The Real Story Behind the Government GPT 5.6 Freeze.

OpenAI just released Chad GPT 5.6, but OpenAI just released Chad GPT 5.6, but not in a normal way. For now, access is not in a normal way. For now, access is not in a normal way. For now, access is restricted to a small group of restricted to a small group of restricted to a small group of government approved partners while government approved partners while government approved partners while Washington reviews the cyber security Washington reviews the cyber security Washington reviews the cyber security risk. That's not a cancellation, but risk. That's not a cancellation, but risk. That's not a cancellation, but it's a tremendous slowdown in frontier it's a tremendous slowdown in frontier it's a tremendous slowdown in frontier availability. And by the end of this availability. And by the end of this availability. And by the end of this video, I want you to understand why that video, I want you to understand why that video, I want you to understand why that delay, the new Siri, claude tag, GLM delay, the new Siri, claude tag, GLM delay, the new Siri, claude tag, GLM 5.2, and Codeex are all about the same 5.2, and Codeex are all about the same 5.2, and Codeex are all about the same underlying thing. a battle for the part underlying thing. a battle for the part underlying thing. a battle for the part of your brain that understands work, not of your brain that understands work, not of your brain that understands work, not your brain in the sci-fi mind control your brain in the sci-fi mind control your brain in the sci-fi mind control sense. The everyday part, which message sense. The everyday part, which message sense. The everyday part, which message matters, which file is current, what the matters, which file is current, what the matters, which file is current, what the customer actually meant, what the team customer actually meant, what the team customer actually meant, what the team decided, what can be shared, what can't decided, what can be shared, what can't decided, what can be shared, what can't be shared, what counts as done. Because be shared, what counts as done. Because be shared, what counts as done. Because if Frontier intelligence slows down, if Frontier intelligence slows down, if Frontier intelligence slows down, even for a few weeks, the next advantage even for a few weeks, the next advantage even for a few weeks, the next advantage is not owning the newest model. It's is not owning the newest model. It's is not owning the newest model. It's having the context that makes any good having the context that makes any good having the context that makes any good model useful. Look at the week through model useful. Look at the week through model useful. Look at the week through that lens, and the news starts to rhyme. that lens, and the news starts to rhyme. that lens, and the news starts to rhyme. Apple's trying to fix Siri by giving it Apple's trying to fix Siri by giving it Apple's trying to fix Siri by giving it access to your messages and photos and access to your messages and photos and access to your messages and photos and email and notes and screen and apps. email and notes and screen and apps. email and notes and screen and apps. Anthropic has launched Claude Tag and Anthropic has launched Claude Tag and Anthropic has launched Claude Tag and Slack where a team can give Claude Slack where a team can give Claude Slack where a team can give Claude access to selected channels and tools access to selected channels and tools access to selected channels and tools and data and code bases. Z.AI's GLM 5.2 and data and code bases. Z.AI's GLM 5.2 and data and code bases. Z.AI's GLM 5.2 has made cheap open frontier-ish has made cheap open frontier-ish has made cheap open frontier-ish intelligence feel much closer to a intelligence feel much closer to a intelligence feel much closer to a reality than it did just a few weeks reality than it did just a few weeks reality than it did just a few weeks ago. And OpenAI has a codeex paper ago. And OpenAI has a codeex paper ago. And OpenAI has a codeex paper showing that inside OpenAI, Codex has showing that inside OpenAI, Codex has showing that inside OpenAI, Codex has become the dominant surface for become the dominant surface for become the dominant surface for workrelated AI output. And those really workrelated AI output. And those really workrelated AI output. And those really do sound like completely different do sound like completely different do sound like completely different stories. I get it. Open AAI being told stories. I get it. Open AAI being told stories. I get it. Open AAI being told to slow the roll out. Apple trying to to slow the roll out. Apple trying to to slow the roll out. Apple trying to make Siri less embarrassing. F dropping make Siri less embarrassing. F dropping make Siri less embarrassing. F dropping putting claw into Slack. How are these putting claw into Slack. How are these putting claw into Slack. How are these related? The the problem is the same

