gesture. gesture. All of that just comes through All of that just comes through All of that just comes through using it. And if you're not using it as using it. And if you're not using it as using it. And if you're not using it as you go, then you're just going to have a you go, then you're just going to have a you go, then you're just going to have a monster cleanup job at the end. And monster cleanup job at the end. And monster cleanup job at the end. And maybe the architecture that grew up maybe the architecture that grew up maybe the architecture that grew up in that one shot in that one shot in that one shot isn't an architecture that can support isn't an architecture that can support isn't an architecture that can support all the tweaks you want to make. So, all the tweaks you want to make. So, all the tweaks you want to make. So, uh that's [snorts] my uh that's [snorts] my uh that's [snorts] my view on it. And I it's very contentious, view on it. And I it's very contentious, view on it. And I it's very contentious, obviously.

obviously. obviously. There's very smart There's very smart There's very smart >> You said contentious? >> You said contentious? >> You said contentious? >> Well, there it is. >> Well, there it is. >> Well, there it is. There's so many really smart people in There's so many really smart people in There's so many really smart people in the even in the industry. the even in the industry. the even in the industry. People that you and I look up to as People that you and I look up to as People that you and I look up to as mentors and heroes of the industry mentors and heroes of the industry mentors and heroes of the industry that are open to the hey, maybe there that are open to the hey, maybe there that are open to the hey, maybe there won't be any software engineers and need won't be any software engineers and need won't be any software engineers and need for software engineers. for software engineers. for software engineers. >> The only way that there would be no >> The only way that there would be no >> The only way that there would be no software engineers is if the if next software engineers is if the if next software engineers is if the if next token prediction can predict it to the token prediction can predict it to the token prediction can predict it to the point where the ambiguity space can be point where the ambiguity space can be point where the ambiguity space can be filled with reasonable filled with reasonable filled with reasonable specificity. And like the whole specificity. And like the whole specificity. And like the whole >> [gasps] >> [gasps] >> [gasps] >> the whole joke we used to tell when I >> the whole joke we used to tell when I >> the whole joke we used to tell when I was teaching was that your computer your was teaching was that your computer your was teaching was that your computer your program runs exactly as you wrote it. It program runs exactly as you wrote it. It program runs exactly as you wrote it. It is deterministic. And if you leave is deterministic. And if you leave is deterministic. And if you leave anything up for for for interpretation, anything up for for for interpretation, anything up for for for interpretation, that's up for interpretation, not up for that's up for interpretation, not up for that's up for interpretation, not up for statistical mean. Are you getting a Teams call? Are you getting a Teams call? >> Yeah, spam. >> Yeah, spam. >> Yeah, spam. >> Sure. You get spam Teams calls? >> Sure. You get spam Teams calls? >> Sure. You get spam Teams calls? >> Oh, yeah, at work. Work spam.

>> Oh, yeah, at work. Work spam. >> Oh, yeah, at work. Work spam. >> Nobody calls me on Teams spamming. >> Nobody calls me on Teams spamming. >> Nobody calls me on Teams spamming. >> [laughter] >> [laughter] >> [laughter] >> I'll start answering and giving them >> I'll start answering and giving them >> I'll start answering and giving them your number. Hey, you know what? I'm not your number. Hey, you know what? I'm not your number. Hey, you know what? I'm not interested, but you know who is? interested, but you know who is? interested, but you know who is? >> Oh, this guy. >> Oh, this guy. >> Oh, this guy. >> [laughter] >> [laughter] >> [laughter] >> Yeah, um >> Yeah, um >> Yeah, um >> always tell all your friends. >> always tell all your friends. >> always tell all your friends. >> Spam Teams, Scott? No, man. Dude, >> Spam Teams, Scott? No, man. Dude, >> Spam Teams, Scott? No, man. Dude, remember when Skype team Skype spam used remember when Skype team Skype spam used remember when Skype team Skype spam used to be a huge problem? Now I'm getting to be a huge problem? Now I'm getting to be a huge problem? Now I'm getting Telegram spam from random people. Telegram spam from random people. Telegram spam from random people. >> Oh, really? Yeah. >> Oh, really? Yeah. >> Oh, really? Yeah. >> And they just say like, "Oh, are you >> And they just say like, "Oh, are you >> And they just say like, "Oh, are you going to the golf course today?" going to the golf course today?" going to the golf course today?" Like, yeah, they really know me. Like, yeah, they really know me. Like, yeah, they really know me. >> That's fishing spam. >> That's fishing spam. >> That's fishing spam. >> Yeah, exactly. Yeah, "Are you going >> Yeah, exactly. Yeah, "Are you going >> Yeah, exactly. Yeah, "Are you going fishing?" fishing?" fishing?" >> Yeah. >> Yeah. >> Yeah. >> Double on target. No, um I the whole >> Double on target. No, um I the whole >> Double on target. No, um I the whole thing is like thing is like thing is like the member of spec driven development. the member of spec driven development. the member of spec driven development. Well, we just do spec driven Well, we just do spec driven Well, we just do spec driven development. Yeah, let's make we make development. Yeah, let's make we make development. Yeah, let's make we make keywords and we could highlight them keywords and we could highlight them keywords and we could highlight them with the special syntax and maybe we can with the special syntax and maybe we can with the special syntax and maybe we can make a compiler to compile like it's a make a compiler to compile like it's a make a compiler to compile like it's a it's programming. it's programming. it's programming. >> Yeah, that's what you get to. >> Yeah, that's what you get to. >> Yeah, that's what you get to. Programming is the is specking. Deep Programming is the is specking. Deep Programming is the is specking. Deep deep specs, you know. deep specs, you know. deep specs, you know. Yeah, and it's like the you know, go Yeah, and it's like the you know, go Yeah, and it's like the you know, go back to the what's the spec? Well, and back to the what's the spec? Well, and back to the what's the spec? Well, and what's the documentation? Check out the what's the documentation? Check out the what's the documentation? Check out the code. code. code. >> And there it is. And see, and I've >> And there it is. And see, and I've >> And there it is. And see, and I've always said I've got blog posts about always said I've got blog posts about always said I've got blog posts about this from 20 years ago. Word documents this from 20 years ago. Word documents this from 20 years ago. Word documents don't break the build. Yeah. So, I was don't break the build. Yeah. So, I was don't break the build. Yeah. So, I was moving spec into code and spec into you moving spec into code and spec into you moving spec into code and spec into you know, deployment as code or in you know, know, deployment as code or in you know, know, deployment as code or in you know, infrastructure as code and infrastructure as code and infrastructure as code and I've got a whole paper I wrote on this I've got a whole paper I wrote on this I've got a whole paper I wrote on this and uh and uh and uh uh uh uh gosh, 2007?

gosh, 2007? gosh, 2007? >> [snorts] >> [snorts] >> [snorts] >> We were trying to do just desired state >> We were trying to do just desired state >> We were trying to do just desired state configuration and making desired state configuration and making desired state configuration and making desired state engines that would go and run over a a engines that would go and run over a a engines that would go and run over a a description of a desired state. description of a desired state. description of a desired state. >> [snorts] >> [snorts] >> [snorts] >> We even tried experimenting we did a >> We even tried experimenting we did a >> We even tried experimenting we did a wacky thing back in the day. I didn't it wacky thing back in the day. I didn't it wacky thing back in the day. I didn't it didn't work, but it was cool. We checked didn't work, but it was cool. We checked didn't work, but it was cool. We checked the system 32 folder into Git. the system 32 folder into Git. the system 32 folder into Git. It was Mercurial and then later Git in It was Mercurial and then later Git in It was Mercurial and then later Git in order to see if we could do like a track order to see if we could do like a track order to see if we could do like a track changes and then we ended up doing like changes and then we ended up doing like changes and then we ended up doing like a a sha of each folder to make sure that a a sha of each folder to make sure that a a sha of each folder to make sure that like DLLs weren't changing. It was all like DLLs weren't changing. It was all like DLLs weren't changing. It was all just part of just part of just part of how do we deterministically clear out how do we deterministically clear out how do we deterministically clear out that these machines in the web farm are that these machines in the web farm are that these machines in the web farm are all the same all the same all the same and that nothing on the machine uh came and that nothing on the machine uh came and that nothing on the machine uh came from somewhere else. from somewhere else. from somewhere else. I hope my whole life has been driving I hope my whole life has been driving I hope my whole life has been driving towards determinism and now it's just towards determinism and now it's just towards determinism and now it's just fine, man. fine, man. fine, man. It's all just fine. It's all just fine. It's all just fine. >> Well, it the >> Well, it the >> Well, it the frustration I've got with it is frustration I've got with it is frustration I've got with it is uh it's a it's an argument you can't uh it's a it's an argument you can't uh it's a it's an argument you can't definitively win. definitively win. definitively win. It's just opinion. It's just opinion. It's just opinion. >> Yep. >> Yep. >> Yep. >> And in intuition, I mean it's it's like >> And in intuition, I mean it's it's like >> And in intuition, I mean it's it's like I can't from first principle. And also, I can't from first principle. And also, I can't from first principle. And also, I do believe that there is I do believe that there is I do believe that there is you know, software engineering and the you know, software engineering and the you know, software engineering and the apps and services are on a complexity apps and services are on a complexity apps and services are on a complexity spectrum.

spectrum. spectrum. >> Mhm. >> Mhm. >> Mhm. >> And at some end of that spectrum on the >> And at some end of that spectrum on the >> And at some end of that spectrum on the simple side, simple side, simple side, I can see, "Hey, you don't need the I can see, "Hey, you don't need the I can see, "Hey, you don't need the software engineer software engineer software engineer to build that thing." to build that thing." to build that thing." It's It's It's because the constraints it's so simple. because the constraints it's so simple. because the constraints it's so simple. There's not that many constraints. The There's not that many constraints. The There's not that many constraints. The requirements are pretty requirements are pretty requirements are pretty well understood. well understood. well understood. >> Mhm. >> Mhm. >> Mhm. >> The models are really good at knowing >> The models are really good at knowing >> The models are really good at knowing that path. that path. that path. And you can just with a very And you can just with a very And you can just with a very with a few sentences with a few sentences with a few sentences give it a spec that's complete enough give it a spec that's complete enough give it a spec that's complete enough that it will do what you that it will do what you that it will do what you are expecting and hoping. are expecting and hoping. are expecting and hoping. >> Yep. >> Yep. >> Yep. Exactly. Exactly. Exactly. >> You get into the medium complexity and >> You get into the medium complexity and >> You get into the medium complexity and that's where that's where that's where things fall apart, right? things fall apart, right? things fall apart, right? >> Yep. Um I was actually meeting with a >> Yep. Um I was actually meeting with a >> Yep. Um I was actually meeting with a bunch of CTOs and Japanese CTOs uh at bunch of CTOs and Japanese CTOs uh at bunch of CTOs and Japanese CTOs uh at this Japanese event yesterday just right this Japanese event yesterday just right this Japanese event yesterday just right before I I got on the plane. before I I got on the plane. before I I got on the plane. And uh they wanted to know if it would And uh they wanted to know if it would And uh they wanted to know if it would go away and we wouldn't need software go away and we wouldn't need software go away and we wouldn't need software engineers. And you know, I of course engineers. And you know, I of course engineers. And you know, I of course have very publicly said we're going to have very publicly said we're going to have very publicly said we're going to double down on teaching teaching double down on teaching teaching double down on teaching teaching juniors. Like whether it's called juniors. Like whether it's called juniors. Like whether it's called programming or specificationing or programming or specificationing or programming or specificationing or whatever, whatever, whatever, if you are unclear, it will become if you are unclear, it will become if you are unclear, it will become mediocre. So then, oh, we had a meeting mediocre. So then, oh, we had a meeting mediocre. So then, oh, we had a meeting uh Derby and I with uh CMU today.

uh Derby and I with uh CMU today. uh Derby and I with uh CMU today. And uh Michael uh helping us just so And uh Michael uh helping us just so And uh Michael uh helping us just so smart. Such a great guy. I could talk to smart. Such a great guy. I could talk to smart. Such a great guy. I could talk to that guy for hours. that guy for hours. that guy for hours. >> There's a lot of smart people at CMU and >> There's a lot of smart people at CMU and >> There's a lot of smart people at CMU and that have come from CMU. that have come from CMU. that have come from CMU. >> I've heard that. Is that Yeah, you uh >> I've heard that. Is that Yeah, you uh >> I've heard that. Is that Yeah, you uh you almost got in there, right? Is that you almost got in there, right? Is that you almost got in there, right? Is that that university? Yeah. that university? Yeah. that university? Yeah. >> [laughter] >> [laughter] >> [laughter] >> Dr. Mark. Um and uh we were trying to >> Dr. Mark. Um and uh we were trying to >> Dr. Mark. Um and uh we were trying to figure out what like an AI augmented figure out what like an AI augmented figure out what like an AI augmented software uh degree would look like and software uh degree would look like and software uh degree would look like and uh we were talking about doubling down uh we were talking about doubling down uh we were talking about doubling down on QA, which is a lost art, doubling on QA, which is a lost art, doubling on QA, which is a lost art, doubling down on testing, so we wanted more of down on testing, so we wanted more of down on testing, so we wanted more of that. And then he actually had a that. And then he actually had a that. And then he actually had a specification class that talks about, specification class that talks about, specification class that talks about, you know, software specification and the you know, software specification and the you know, software specification and the importance of that, but that was not importance of that, but that was not importance of that, but that was not related to AI. related to AI. related to AI. Uh like clarity of intent is what we Uh like clarity of intent is what we Uh like clarity of intent is what we want to teach people. So we're going to want to teach people. So we're going to want to teach people. So we're going to kind of double down that. kind of double down that. kind of double down that. >> Here's Okay, here's a >> Here's Okay, here's a >> Here's Okay, here's a controversial take. controversial take. controversial take. >> Okay, what you got? >> Okay, what you got? >> Okay, what you got? >> How do you make a junior engineer a >> How do you make a junior engineer a >> How do you make a junior engineer a senior into you give them senior senior into you give them senior senior into you give them senior capability? capability? capability? A skills file. A skills file. A skills file. >> You're going to make a markdown file >> You're going to make a markdown file >> You're going to make a markdown file that's going to allow someone to go to that's going to allow someone to go to that's going to allow someone to go to the next level? the next level? the next level? >> Yeah. Thoughts? >> Yeah. Thoughts? >> Yeah. Thoughts? Cuz I've heard smart people say that.