related? The the problem is the same related? The the problem is the same underneath all of them. The model can be underneath all of them. The model can be underneath all of them. The model can be smart and still not know what's going smart and still not know what's going smart and still not know what's going on. If you use AI every day, you already on. If you use AI every day, you already on. If you use AI every day, you already know the feeling. You can open Codeex or know the feeling. You can open Codeex or know the feeling. You can open Codeex or Chad GBC or Claude or Gemini and and the Chad GBC or Claude or Gemini and and the Chad GBC or Claude or Gemini and and the model's very capable, right? It can model's very capable, right? It can model's very capable, right? It can write. It can reason. It can summarize. write. It can reason. It can summarize. write. It can reason. It can summarize. It can help you think. But before it can It can help you think. But before it can It can help you think. But before it can do something useful, you have to carry do something useful, you have to carry do something useful, you have to carry that entire situation into the context that entire situation into the context that entire situation into the context window. Often through uploading files to window. Often through uploading files to window. Often through uploading files to the chat box. You paste the email, you the chat box. You paste the email, you the chat box. You paste the email, you paste the memo, you explain who the paste the memo, you explain who the paste the memo, you explain who the client is, you explain which version of client is, you explain which version of client is, you explain which version of the deck is current, you explain that the deck is current, you explain that the deck is current, you explain that the Slack thread from yesterday changed the Slack thread from yesterday changed the Slack thread from yesterday changed that decision. This is what prompting that decision. This is what prompting that decision. This is what prompting has become as we've asked these models has become as we've asked these models has become as we've asked these models to do more. And then after all of that, to do more. And then after all of that, to do more. And then after all of that, when you put all of that in, the AI when you put all of that in, the AI when you put all of that in, the AI finally becomes extremely useful. And finally becomes extremely useful. And finally becomes extremely useful. And that's a really big friction point. And that's a really big friction point. And that's a really big friction point. And that is what we've described as an agent that is what we've described as an agent that is what we've described as an agent problem, right? A problem that you want problem, right? A problem that you want problem, right? A problem that you want agents to fix by going after the context agents to fix by going after the context agents to fix by going after the context window. And that's the promise that window. And that's the promise that window. And that's the promise that we've all been trying to realize with we've all been trying to realize with we've all been trying to realize with agents for the last few months. So I'm agents for the last few months. So I'm agents for the last few months. So I'm going to walk you through three surfaces going to walk you through three surfaces going to walk you through three surfaces here. I'm going to walk you through here. I'm going to walk you through here. I'm going to walk you through Apple's Siri, Claude Tag, and Codeex in Apple's Siri, Claude Tag, and Codeex in Apple's Siri, Claude Tag, and Codeex in terms of execution inside OpenAI. And terms of execution inside OpenAI. And terms of execution inside OpenAI. And I'm going to walk you through the I'm going to walk you through the I'm going to walk you through the pressure points around them. GLM 5.2 on pressure points around them. GLM 5.2 on pressure points around them. GLM 5.2 on the one hand, the delay of chat GPT 5.6 the one hand, the delay of chat GPT 5.6 the one hand, the delay of chat GPT 5.6 six on the other. And throughout we're six on the other. And throughout we're six on the other. And throughout we're going to uncover the story of why going to uncover the story of why going to uncover the story of why intelligence is getting cheaper. The intelligence is getting cheaper. The intelligence is getting cheaper. The newest frontier intelligence is coming newest frontier intelligence is coming newest frontier intelligence is coming out more slowly and what that means for out more slowly and what that means for out more slowly and what that means for all of us as far as context goes.

all of us as far as context goes. all of us as far as context goes. Fundamentally, the next useful AI Fundamentally, the next useful AI Fundamentally, the next useful AI product is probably not going to be the product is probably not going to be the product is probably not going to be the one that wins a benchmark. It's going to one that wins a benchmark. It's going to one that wins a benchmark. It's going to be the one that knows where the work is, be the one that knows where the work is, be the one that knows where the work is, what it's allowed to see, what it's what it's allowed to see, what it's what it's allowed to see, what it's allowed to do, and it's going to be allowed to do, and it's going to be allowed to do, and it's going to be something that knows that seamlessly. something that knows that seamlessly. something that knows that seamlessly. So, let's start with Siri because Siri So, let's start with Siri because Siri So, let's start with Siri because Siri is something that for better or worse, is something that for better or worse, is something that for better or worse, we all understand. Siri has been bad for we all understand. Siri has been bad for we all understand. Siri has been bad for so long that it's become a punchline. so long that it's become a punchline. so long that it's become a punchline. You can ask it something normal and half You can ask it something normal and half You can ask it something normal and half the time it either misunderstands the the time it either misunderstands the the time it either misunderstands the question or gives you a web search that question or gives you a web search that question or gives you a web search that makes you wonder why you bother speaking makes you wonder why you bother speaking makes you wonder why you bother speaking out loud. So, the easy headline for a out loud. So, the easy headline for a out loud. So, the easy headline for a long time has been, "Apple's finally long time has been, "Apple's finally long time has been, "Apple's finally trying to do something