>> I am starting to get >> I am starting to get a offended and just a little bit a offended and just a little bit a offended and just a little bit irritated irritated irritated that people think that that people think that that people think that telling an LLM you're an expert at blank telling an LLM you're an expert at blank telling an LLM you're an expert at blank does stuff. does stuff. does stuff. Like it nudges word vectors in a certain Like it nudges word vectors in a certain Like it nudges word vectors in a certain direction, but it is not direction, but it is not direction, but it is not a it is not true that you can a it is not true that you can a it is not true that you can you're a you're an amazing senior and you're a you're an amazing senior and you're a you're an amazing senior and you make no mistakes. Like I'm just I'm you make no mistakes. Like I'm just I'm you make no mistakes. Like I'm just I'm so tired of that. And the idea that so tired of that. And the idea that so tired of that. And the idea that someone's going to upload a markdown someone's going to upload a markdown someone's going to upload a markdown file that is somehow unique or genius file that is somehow unique or genius file that is somehow unique or genius and then someone says, "That's the real and then someone says, "That's the real and then someone says, "That's the real unlock." unlock." unlock." That's that's not a thing. That's that's not a thing. That's that's not a thing. >> have you a markdown file of Scott >> have you a markdown file of Scott >> have you a markdown file of Scott Hanselman. Hanselman. Hanselman. >> Yeah, like that's that's a that's >> Yeah, like that's that's a that's >> Yeah, like that's that's a that's I am more than the sum of my context. I am more than the sum of my context. I am more than the sum of my context. And you need need a lot more. And you need need a lot more. And you need need a lot more. >> Yeah. >> Yeah. >> Yeah. >> Like someone actually emailed me today >> Like someone actually emailed me today >> Like someone actually emailed me today after seeing our talk and wanted to know after seeing our talk and wanted to know after seeing our talk and wanted to know if I would take my blog posts, my if I would take my blog posts, my if I would take my blog posts, my 20-plus years of blog posts and turn it 20-plus years of blog posts and turn it 20-plus years of blog posts and turn it into some kind of a skill into some kind of a skill into some kind of a skill and then have the blog start chewing out and then have the blog start chewing out and then have the blog start chewing out blog posts and I just like blog posts and I just like blog posts and I just like I am I haven't answered them. I need to I am I haven't answered them. I need to I am I haven't answered them. I need to figure out figure out figure out if I'm offended or not.

if I'm offended or not. if I'm offended or not. >> [laughter] >> [laughter] >> [laughter] >> It's You sound a little bit offended. >> It's You sound a little bit offended. >> It's You sound a little bit offended. >> You sound offended. It's just >> You sound offended. It's just >> You sound offended. It's just I I think that we I I think that we I I think that we Well, this is the thing, right? We Well, this is the thing, right? We Well, this is the thing, right? We believe believe believe what we believe. what we believe. what we believe. Uh it's based on our intuition, it's Uh it's based on our intuition, it's Uh it's based on our intuition, it's based on a little bit of science, based based on a little bit of science, based based on a little bit of science, based on experience, but like we could be on experience, but like we could be on experience, but like we could be wrong. Someone would watch this video in wrong. Someone would watch this video in wrong. Someone would watch this video in a couple years and say we're wrong. a couple years and say we're wrong. a couple years and say we're wrong. >> I try to check myself on this one. Is it >> I try to check myself on this one. Is it >> I try to check myself on this one. Is it Is it I believe these things cuz I want Is it I believe these things cuz I want Is it I believe these things cuz I want them to be true? them to be true? them to be true? >> Yeah. >> Yeah. >> Yeah. >> And I and I try and I check myself like >> And I and I try and I check myself like >> And I and I try and I check myself like from a neutral perspective. from a neutral perspective. from a neutral perspective. I mean, I don't even I mean, I don't even I mean, I don't even it's not even I think do I want cuz it it's not even I think do I want cuz it it's not even I think do I want cuz it doesn't it's not relevant what I want. doesn't it's not relevant what I want. doesn't it's not relevant what I want. These things will evolve. These things will evolve. These things will evolve. >> Yep. They're going to get better. >> Yep. They're going to get better. >> Yep. They're going to get better. >> You know, efficiency >> You know, efficiency >> You know, efficiency is everybody's goal is everybody's goal is everybody's goal >> Mhm. >> Mhm. >> Mhm. >> of carrying out tasks that these things >> of carrying out tasks that these things >> of carrying out tasks that these things are designed to to carry out and are designed to to carry out and are designed to to carry out and you can't stand in the way of that. you can't stand in the way of that. you can't stand in the way of that. >> Yeah. >> Yeah. >> Yeah. >> You can't stop and say, "No, don't be >> You can't stop and say, "No, don't be >> You can't stop and say, "No, don't be efficient." efficient." efficient." >> Yeah. >> Yeah. >> Yeah. >> Because we want humans to do stuff and >> Because we want humans to do stuff and >> Because we want humans to do stuff and that's just not going to work. that's just not going to work. that's just not going to work. >> Well, when I was talking to the >> Well, when I was talking to the >> Well, when I was talking to the Japanese, I used the the analogy of how Japanese, I used the the analogy of how Japanese, I used the the analogy of how do you have Japanese carpentry and do you have Japanese carpentry and do you have Japanese carpentry and tradition tradition tradition exist in a world where IKEA also exists?

exist in a world where IKEA also exists? exist in a world where IKEA also exists? Like where the goal is furniture. Like where the goal is furniture. Like where the goal is furniture. The goal is the goal is good quality The goal is the goal is good quality The goal is the goal is good quality furniture. furniture. furniture. But the goals are different because IKEA But the goals are different because IKEA But the goals are different because IKEA is good quality furniture in a flat pack is good quality furniture in a flat pack is good quality furniture in a flat pack at scale, at speed to make a profit. at scale, at speed to make a profit. at scale, at speed to make a profit. Well, Japanese carpentry arguably does Well, Japanese carpentry arguably does Well, Japanese carpentry arguably does not have at scale, at a profit like at not have at scale, at a profit like at not have at scale, at a profit like at its heart. its heart. its heart. And that resonated with the Japanese. And that resonated with the Japanese. And that resonated with the Japanese. And I I think that that they are going And I I think that that they are going And I I think that that they are going to continue to grind out really quality to continue to grind out really quality to continue to grind out really quality carpenters. They just won't be able to carpenters. They just won't be able to carpenters. They just won't be able to do as many. They would they can't win. do as many. They would they can't win. do as many. They would they can't win. >> Well, actually, I mean, >> Well, actually, I mean, >> Well, actually, I mean, uh trying to map that analogy into what uh trying to map that analogy into what uh trying to map that analogy into what we've been talking about, the you know, we've been talking about, the you know, we've been talking about, the you know, you talk about that you talk about that you talk about that spectrum of software complexity. spectrum of software complexity. spectrum of software complexity. >> Yeah. >> Yeah. >> Yeah. >> On the >> On the >> On the the low end of that, the low complexity the low end of that, the low complexity the low end of that, the low complexity side of that, that's where you can have side of that, that's where you can have side of that, that's where you can have IKEA. You know, the IKEA. You know, the IKEA. You know, the >> That's great point. >> That's great point. >> That's great point. >> the agent with skills that can crank out >> the agent with skills that can crank out >> the agent with skills that can crank out your web plus database systems. your web plus database systems. your web plus database systems. >> Exactly. If the web plus database >> Exactly. If the web plus database >> Exactly. If the web plus database doesn't matter. Like I always talk about doesn't matter. Like I always talk about doesn't matter. Like I always talk about my podcast admin site my podcast admin site my podcast admin site for an audience of one.

for an audience of one. for an audience of one. If you if I showed it to you, you would If you if I showed it to you, you would If you if I showed it to you, you would be like, "This is Opus 45." Like it it be like, "This is Opus 45." Like it it be like, "This is Opus 45." Like it it it looks, it smells, it works, it runs. it looks, it smells, it works, it runs. it looks, it smells, it works, it runs. It's got a god object. Like it's classic It's got a god object. Like it's classic It's got a god object. Like it's classic uh late 2025 software. But it also works uh late 2025 software. But it also works uh late 2025 software. But it also works great and it's got a thousand tests and great and it's got a thousand tests and great and it's got a thousand tests and it does exactly what I need it to do. it does exactly what I need it to do. it does exactly what I need it to do. >> Yeah. >> Yeah. >> Yeah. >> Would What would I ship it? No. >> Would What would I ship it? No. >> Would What would I ship it? No. >> Um >> Um >> Um Would ship What do you mean? You do ship Would ship What do you mean? You do ship Would ship What do you mean? You do ship it. It is it. It is it. It is >> Let me rephrase. Would I ship it to an >> Let me rephrase. Would I ship it to an >> Let me rephrase. Would I ship it to an audience of two? audience of two? audience of two? Probably. Probably. Probably. >> Yeah, yeah. Okay, but if you take a look >> Yeah, yeah. Okay, but if you take a look >> Yeah, yeah. Okay, but if you take a look at line of business apps at line of business apps at line of business apps >> Mhm. >> Mhm. >> Mhm. >> or bespoke systems >> or bespoke systems >> or bespoke systems >> Sure. >> Sure. >> Sure. >> So, for example, the um uh uh the the the the state retirement systems the state retirement systems the state retirement systems >> Yeah. >> Yeah. >> Yeah. >> for state employees >> for state employees >> for state employees are almost always bespoke systems. are almost always bespoke systems. are almost always bespoke systems. >> Right. >> Right. >> Right. >> And they're implemented typically by >> And they're implemented typically by >> And they're implemented typically by engineers engineers engineers that that that are job engineers. They're not systems are job engineers. They're not systems are job engineers. They're not systems people. people. people. They know basic They know basic They know basic software, you know, programming and software, you know, programming and software, you know, programming and software engineering.

software engineering. software engineering. >> Okay. >> Okay. >> Okay. >> That kind of stuff. >> That kind of stuff. >> That kind of stuff. The need for that is everywhere. The need for that is everywhere. The need for that is everywhere. And it's not like shipped to lots of And it's not like shipped to lots of And it's not like shipped to lots of people. It's people. It's people. It's bespoke because there's bespoke systems bespoke because there's bespoke systems bespoke because there's bespoke systems in these and every state has their own in these and every state has their own in these and every state has their own bespoke data formats and bespoke data formats and bespoke data formats and >> Yeah. >> Yeah. >> Yeah. >> and software that they've been they've >> and software that they've been they've >> and software that they've been they've built their built their built their their stack on top of. And so, you can't their stack on top of. And so, you can't their stack on top of. And so, you can't by definition create something there by definition create something there by definition create something there that is generic. that is generic. that is generic. >> And they have small audiences. >> And they have small audiences. >> And they have small audiences. >> And they have small audiences. And this >> And they have small audiences. And this >> And they have small audiences. And this is where is where is where they're typically on the very low end of they're typically on the very low end of they're typically on the very low end of the the the the complexity spectrum. the complexity spectrum. the complexity spectrum. >> Text box is over data, as I've always >> Text box is over data, as I've always >> Text box is over data, as I've always said. said. said. >> So, >> So, >> So, that I could see that I could see that I could see being the IKEA side of things. being the IKEA side of things. being the IKEA side of things. >> Mhm. >> Mhm. >> Mhm. >> [snorts] >> [snorts] >> [snorts] >> Yeah. And but then but then it becomes a >> Yeah. And but then but then it becomes a >> Yeah. And but then but then it becomes a thing about then it's all about money thing about then it's all about money thing about then it's all about money again. So, then who can afford a again. So, then who can afford a again. So, then who can afford a Japanese carpenter? Who can afford uh Japanese carpenter? Who can afford uh Japanese carpenter? Who can afford uh having a house built from scratch, from having a house built from scratch, from having a house built from scratch, from spec, from, you know, spec, from, you know, spec, from, you know, custom houses versus McMansions.

custom houses versus McMansions. custom houses versus McMansions. >> Okay, so >> Okay, so >> Okay, so let's get even more controversial. let's get even more controversial. let's get even more controversial. What's cheaper? What's cheaper? What's cheaper? A Claude subscription A Claude subscription A Claude subscription with a senior programmer behind it. with a senior programmer behind it. with a senior programmer behind it. >> Right. >> Right. >> Right. >> Or a bunch of >> Or a bunch of >> Or a bunch of >> Yeah, hiring a team of senior >> Yeah, hiring a team of senior >> Yeah, hiring a team of senior programmers. programmers. programmers. >> Yeah. >> Yeah. >> Yeah. >> [snorts] >> [snorts] >> [snorts] [sighs] [sighs] [sighs] >> It's >> It's >> It's complicated, man. It's complicated and complicated, man. It's complicated and complicated, man. It's complicated and it's expensive and it all comes down to it's expensive and it all comes down to it's expensive and it all comes down to driving cost down. I think the question driving cost down. I think the question driving cost down. I think the question is can you drive cost down to is can you drive cost down to is can you drive cost down to to nothing, to zero, to to nothing, to zero, to to nothing, to zero, to >> Okay, so this this is also something >> Okay, so this this is also something >> Okay, so this this is also something that I think we've talked about this that I think we've talked about this that I think we've talked about this before and I've I've mentioned it. before and I've I've mentioned it. before and I've I've mentioned it. >> Yeah. >> Yeah. >> Yeah. >> Like from the very first start when I >> Like from the very first start when I >> Like from the very first start when I was getting this boost, was getting this boost, was getting this boost, I could immediately I immediately saw I could immediately I immediately saw I could immediately I immediately saw the bottleneck is evaluation. In the bottleneck is evaluation. In the bottleneck is evaluation. In evaluation are two things, tests, the evaluation are two things, tests, the evaluation are two things, tests, the actual tests themselves or evaluation, actual tests themselves or evaluation, actual tests themselves or evaluation, and my oversight of the output, whether and my oversight of the output, whether and my oversight of the output, whether it's the test output or the output of it's the test output or the output of it's the test output or the output of the system, the code itself and how it the system, the code itself and how it the system, the code itself and how it behaves, behaves, behaves, is evaluation. is evaluation. is evaluation. >> Right. >> Right. >> Right. >> The bottleneck that's going to that is a >> The bottleneck that's going to that is a >> The bottleneck that's going to that is a bottleneck bottleneck bottleneck >> right? Your responsibility.