with Siri. We trying to do something with Siri. We trying to do something with Siri. We don't know if it's actually good or not. don't know if it's actually good or not. don't know if it's actually good or not. We have a little skepticism, but Apple's We have a little skepticism, but Apple's We have a little skepticism, but Apple's relaunching Siri effectively." And I get relaunching Siri effectively." And I get relaunching Siri effectively." And I get where that story exists, right? Apple where that story exists, right? Apple where that story exists, right? Apple itself is talking about Siri as a itself is talking about Siri as a itself is talking about Siri as a conversational AI assistant, is conversational AI assistant, is conversational AI assistant, is promising more natural conversations, promising more natural conversations, promising more natural conversations, richer answers, a dedicated Siri app, richer answers, a dedicated Siri app, richer answers, a dedicated Siri app, and that's all a part of the story, and and that's all a part of the story, and and that's all a part of the story, and it may well work. I've gotten my hands it may well work. I've gotten my hands it may well work. I've gotten my hands on it a little bit. I've played with it, on it a little bit. I've played with it, on it a little bit. I've played with it, but I don't think that Siri becomes chat but I don't think that Siri becomes chat but I don't think that Siri becomes chat GPT is the story here. I think the story GPT is the story here. I think the story GPT is the story here. I think the story here is that Apple is trying to make here is that Apple is trying to make here is that Apple is trying to make Siri useful by connecting it to the Siri useful by connecting it to the Siri useful by connecting it to the context in your life. like when is my context in your life. like when is my context in your life. like when is my mom landing on the plane requires mom landing on the plane requires mom landing on the plane requires context from calendar, flight number, context from calendar, flight number, context from calendar, flight number, email confirmation, uh whether another email confirmation, uh whether another email confirmation, uh whether another family member said they might go pick up family member said they might go pick up family member said they might go pick up mom instead, whether the airplane uh is mom instead, whether the airplane uh is mom instead, whether the airplane uh is late or not. And so the challenge for late or not. And so the challenge for late or not. And so the challenge for Apple is to find a way to privately and Apple is to find a way to privately and Apple is to find a way to privately and securely connect Siri to where the

securely connect Siri to where the securely connect Siri to where the context lives on your phone, right? context lives on your phone, right? context lives on your phone, right? Photos, calendar, notes, email, app Photos, calendar, notes, email, app Photos, calendar, notes, email, app state, screen, etc. And if Siri can do state, screen, etc. And if Siri can do state, screen, etc. And if Siri can do that, Siri's intelligence level doesn't that, Siri's intelligence level doesn't that, Siri's intelligence level doesn't have to be super high for Siri to be have to be super high for Siri to be have to be super high for Siri to be incredibly useful. Ultimately, the incredibly useful. Ultimately, the incredibly useful. Ultimately, the question of Siri's capability may be the question of Siri's capability may be the question of Siri's capability may be the wrong one. And Apple's answer is not a wrong one. And Apple's answer is not a wrong one. And Apple's answer is not a capability answer. It's a context capability answer. It's a context capability answer. It's a context answer. It's an answer about where answer. It's an answer about where answer. It's an answer about where intelligence lives and they're trying to intelligence lives and they're trying to intelligence lives and they're trying to push it as close to your systems as push it as close to your systems as push it as close to your systems as possible. So ondevice processing is possible. So ondevice processing is possible. So ondevice processing is Apple's goal wherever possible and then Apple's goal wherever possible and then Apple's goal wherever possible and then private cloud where it's not. So, one of private cloud where it's not. So, one of private cloud where it's not. So, one of the things that's really interesting the things that's really interesting the things that's really interesting from a product shape here for Apple's from a product shape here for Apple's from a product shape here for Apple's solution is that Apple is essentially solution is that Apple is essentially solution is that Apple is essentially saying your assistant gets better when saying your assistant gets better when saying your assistant gets better when it's close to you. And very it's close to you. And very it's close to you. And very conveniently, when it's close to you, we conveniently, when it's close to you, we conveniently, when it's close to you, we can construct a privacy architecture can construct a privacy architecture can construct a privacy architecture that means it's only yours. And that's a that means it's only yours. And that's a that means it's only yours. And that's a consumer answer to this context problem, consumer answer to this context problem, consumer answer to this context problem, right? It's an answer where Siri doesn't right? It's an answer where Siri doesn't right? It's an answer where Siri doesn't have to be able to be that smart to use have to be able to be that smart to use have to be able to be that smart to use your phone to be extremely useful. And your phone to be extremely useful. And your phone to be extremely useful. And so suddenly instead of Apple's advantage so suddenly instead of Apple's advantage so suddenly instead of Apple's advantage coming from the app store ecosystem or coming from the app store ecosystem or coming from the app store ecosystem or from the hardware, Apple's advantage from the hardware, Apple's advantage from the hardware, Apple's advantage comes from the fact that we have Apple comes from the fact that we have Apple comes from the fact that we have Apple products and we have context that lives products and we have context that lives products and we have context that lives inside the iPhone and Apple can access inside the iPhone and Apple can access inside the iPhone and Apple can access that context in ways that are very that context in ways that are very that context in ways that are very useful to us. Keep that in mind as we useful to us. Keep that in mind as we useful to us. Keep that in mind as we walk over to the work side and we talk walk over to the work side and we talk walk over to the work side and we talk about Claude Tag. Now Enthropic product about Claude Tag. Now Enthropic product about Claude Tag. Now Enthropic product announcement is pretty plain on the announcement is pretty plain on the announcement is pretty plain on the surface. Claude tag starts in Slack. A surface. Claude tag starts in Slack. A surface. Claude tag starts in Slack. A team can grant Claude access to selected team can grant Claude access to selected team can grant Claude access to selected channels, to tools, to data, to code channels, to tools, to data, to code channels, to tools, to data, to code bases. It can tag it in, and then Claude

bases. It can tag it in, and then Claude bases. It can tag it in, and then Claude just works through tasks and stages as just works through tasks and stages as just works through tasks and stages as it's tagged in, and it can respond in it's tagged in, and it can respond in it's tagged in, and it can respond in the thread. It can remember relevant the thread. It can remember relevant the thread. It can remember relevant information from channels it's in, and information from channels it's in, and information from channels it's in, and it can operate inside of particular it can operate inside of particular it can operate inside of particular permission scopes, particular spend permission scopes, particular spend permission scopes, particular spend limits, particular logs it can touch. limits, particular logs it can touch. limits, particular logs it can touch. That sounds like a Slackbot, but don't That sounds like a Slackbot, but don't That sounds like a Slackbot, but don't say that too fast because Slack has had say that too fast because Slack has had say that too fast because Slack has had bots for a long time. And the bots for a long time. And the bots for a long time. And the interesting thing is that Anthropic is interesting thing is that Anthropic is interesting thing is that Anthropic is trying to put the assistant inside your trying to put the assistant inside your trying to put the assistant inside your team's context. So on your phone, the team's context. So on your phone, the team's context. So on your phone, the context is private and messy because context is private and messy because context is private and messy because it's your life. In a company, the it's your life. In a company, the it's your life. In a company, the context is shared and permissioned and context is shared and permissioned and context is shared and permissioned and political and stale and halfwritten and political and stale and halfwritten and political and stale and halfwritten and in six places. And so the thing that's in six places. And so the thing that's in six places. And so the thing that's interesting about all of this is that interesting about all of this is that interesting about all of this is that work is happening in those messy places. work is happening in those messy places. work is happening in those messy places. And for a long time, AI has been kind of And for a long time, AI has been kind of And for a long time, AI has been kind of separate from that except in a few separate from that except in a few separate from that except in a few instances. Devon has been very instances. Devon has been very instances. Devon has been very successful with this from a coding successful with this from a coding successful with this from a coding perspective, but there's not a lot of perspective, but there's not a lot of perspective, but there's not a lot of great off-the-shelf instances for really great off-the-shelf instances for really great off-the-shelf instances for really intelligent AI coming into that kind of intelligent AI coming into that kind of intelligent AI coming into that kind of messy context. And so when Anthropic messy context. And so when Anthropic messy context. And so when Anthropic says Claude tag can build context over says Claude tag can build context over says Claude tag can build context over time, that is not only a real claim, but time, that is not only a real claim, but time, that is not only a real claim, but the heart of where the company is going the heart of where the company is going the heart of where the company is going to go. It's a very powerful and to go. It's a very powerful and to go. It's a very powerful and dangerous statement because the more dangerous statement because the more dangerous statement because the more useful claude becomes in Slack, the more useful claude becomes in Slack, the more useful claude becomes in Slack, the more it need access to messy stuff companies it need access to messy stuff companies it need access to messy stuff companies are bad at governing like engineering are bad at governing like engineering are bad at governing like engineering decisions and customer tickets and decisions and customer tickets and decisions and customer tickets and pricing debates and people information.