>> right? Your responsibility. >> right? Your responsibility. >> that that that will not go to zero. >> that that that will not go to zero. >> that that that will not go to zero. It just AI cannot make it go to zero. It just AI cannot make it go to zero. It just AI cannot make it go to zero. Tests take time to run. Tests take time to run. Tests take time to run. And the more complex your system, the And the more complex your system, the And the more complex your system, the more tests you've got. more tests you've got. more tests you've got. A moderately complex system, A moderately complex system, A moderately complex system, service and Azure has tests that take service and Azure has tests that take service and Azure has tests that take hours to run. hours to run. hours to run. And so it you know, when you're doing a And so it you know, when you're doing a And so it you know, when you're doing a web plus database and you're sitting or web plus database and you're sitting or web plus database and you're sitting or your inner loop is basically the whole your inner loop is basically the whole your inner loop is basically the whole thing. thing. thing. >> Yeah. >> Yeah. >> Yeah. >> And it's a very small set of tests, >> And it's a very small set of tests, >> And it's a very small set of tests, then then then that's perfect for AI and you iterating that's perfect for AI and you iterating that's perfect for AI and you iterating over it and the AI or the AI iterating over it and the AI or the AI iterating over it and the AI or the AI iterating over it and making mistakes and going over it and making mistakes and going over it and making mistakes and going I'm going to fix that and trying again I'm going to fix that and trying again I'm going to fix that and trying again and doing it over. and doing it over. and doing it over. When you get into the point where it's When you get into the point where it's When you get into the point where it's like, well, I'm going to make this like, well, I'm going to make this like, well, I'm going to make this change change change and you know what, when I commit it, I'm and you know what, when I commit it, I'm and you know what, when I commit it, I'm kicking off 5 hours of tests. kicking off 5 hours of tests. kicking off 5 hours of tests. >> Right. >> Right. >> Right. And then you do 5 hours of test in And then you do 5 hours of test in And then you do 5 hours of test in parallel across 10 PRs across across parallel across 10 PRs across across parallel across 10 PRs across across across across across >> The stakes are high >> The stakes are high >> The stakes are high for something failing in the middle of for something failing in the middle of for something failing in the middle of those that 5 hours. those that 5 hours. those that 5 hours. Um so what that means is that you want Um so what that means is that you want Um so what that means is that you want to invest more on on to invest more on on to invest more on on evaluation before you even do the tests evaluation before you even do the tests evaluation before you even do the tests in what the output is.

in what the output is. in what the output is. >> But when you say invest more in >> But when you say invest more in >> But when you say invest more in evaluation, then it's just it's it's evaluation, then it's just it's it's evaluation, then it's just it's it's more it's more AI, it's more models, more it's more AI, it's more models, more it's more AI, it's more models, it's material reviews. it's material reviews. it's material reviews. >> It is it is more of that and that takes >> It is it is more of that and that takes >> It is it is more of that and that takes time, too, time, too, time, too, by the way, and money. by the way, and money. by the way, and money. >> Yep. >> Yep. >> Yep. >> So >> So >> So >> Is it faster than So I think everybody I >> Is it faster than So I think everybody I >> Is it faster than So I think everybody I I I think yeah I I I think yeah I I I think yeah I >> You hear this, too. Cost of software is >> You hear this, too. Cost of software is >> You hear this, too. Cost of software is going to zero. going to zero. going to zero. >> Yep. >> Yep. >> Yep. >> Well, only because the people that are >> Well, only because the people that are >> Well, only because the people that are playing with it and vibe coding crap, playing with it and vibe coding crap, playing with it and vibe coding crap, they only see this tiny slice they only see this tiny slice they only see this tiny slice on a very simple side of the spectrum of on a very simple side of the spectrum of on a very simple side of the spectrum of complex software systems complexity. complex software systems complexity. complex software systems complexity. >> Mhm. >> Mhm. >> Mhm. >> And they go, "Wow, >> And they go, "Wow, >> And they go, "Wow, this thing that in even on the small this thing that in even on the small this thing that in even on the small projects we got, this would have taken projects we got, this would have taken projects we got, this would have taken me months. Now it took me hours. Cost of me months. Now it took me hours. Cost of me months. Now it took me hours. Cost of software is going to zero." software is going to zero." software is going to zero." >> Mhm. >> Mhm. >> Mhm. >> But >> But >> But the things that we're working on the things that we're working on the things that we're working on on the low end of the complexity on the low end of the complexity on the low end of the complexity spectrum, spectrum, spectrum, we're not shipping it to production, we're not shipping it to production, we're not shipping it to production, we're not maintaining it. we're not maintaining it. we're not maintaining it. And And And we don't have and the system is we don't have and the system is we don't have and the system is because it's simple doesn't have tests because it's simple doesn't have tests because it's simple doesn't have tests that run hours. that run hours. that run hours. >> Right. And those And the cost of those >> Right. And those And the cost of those >> Right. And those And the cost of those little utilities, those little command little utilities, those little command little utilities, those little command line utilities, the little vibe coded line utilities, the little vibe coded line utilities, the little vibe coded thing though that is being driven to thing though that is being driven to thing though that is being driven to done near zero.

done near zero. done near zero. >> So that side that side of things, again, >> So that side that side of things, again, >> So that side that side of things, again, it's that Kia side of things. I think it's that Kia side of things. I think it's that Kia side of things. I think we'll just start calling it that, the we'll just start calling it that, the we'll just start calling it that, the Kia side of things. Kia side of things. Kia side of things. >> With all respect to our friends at IKEA. >> With all respect to our friends at IKEA. >> With all respect to our friends at IKEA. >> Yeah. That is the cost of that is going >> Yeah. That is the cost of that is going >> Yeah. That is the cost of that is going to zero. >> So then the question is will Japanese >> So then the question is will Japanese carpentry be driven to zero? carpentry be driven to zero? carpentry be driven to zero? >> No, it's not going to zero. >> No, it's not going to zero. >> No, it's not going to zero. >> Right. That's where we are saying we >> Right. That's where we are saying we >> Right. That's where we are saying we believe it will not go to zero. believe it will not go to zero. believe it will not go to zero. >> Yeah. >> Yeah. >> Yeah. >> And that's we don't know. >> And that's we don't know. >> And that's we don't know. We don't know and we we could be wrong. We don't know and we we could be wrong. We don't know and we we could be wrong. And the question you And the question you And the question you >> I know. >> I know. >> I know. >> Yeah. You know. You're you're you're >> Yeah. You know. You're you're you're >> Yeah. You know. You're you're you're convinced. convinced. convinced. >> I thought you always I thought I thought >> I thought you always I thought I thought >> I thought you always I thought I thought the saying was Mark is always right. the saying was Mark is always right. the saying was Mark is always right. >> Uh the the saying is, as I recall, he >> Uh the the saying is, as I recall, he >> Uh the the saying is, as I recall, he who is most likely to make declarative who is most likely to make declarative who is most likely to make declarative statements is most likely to call be statements is most likely to call be statements is most likely to call be called a fool in retrospect. So I just called a fool in retrospect. So I just called a fool in retrospect. So I just kind of worry. kind of worry. kind of worry. >> declarative statement. >> declarative statement. >> declarative statement. >> Yeah, well, we are making declarative >> Yeah, well, we are making declarative >> Yeah, well, we are making declarative statement. statement. statement. >> didn't you just make one by making that >> didn't you just make one by making that >> didn't you just make one by making that statement? statement? statement? >> There you go. It's turtles all the way >> There you go. It's turtles all the way >> There you go. It's turtles all the way down, right? down, right? down, right? >> Yeah. >> Yeah. >> Yeah. >> Yeah, man, I don't know. >> Yeah, man, I don't know. >> Yeah, man, I don't know. Uh Uh Uh >> No, I >> No, I >> No, I So, look, I think So, look, I think So, look, I think it's one of those things as as uh we it's one of those things as as uh we it's one of those things as as uh we talked about this about this the skill talked about this about this the skill talked about this about this the skill of of of of uh of uh of uh success in a complicated world success in a complicated world success in a complicated world >> Mhm.