pricing debates and people information. pricing debates and people information. Enthropic knows this, which is why the Enthropic knows this, which is why the Enthropic knows this, which is why the launch language spends so much time on launch language spends so much time on launch language spends so much time on scopes and permissions, on admin scopes and permissions, on admin scopes and permissions, on admin controls, on channel defined memories. controls, on channel defined memories. controls, on channel defined memories. uh they recognize that they're going to uh they recognize that they're going to uh they recognize that they're going to have to earn that trust because if you have to earn that trust because if you have to earn that trust because if you put an AI teammate in Slack and it put an AI teammate in Slack and it put an AI teammate in Slack and it breaks boundaries, you've created a breaks boundaries, you've created a breaks boundaries, you've created a context leak. You've created a corporate context leak. You've created a corporate context leak. You've created a corporate liability. And so what Anthropic is liability. And so what Anthropic is liability. And so what Anthropic is saying is you can trust us with this saying is you can trust us with this saying is you can trust us with this context because you're in charge the context because you're in charge the context because you're in charge the whole time. And I think that Claude tag whole time. And I think that Claude tag whole time. And I think that Claude tag is a much better signal than most out is a much better signal than most out is a much better signal than most out there of this whole AI co-worker there of this whole AI co-worker there of this whole AI co-worker phenomenon because we have a lot of phenomenon because we have a lot of phenomenon because we have a lot of startups that are in this space. And one startups that are in this space. And one startups that are in this space. And one of the things that Anthropic is doing of the things that Anthropic is doing of the things that Anthropic is doing very intentionally here is they're very intentionally here is they're very intentionally here is they're saying, "You fed us formal context saying, "You fed us formal context saying, "You fed us formal context through prompts, uh, through co-work, through prompts, uh, through co-work, through prompts, uh, through co-work, through claude code for a while. Now through claude code for a while. Now through claude code for a while. Now trust us with informal context and trust us with informal context and trust us with informal context and enable us to be a co-worker that's more enable us to be a co-worker that's more enable us to be a co-worker that's more useful as a result." And no other useful as a result." And no other useful as a result." And no other company can say that in the same way. company can say that in the same way. company can say that in the same way. This is anthropic doing for work what This is anthropic doing for work what This is anthropic doing for work what Apple is doing for your phone. Now, Apple is doing for your phone. Now, Apple is doing for your phone. Now, let's bring in codecs. Now, a codec let's bring in codecs. Now, a codec let's bring in codecs. Now, a codec study is easy to dismiss. It doesn't study is easy to dismiss. It doesn't study is easy to dismiss. It doesn't feel like it's news, but I think the feel like it's news, but I think the feel like it's news, but I think the codec paper is really useful in this codec paper is really useful in this codec paper is really useful in this conversation because software is showing conversation because software is showing conversation because software is showing us the assistant context problem in its us the assistant context problem in its us the assistant context problem in its cleanest form. And Codex is a piece of cleanest form. And Codex is a piece of cleanest form. And Codex is a piece of software. And the study that's released software. And the study that's released software. And the study that's released is essentially how actual employees at is essentially how actual employees at is essentially how actual employees at OpenAI chose or did not choose to adopt OpenAI chose or did not choose to adopt OpenAI chose or did not choose to adopt Codeex over the course of time and what

Codeex over the course of time and what Codeex over the course of time and what they used codeex for. In other words, they used codeex for. In other words, they used codeex for. In other words, what context did they trust codeex with? what context did they trust codeex with? what context did they trust codeex with? And this is fascinating to me because And this is fascinating to me because And this is fascinating to me because you might think that at OpenAI it's a you might think that at OpenAI it's a you might think that at OpenAI it's a requirement and everyone's mandated to requirement and everyone's mandated to requirement and everyone's mandated to use Codeex. That wasn't how it worked. use Codeex. That wasn't how it worked. use Codeex. That wasn't how it worked. Codeex had to earn everyone's trust. Codeex had to earn everyone's trust. Codeex had to earn everyone's trust. Codeex had to earn trust first with Codeex had to earn trust first with Codeex had to earn trust first with engineers and then with other knowledge engineers and then with other knowledge engineers and then with other knowledge workers at OpenAI. And so the thing that workers at OpenAI. And so the thing that workers at OpenAI. And so the thing that matters the most to me when I read this matters the most to me when I read this matters the most to me when I read this study is that even at a company that is study is that even at a company that is study is that even at a company that is one of the most AI native companies on one of the most AI native companies on one of the most AI native companies on the planet, you still have to think the planet, you still have to think the planet, you still have to think about where you trust a particular AI about where you trust a particular AI about where you trust a particular AI application with context and it's not a application with context and it's not a application with context and it's not a zero to one light flip switch. But it is zero to one light flip switch. But it is zero to one light flip switch. But it is true that you can see tipping points and true that you can see tipping points and true that you can see tipping points and one of the tipping points that's evident one of the tipping points that's evident one of the tipping points that's evident in the data from OpenAI and that I have in the data from OpenAI and that I have in the data from OpenAI and that I have seen personally is that codec got a lot seen personally is that codec got a lot seen personally is that codec got a lot more useful in the last couple months more useful in the last couple months more useful in the last couple months after 5.5 was released and you can see after 5.5 was released and you can see after 5.5 was released and you can see that in the adoption data that shows that in the adoption data that shows that in the adoption data that shows that the popular adoption of codecs that the popular adoption of codecs that the popular adoption of codecs after 5.5 in non- tech circles in open after 5.5 in non- tech circles in open after 5.5 in non- tech circles in open AAI skyrocketed. Now, the thing that AAI skyrocketed. Now, the thing that AAI skyrocketed. Now, the thing that stands out to me when you put that in stands out to me when you put that in stands out to me when you put that in the context conversation we've been the context conversation we've been the context conversation we've been having is that Codeex has with 5.5 having is that Codeex has with 5.5 having is that Codeex has with 5.5 earned the trust to get legal stuff, earned the trust to get legal stuff, earned the trust to get legal stuff, recruiting stuff, sales stuff, like all recruiting stuff, sales stuff, like all recruiting stuff, sales stuff, like all of that dirty context fed into it in the of that dirty context fed into it in the of that dirty context fed into it in the way it earned trust with engineers for

way it earned trust with engineers for way it earned trust with engineers for code. And there's a lot more in that code. And there's a lot more in that code. And there's a lot more in that study. I encourage you to read it. I can study. I encourage you to read it. I can study. I encourage you to read it. I can link it. Uh, codeex is doing the link it. Uh, codeex is doing the link it. Uh, codeex is doing the opposite of cloud tag. So if claude tag opposite of cloud tag. So if claude tag opposite of cloud tag. So if claude tag is basically saying you work in slack so is basically saying you work in slack so is basically saying you work in slack so tag claude in codeex is saying your work tag claude in codeex is saying your work tag claude in codeex is saying your work is sensitive your work is important make is sensitive your work is important make is sensitive your work is important make sure you point codeex at the local files sure you point codeex at the local files sure you point codeex at the local files you care about for that work and codeex you care about for that work and codeex you care about for that work and codeex can take care of the rest and so that's can take care of the rest and so that's can take care of the rest and so that's a frame that has codeex as your a frame that has codeex as your a frame that has codeex as your launchpad codeex as your headquarters launchpad codeex as your headquarters launchpad codeex as your headquarters whereas Claude's frame is more let whereas Claude's frame is more let whereas Claude's frame is more let claude come to where you already are and claude come to where you already are and claude come to where you already are and you can give it the messy context in you can give it the messy context in you can give it the messy context in both cases there's some mess but Claude both cases there's some mess but Claude both cases there's some mess but Claude is saying that they can tackle the human is saying that they can tackle the human is saying that they can tackle the human conversation and the context and still conversation and the context and still conversation and the context and still do useful work and codeex is saying you do useful work and codeex is saying you do useful work and codeex is saying you know what give us the files give us the know what give us the files give us the know what give us the files give us the jobs and we can produce great outputs jobs and we can produce great outputs jobs and we can produce great outputs for you whether you're in legal or sales for you whether you're in legal or sales for you whether you're in legal or sales or HR I love that distinction not or HR I love that distinction not or HR I love that distinction not because I don't think that open AAI will because I don't think that open AAI will because I don't think that open AAI will release a tag codec soon these models release a tag codec soon these models release a tag codec soon these models tend to copy each other but because I tend to copy each other but because I tend to copy each other but because I think it shows the difference in product think it shows the difference in product think it shows the difference in product shape around context that these two labs shape around context that these two labs shape around context that these two labs have Claude has always been a we come to have Claude has always been a we come to have Claude has always been a we come to you, we wrap our interface around you you, we wrap our interface around you you, we wrap our interface around you kind of products. Uh Claude code was kind of products. Uh Claude code was kind of products. Uh Claude code was really exciting and co-work was really really exciting and co-work was really really exciting and co-work was really exciting partly because they basically exciting partly because they basically exciting partly because they basically said just type what you want into the said just type what you want into the said just type what you want into the terminal, type what you want into terminal, type what you want into terminal, type what you want into co-work uh and we will just take care of co-work uh and we will just take care of co-work uh and we will just take care of it for you. Now they're taking the next it for you. Now they're taking the next it for you. Now they're taking the next step into Slack. It's in a sandbox. It's step into Slack. It's in a sandbox. It's step into Slack. It's in a sandbox. It's just going to do the work there and then