>> Mhm. >> Mhm. >> uh which can be an organizational world >> uh which can be an organizational world >> uh which can be an organizational world or technology world or technology world or technology world is always leaving room for being wrong. is always leaving room for being wrong. is always leaving room for being wrong. >> Yep. >> Yep. >> Yep. It is. And It is. And It is. And it depends though is tiresome after so it depends though is tiresome after so it depends though is tiresome after so many years though. Like you and I like many years though. Like you and I like many years though. Like you and I like all things it's like, well, it depends. all things it's like, well, it depends. all things it's like, well, it depends. Does this solution work Does this solution work Does this solution work for some set of specifications? Yes. for some set of specifications? Yes. for some set of specifications? Yes. >> Yeah. Okay. I mean, that just you just >> Yeah. Okay. I mean, that just you just >> Yeah. Okay. I mean, that just you just tre- you know, tre- you know, tre- you know, triggered something in me which triggered something in me which triggered something in me which >> what do you got? >> what do you got? >> what do you got? >> Well, this is the um >> Well, this is the um >> Well, this is the um the of course everything has nuance. Of the of course everything has nuance. Of the of course everything has nuance. Of course it does. And what but but bothers course it does. And what but but bothers course it does. And what but but bothers me is you've got to make you've got to me is you've got to make you've got to me is you've got to make you've got to make a statement about make a statement about make a statement about directionally. And for the most part. directionally. And for the most part. directionally. And for the most part. And what I can't just annoys me is And what I can't just annoys me is And what I can't just annoys me is you're in meetings or you're talking you're in meetings or you're talking you're in meetings or you're talking with other groups and you you make this with other groups and you you make this with other groups and you you make this hey, this is the way things you know, hey, this is the way things you know, hey, this is the way things you know, this is this is this is what's going to work. This is what we what's going to work. This is what we what's going to work. This is what we should do. should do. should do. And then you have the people that are And then you have the people that are And then you have the people that are like, wait, wait, wait. like, wait, wait, wait. like, wait, wait, wait. What about all these What about all these What about all these you know, I don't necessarily agree with you know, I don't necessarily agree with you know, I don't necessarily agree with that. All of these nuances.

that. All of these nuances. that. All of these nuances. >> Mhm. >> Mhm. >> Mhm. >> And I'm like, of course. That's it's >> And I'm like, of course. That's it's >> And I'm like, of course. That's it's obvious those nuances are obvious. obvious those nuances are obvious. obvious those nuances are obvious. You're missing the point of what I'm You're missing the point of what I'm You're missing the point of what I'm trying to say here which is this is the trying to say here which is this is the trying to say here which is this is the direction. direction. direction. And we got to take into account all And we got to take into account all And we got to take into account all those nuances on the way there, but those nuances on the way there, but those nuances on the way there, but unless you can plant a flag or point at unless you can plant a flag or point at unless you can plant a flag or point at something and say, let's go there, something and say, let's go there, something and say, let's go there, you're never going to make any progress. you're never going to make any progress. you're never going to make any progress. And if you sit here all the time with And if you sit here all the time with And if you sit here all the time with your nuances and edge cases and your nuances and edge cases and your nuances and edge cases and conditions and conditions and conditions and like like like that's not helpful, especially in the that's not helpful, especially in the that's not helpful, especially in the early stages of figuring out what you early stages of figuring out what you early stages of figuring out what you want to do. want to do. want to do. >> Well, and I think figuring out what >> Well, and I think figuring out what >> Well, and I think figuring out what you're going to do is also about the you're going to do is also about the you're going to do is also about the definition of done. That was another big definition of done. That was another big definition of done. That was another big thing when I was teaching is that early thing when I was teaching is that early thing when I was teaching is that early in career people who are developing in career people who are developing in career people who are developing their sense of taste don't quite know their sense of taste don't quite know their sense of taste don't quite know yet what done looks like. So, they'll yet what done looks like. So, they'll yet what done looks like. So, they'll declare victory well before, but then declare victory well before, but then declare victory well before, but then people who are a little bit farther people who are a little bit farther people who are a little bit farther along in their career and I think I want along in their career and I think I want along in their career and I think I want to say that you're like this a little to say that you're like this a little to say that you're like this a little bit is that I would have called done a bit is that I would have called done a bit is that I would have called done a week ago and you're like, "No, this is week ago and you're like, "No, this is week ago and you're like, "No, this is not even close to done." Like, "We're at not even close to done." Like, "We're at not even close to done." Like, "We're at 98% We're at 99%." Like, "Yes, but I 98% We're at 99%." Like, "Yes, but I 98% We're at 99%." Like, "Yes, but I found this edge case."

found this edge case." found this edge case." Like with that with your scroller Like with that with your scroller Like with that with your scroller stitcher thing, you know what I mean? stitcher thing, you know what I mean? stitcher thing, you know what I mean? Like I declared done at 80%. Like I declared done at 80%. Like I declared done at 80%. >> Well >> Well >> Well >> Yeah, but actually um >> Yeah, but actually um >> Yeah, but actually um I think that that has served me well. I think that that has served me well. I think that that has served me well. Like System Turtles Like System Turtles Like System Turtles >> served you well. To your credit, they >> served you well. To your credit, they >> served you well. To your credit, they exist in turtles, right? exist in turtles, right? exist in turtles, right? >> Um >> Um >> Um >> I was like I can't it can't be done >> I was like I can't it can't be done >> I was like I can't it can't be done until I believe that it's done. If until I believe that it's done. If until I believe that it's done. If there's If there's weaknesses in it that there's If there's weaknesses in it that there's If there's weaknesses in it that I run into, I've got to I've got to I run into, I've got to I've got to I run into, I've got to I've got to address them because somebody else is address them because somebody else is address them because somebody else is going to run into it and go, "This is going to run into it and go, "This is going to run into it and go, "This is crappy software." crappy software." crappy software." Or they got a bad experience and it Or they got a bad experience and it Or they got a bad experience and it taints the reputation of the whole taints the reputation of the whole taints the reputation of the whole thing. thing. thing. >> Well, so I I was working on the the >> Well, so I I was working on the the >> Well, so I I was working on the the Windows companion app that I showed you Windows companion app that I showed you Windows companion app that I showed you for OpenClaw for OpenClaw for OpenClaw and I'm like and I'm like and I'm like this is alpha, [clears throat] you know this is alpha, [clears throat] you know this is alpha, [clears throat] you know what I mean? Like I feel okay about it. what I mean? Like I feel okay about it. what I mean? Like I feel okay about it. I've been using it I've been I've been I've been using it I've been I've been I've been using it I've been I've been using it for weeks without any problems, using it for weeks without any problems, using it for weeks without any problems, but I know that there's going to be but I know that there's going to be but I know that there's going to be somebody on a weird OS or a weird build somebody on a weird OS or a weird build somebody on a weird OS or a weird build or a weird this or in a different a or a weird this or in a different a or a weird this or in a different a non-English language and I haven't non-English language and I haven't non-English language and I haven't thought about that and then and I'm thought about that and then and I'm thought about that and then and I'm going to get nailed. And it's still going to get nailed. And it's still going to get nailed. And it's still alpha. alpha. alpha. And but you know, but the OpenClaw folks And but you know, but the OpenClaw folks And but you know, but the OpenClaw folks are like, "Nope, they'll be fine." You are like, "Nope, they'll be fine." You are like, "Nope, they'll be fine." You know, and of course know, and of course know, and of course that is the whole move fast and break that is the whole move fast and break that is the whole move fast and break things thing, which is a cultural choice things thing, which is a cultural choice things thing, which is a cultural choice to ship software to ship software to ship software >> And I think I think it's fine to you >> And I think I think it's fine to you >> And I think I think it's fine to you know, when there's expectation around know, when there's expectation around know, when there's expectation around it. Well, this is an open source it. Well, this is an open source it. Well, this is an open source project. It's you know, single guy project. It's you know, single guy project. It's you know, single guy behind it.