just going to do the work there and then just going to do the work there and then you'll get an output. It's almost like you'll get an output. It's almost like you'll get an output. It's almost like bring your wheelbarrow of work and let bring your wheelbarrow of work and let bring your wheelbarrow of work and let us do the work and then we'll we'll give us do the work and then we'll we'll give us do the work and then we'll we'll give you an output. And it's gotten much more you an output. And it's gotten much more you an output. And it's gotten much more wide ranging as computer use has come in wide ranging as computer use has come in wide ranging as computer use has come in in the last couple months and that's in the last couple months and that's in the last couple months and that's made it much more useful and you can see made it much more useful and you can see made it much more useful and you can see that in the study but it's still that in the study but it's still that in the study but it's still fundamentally a fileshaped tool and fundamentally a fileshaped tool and fundamentally a fileshaped tool and claude is kind of a chatshaped tool and claude is kind of a chatshaped tool and claude is kind of a chatshaped tool and I realize that that is a gross I realize that that is a gross I realize that that is a gross simplification because both of them simplification because both of them simplification because both of them tackle files, both of them do chat, tackle files, both of them do chat, tackle files, both of them do chat, right? So I'm not saying it's one or the right? So I'm not saying it's one or the right? So I'm not saying it's one or the other. It's not a light bulb on off other. It's not a light bulb on off other. It's not a light bulb on off conversation. Claude has for a long time conversation. Claude has for a long time conversation. Claude has for a long time thought of the problem of context as thought of the problem of context as thought of the problem of context as conversational in the way they've conversational in the way they've conversational in the way they've designed their product. And Codex for a designed their product. And Codex for a designed their product. And Codex for a long time has thought about the problem long time has thought about the problem long time has thought about the problem of context in terms of files and it's of context in terms of files and it's of context in terms of files and it's been a fileshaped answer. And you can been a fileshaped answer. And you can been a fileshaped answer. And you can still see the legacy of that context in still see the legacy of that context in still see the legacy of that context in these moments this week. Now this brings these moments this week. Now this brings these moments this week. Now this brings us to the next OpenAI story which is us to the next OpenAI story which is us to the next OpenAI story which is around this Chad GBT 5.6 delay. If a around this Chad GBT 5.6 delay. If a around this Chad GBT 5.6 delay. If a frontier model spends the next few weeks frontier model spends the next few weeks frontier model spends the next few weeks or months in a restricted preview, which or months in a restricted preview, which or months in a restricted preview, which it looks like almost all of them will, it looks like almost all of them will, it looks like almost all of them will, we in the world do not pause and wait.

we in the world do not pause and wait. we in the world do not pause and wait. Companies still have Claude and they Companies still have Claude and they Companies still have Claude and they have open AAI models and they now have have open AAI models and they now have have open AAI models and they now have GLM 5.2 and they'll have whatever new GLM 5.2 and they'll have whatever new GLM 5.2 and they'll have whatever new open-source model is coming right after open-source model is coming right after open-source model is coming right after that. Maybe a new Deep Seek, who knows? that. Maybe a new Deep Seek, who knows? that. Maybe a new Deep Seek, who knows? and they will have a anticipation but and they will have a anticipation but and they will have a anticipation but not the reality of future frontier work not the reality of future frontier work not the reality of future frontier work from anthropic and open which by the way from anthropic and open which by the way from anthropic and open which by the way are still developing and still are still developing and still are still developing and still accumulating knowledge very rapidly accumulating knowledge very rapidly accumulating knowledge very rapidly internally. they're just not able to internally. they're just not able to internally. they're just not able to release it as fast. And so the release it as fast. And so the release it as fast. And so the government restriction is putting government restriction is putting government restriction is putting friction at the frontier of friction at the frontier of friction at the frontier of intelligence. And it means that there is intelligence. And it means that there is intelligence. And it means that there is more pressure on anthropic and open AI more pressure on anthropic and open AI more pressure on anthropic and open AI to release features like claw tag to release features like claw tag to release features like claw tag because you have to increase the utility because you have to increase the utility because you have to increase the utility of the intelligence you already have to of the intelligence you already have to of the intelligence you already have to bring it closer to context so that you bring it closer to context so that you bring it closer to context so that you get more value for the customer. If you get more value for the customer. If you get more value for the customer. If you can spend, you know, 2 minutes tagging can spend, you know, 2 minutes tagging can spend, you know, 2 minutes tagging in Claude or 30 seconds tagging in in Claude or 30 seconds tagging in in Claude or 30 seconds tagging in Claude instead of 10 minutes briefing Claude instead of 10 minutes briefing Claude instead of 10 minutes briefing the AI, you've saved yourself a lot of the AI, you've saved yourself a lot of the AI, you've saved yourself a lot of time. You can add that up, right? If time. You can add that up, right? If time. You can add that up, right? If it's something where it becomes a it's something where it becomes a it's something where it becomes a seamless part of your work, then you seamless part of your work, then you seamless part of your work, then you perceive a lot more utility from that perceive a lot more utility from that perceive a lot more utility from that even if the model didn't get smarter.