behind it. behind it. You know, I expect it to have rough You know, I expect it to have rough You know, I expect it to have rough edges. We're moving fast in a edges. We're moving fast in a edges. We're moving fast in a fast-moving world. fast-moving world. fast-moving world. If it's coming from Microsoft and it's If it's coming from Microsoft and it's If it's coming from Microsoft and it's going to lots of people going to lots of people going to lots of people >> That's what I said. >> That's what I said. >> That's what I said. >> You the bar is different. >> You the bar is different. >> You the bar is different. >> Yeah, because I because we we we we went >> Yeah, because I because we we we we went >> Yeah, because I because we we we we went to leadership and we got a couple people to leadership and we got a couple people to leadership and we got a couple people to help out and the team's doing a great to help out and the team's doing a great to help out and the team's doing a great job, but like even a you know, a tiny job, but like even a you know, a tiny job, but like even a you know, a tiny [snorts] team in a couple of weeks half [snorts] team in a couple of weeks half [snorts] team in a couple of weeks half time, you know what I mean? Like they time, you know what I mean? Like they time, you know what I mean? Like they have other jobs. They have other stuff have other jobs. They have other stuff have other jobs. They have other stuff to do. to do. to do. You know, what's the what's the old You know, what's the what's the old You know, what's the what's the old joke? Nine women can't have a baby in a joke? Nine women can't have a baby in a joke? Nine women can't have a baby in a month. Like there is a certain amount of month. Like there is a certain amount of month. Like there is a certain amount of baking that just needs to happen. baking that just needs to happen. baking that just needs to happen. And uh And uh And uh we're continuing to bake and I'm proud we're continuing to bake and I'm proud we're continuing to bake and I'm proud of the software, but I have no doubt of the software, but I have no doubt of the software, but I have no doubt that there's dragons deep inside it that there's dragons deep inside it that there's dragons deep inside it somewhere. somewhere. somewhere. >> So Simon Willison who we got to >> So Simon Willison who we got to >> So Simon Willison who we got to hang out with hang out with hang out with >> Yeah, yeah. >> Yeah, yeah. >> Yeah, yeah. >> at Build. >> at Build. >> at Build. Uh Uh Uh One of the blog posts that we point at One of the blog posts that we point at One of the blog posts that we point at is is is and what he what he said at Build in our and what he what he said at Build in our and what he what he said at Build in our discussions with him discussions with him discussions with him and also in his talk is the only and also in his talk is the only and also in his talk is the only software he'll trust is software that software he'll trust is software that software he'll trust is software that has been used by lots of people for some has been used by lots of people for some has been used by lots of people for some period of time.

period of time. period of time. >> That's a great point. And he has whole >> That's a great point. And he has whole >> That's a great point. And he has whole blog post talking about how you are blog post talking about how you are blog post talking about how you are ultimately responsible for shipping, ultimately responsible for shipping, ultimately responsible for shipping, which is on you and it doesn't matter which is on you and it doesn't matter which is on you and it doesn't matter where the code came from, whether it was where the code came from, whether it was where the code came from, whether it was from open source. And I'm starting to from open source. And I'm starting to from open source. And I'm starting to use the analogy of if the LLM makes use the analogy of if the LLM makes use the analogy of if the LLM makes makes a bunch of code, you should treat makes a bunch of code, you should treat makes a bunch of code, you should treat it with the same kind of distrust that it with the same kind of distrust that it with the same kind of distrust that you take a random open source person's you take a random open source person's you take a random open source person's pull request. pull request. pull request. >> Yep. >> Yep. >> Yep. >> You just you just don't know. Yeah. Yeah. >> Well, okay. >> Well, okay. >> Well, okay. Well, we're I don't know what our Well, we're I don't know what our Well, we're I don't know what our >> That's a show right here. I don't know >> That's a show right here. I don't know >> That's a show right here. I don't know what the actual like what the actual like what the actual like >> I had more to talk >> I had more to talk >> I had more to talk >> Oh, please continue talking. We'll I'll >> Oh, please continue talking. We'll I'll >> Oh, please continue talking. We'll I'll I'll step out. I'll step out. I'll step out. >> [laughter] >> Well, you're you're you're >> Well, you're you're you're you're good. Thanks for your you're good. Thanks for your you're good. Thanks for your contribution, Scott. contribution, Scott. contribution, Scott. >> No, no, no. Yeah, you're absolutely >> No, no, no. Yeah, you're absolutely >> No, no, no. Yeah, you're absolutely right. You're right to push back. right. You're right to push back. right. You're right to push back. >> [laughter] >> [laughter] >> [laughter] >> No, what what were you going to say? >> No, what what were you going to say? >> No, what what were you going to say? >> Um >> Um >> Um I was going to say what Well, you I was going to say what Well, you I was going to say what Well, you interrupted me. >> [laughter] >> [laughter] >> Must be my fault. >> Must be my fault. >> Must be my fault. >> Uh >> Uh >> Uh >> what I'm trying to say?

>> what I'm trying to say? >> what I'm trying to say? >> Simon Willison, you're responsible for >> Simon Willison, you're responsible for >> Simon Willison, you're responsible for >> for what you ship. >> for what you ship. >> for what you ship. >> for what you ship. >> for what you ship. >> for what you ship. And I didn't interrupt you. I yes and I And I didn't interrupt you. I yes and I And I didn't interrupt you. I yes and I supported you supported you supported you with an interjection. with an interjection. with an interjection. >> all right. I can't remember and you >> all right. I can't remember and you >> all right. I can't remember and you know, it's one of those things now that know, it's one of those things now that know, it's one of those things now that I my my brain's like haha. I my my brain's like haha. I my my brain's like haha. I know what you want what you're looking I know what you want what you're looking I know what you want what you're looking for but for but for but >> And now [clears throat] you're you see >> And now [clears throat] you're you see >> And now [clears throat] you're you see your your brain is just trolling you at your your brain is just trolling you at your your brain is just trolling you at this point? this point? this point? >> Okay. Well, >> Okay. Well, >> Okay. Well, maybe when it when you come back after maybe when it when you come back after maybe when it when you come back after that we let the we go to these sponsors that we let the we go to these sponsors that we let the we go to these sponsors or whatever we should get Pike get or whatever we should get Pike get or whatever we should get Pike get sponsors at some point. sponsors at some point. sponsors at some point. >> Yeah. >> Yeah. >> Yeah. >> Uh we'll we'll return back and we'll go >> Uh we'll we'll return back and we'll go >> Uh we'll we'll return back and we'll go the next episode Mark will remember what the next episode Mark will remember what the next episode Mark will remember what he was going to talk about. Cuz that he was going to talk about. Cuz that he was going to talk about. Cuz that could be a whole next episode. could be a whole next episode. could be a whole next episode. >> to remember. >> to remember. >> to remember. >> This one's only Well, yeah, we need a >> This one's only Well, yeah, we need a >> This one's only Well, yeah, we need a memory system. We need to work on that. memory system. We need to work on that. memory system. We need to work on that. Um we'll find the markdown file the Mark Um we'll find the markdown file the Mark Um we'll find the markdown file the Mark Susterovich dot md file that has all the Susterovich dot md file that has all the Susterovich dot md file that has all the things that needs to things that needs to things that needs to >> the the notebook that I talked about. >> the the notebook that I talked about. >> the the notebook that I talked about. >> Yeah? We should talk about that. >> Yeah? We should talk about that. >> Yeah? We should talk about that. >> Previ- No, we already did. >> Previ- No, we already did. >> Previ- No, we already did. >> Did we do that? >> Did we do that? >> Did we do that? >> You do you not remember? >> You do you not remember? >> You do you not remember? >> I do. I don't know what you're >> I do. I don't know what you're >> I do. I don't know what you're >> Yeah, it is. We we covered this at the >> Yeah, it is. We we covered this at the >> Yeah, it is. We we covered this at the beginning of the show. beginning of the show. beginning of the show. >> I don't know what day it is. >> I don't know what day it is. >> I don't know what day it is. >> We talked about keeping task lists.

>> We talked about keeping task lists. >> We talked about keeping task lists. >> Oh, yeah, yeah, yeah. The to-dos. Yeah, >> Oh, yeah, yeah, yeah. The to-dos. Yeah, >> Oh, yeah, yeah, yeah. The to-dos. Yeah, that was good. That was good. that was good. That was good. that was good. That was good. >> Yeah. >> Yeah. >> Yeah. >> Yeah, I've got my Moleskine notebooks or >> Yeah, I've got my Moleskine notebooks or >> Yeah, I've got my Moleskine notebooks or Moleskine Moleskina? Moleskine Moleskina? Moleskine Moleskina? Moleskina? Moleskina? Moleskina? >> I don't know. Moleskin Moles- Moleskin >> I don't know. Moleskin Moles- Moleskin >> I don't know. Moleskin Moles- Moleskin Moles- Moles- Moles- >> You called me out in public because I >> You called me out in public because I >> You called me out in public because I said pleb and it's pleeb? said pleb and it's pleeb? said pleb and it's pleeb? >> Yeah. >> Yeah. >> Yeah. >> Plebian? >> Plebian? >> Plebian? >> Actually look that up. >> Actually look that up. >> Actually look that up. >> Okay, hang on. We got to find that out >> Okay, hang on. We got to find that out >> Okay, hang on. We got to find that out now because now because now because I've been saying pleb I've been saying pleb I've been saying pleb >> you can have AI say it out loud. >> you can have AI say it out loud. >> you can have AI say it out loud. Right? Right? Right? If you go to dictionary.com now they If you go to dictionary.com now they If you go to dictionary.com now they have a have a have a >> Thank you. I am familiar with how you >> Thank you. I am familiar with how you >> Thank you. I am familiar with how you use the internet but I will go ahead and use the internet but I will go ahead and use the internet but I will go ahead and hit dictionary.com. hit dictionary.com. hit dictionary.com. >> Were you familiar with that site? >> Were you familiar with that site? >> Were you familiar with that site? >> Yeah. >> Yeah. >> Yeah. Plebian. Yeah, it is plebian. Plebian. Yeah, it is plebian. Plebian. Yeah, it is plebian. >> Oh, that ple- >> Oh, that ple- >> Oh, that ple- plebian. No. plebian. No. plebian. No. >> No, no, no. Can you hear this? Is it cutting it out? Is it cutting it out? >> Is it pleb or pleeb? >> Is it pleb or pleeb? >> Is it pleb or pleeb? >> It's Well, that's the thing. When you >> It's Well, that's the thing. When you >> It's Well, that's the thing. When you shorten it shorten it shorten it >> It's pleeb. >> It's pleeb. >> It's pleeb. >> p l e b >> p l e b >> p l e b Looks like pleb. Looks like pleb. Looks like pleb. But let's actually look that up in But let's actually look that up in But let's actually look that up in dictionary.com. dictionary.com. dictionary.com. Nope. It says see pub- plebian plebian.

Nope. It says see pub- plebian plebian. Nope. It says see pub- plebian plebian. A member of the Roman plebs. A member of the Roman plebs. A member of the Roman plebs. Aha! Aha! Aha! >> What? >> You hear this? >> You hear this? >> Pleb. >> Pleb. >> Pleb. >> It's plebs when you shorten it plebeian >> It's plebs when you shorten it plebeian >> It's plebs when you shorten it plebeian is plebs the common people of ancient is plebs the common people of ancient is plebs the common people of ancient Rome. Rome. Rome. >> It's plebeian though. >> It's plebeian though. >> It's plebeian though. >> It's it's ple >> It's it's ple >> It's it's ple >> No, it's plebeian. >> No, it's plebeian. >> No, it's plebeian. >> Is it plebeian? >> Is it plebeian? >> Is it plebeian? >> Plebeian, uh-huh. >> Plebeian, uh-huh. >> Plebeian, uh-huh. >> Plebeian, it's not plebeian. >> Plebeian, it's not plebeian. >> Plebeian, it's not plebeian. >> Okay, but it's plebeian. >> Okay, but it's plebeian. >> Okay, but it's plebeian. >> Yes, but I was right is all that >> Yes, but I was right is all that >> Yes, but I was right is all that matters. matters. matters. >> Oh, and it's a pleb. >> Oh, and it's a pleb. >> Oh, and it's a pleb. >> Yeah, funny how that is. >> Yeah, funny how that is. >> Yeah, funny how that is. >> All right, all right, all right, all >> All right, all right, all right, all >> All right, all right, all right, all right, right, right, plebeian pleb. plebeian pleb. plebeian pleb. >> We have to open with this part of the >> We have to open with this part of the >> We have to open with this part of the >> Yeah. >> Yeah. >> Yeah. >> [laughter] >> [laughter] >> [laughter] >> I want your right in Mark's voice to be >> I want your right in Mark's voice to be >> I want your right in Mark's voice to be my my my my ringtone now. So, this is excellent. my ringtone now. So, this is excellent. my ringtone now. So, this is excellent. >> [music] >> [music] >> [music] >> Did I say you're right? >> Did I say you're right? >> Did I say you're right? >> You did. >> You did. >> You did. >> I'm cooking it. >> I'm cooking it. >> I'm cooking it. >> Cooking it. >> Cooking it. >> Cooking it. >> [laughter] >> [laughter] >> [laughter] >> Becoming a Tik Tok. >> Becoming a Tik Tok. >> Becoming a Tik Tok. >> It was magic. It was it was phenomenal. >> It was magic. It was it was phenomenal. >> It was magic. It was it was phenomenal. >> If you made it this far into the show, >> If you made it this far into the show, >> If you made it this far into the show, like and subscribe, add a comment, and like and subscribe, add a comment, and like and subscribe, add a comment, and you are better than your peers. you are better than your peers. you are better than your peers. Is that all right? That's our new cuz Is that all right? That's our new cuz Is that all right? That's our new cuz people people people >> You're in the top 1% of 1%.

>> You're in the top 1% of 1%. >> You're in the top 1% of 1%. >> 1% of human beings on planet Earth if >> 1% of human beings on planet Earth if >> 1% of human beings on planet Earth if you made [music] it to the end. you made [music] it to the end. you made [music] it to the end. >> 1% of 1%. >> 1% of 1%. >> 1% of 1%. >> Oh, at least. >> Oh, at least. >> Oh, at least. Absolutely, you're on the far right of Absolutely, you're on the far right of Absolutely, you're on the far right of the bell curve. the bell curve. the bell curve. >> Yeah, 100% agree. >> Yeah, 100% agree. >> Yeah, 100% agree. >> All right, cool. Like and subscribe, >> All right, cool. Like and subscribe, >> All right, cool. Like and subscribe, please share, and please comment, and we please share, and please comment, and we please share, and please comment, and we will see you again next week.