even if the model didn't get smarter. even if the model didn't get smarter. What that means is that we are in the What that means is that we are in the What that means is that we are in the middle of a context war and that is the middle of a context war and that is the middle of a context war and that is the way you should read the news for the way you should read the news for the way you should read the news for the next few weeks. I think you should be next few weeks. I think you should be next few weeks. I think you should be looking at it and saying Apple is looking at it and saying Apple is looking at it and saying Apple is battling for your personal context which battling for your personal context which battling for your personal context which because we bring our devices to work because we bring our devices to work because we bring our devices to work becomes a work context conversation. Uh becomes a work context conversation. Uh becomes a work context conversation. Uh anthropic and open AI are definitely anthropic and open AI are definitely anthropic and open AI are definitely battling over work context. They have battling over work context. They have battling over work context. They have different shapes for how they do that. different shapes for how they do that. different shapes for how they do that. And one of the most interesting things And one of the most interesting things And one of the most interesting things here is that effectively the government here is that effectively the government here is that effectively the government slowdown is giving opensource models slowdown is giving opensource models slowdown is giving opensource models time to catch up in public even if time to catch up in public even if time to catch up in public even if they're not catching up in private. So they're not catching up in private. So they're not catching up in private. So anthropic and open AI may maintain their anthropic and open AI may maintain their anthropic and open AI may maintain their 6, seven, 8-month lead over open- source 6, seven, 8-month lead over open- source 6, seven, 8-month lead over open- source models privately, but the public models models privately, but the public models models privately, but the public models we have access to may start to close we have access to may start to close we have access to may start to close because the US government is slowing because the US government is slowing because the US government is slowing down frontier model releases. And that down frontier model releases. And that down frontier model releases. And that leads to a tremendous amount of pressure leads to a tremendous amount of pressure leads to a tremendous amount of pressure on utility in the context layer. There's on utility in the context layer. There's on utility in the context layer. There's going to be a huge war over how quickly going to be a huge war over how quickly going to be a huge war over how quickly and easily an AI model can apply and easily an AI model can apply and easily an AI model can apply intelligence to that context. And so intelligence to that context. And so intelligence to that context. And so look at an Apple and Enthropic and look at an Apple and Enthropic and look at an Apple and Enthropic and OpenAI as being in the same boat even OpenAI as being in the same boat even OpenAI as being in the same boat even though we don't typically put them in though we don't typically put them in though we don't typically put them in that boat. And think about your context.

that boat. And think about your context. that boat. And think about your context. Think about what context you're Think about what context you're Think about what context you're comfortable giving to these companies. comfortable giving to these companies. comfortable giving to these companies. Think about what context you want to Think about what context you want to Think about what context you want to retain. And think about whether you are retain. And think about whether you are retain. And think about whether you are willing to put the time in to actually willing to put the time in to actually willing to put the time in to actually build elements of a harness that allow build elements of a harness that allow build elements of a harness that allow you to decide where to route your you to decide where to route your you to decide where to route your context. And so when I've talked about context. And so when I've talked about context. And so when I've talked about open brain and and open engine most open brain and and open engine most open brain and and open engine most recently, a lot of what I'm doing is recently, a lot of what I'm doing is recently, a lot of what I'm doing is basically building pieces of a harness basically building pieces of a harness basically building pieces of a harness in public so that you have more choices. in public so that you have more choices. in public so that you have more choices. And I'm not the only one doing it. There And I'm not the only one doing it. There And I'm not the only one doing it. There are others who are doing it. It's a good are others who are doing it. It's a good are others who are doing it. It's a good work. I'm glad it's widespread. There's work. I'm glad it's widespread. There's work. I'm glad it's widespread. There's there's a big movement around this. I there's a big movement around this. I there's a big movement around this. I think it's important that we have think it's important that we have think it's important that we have choice. We shouldn't have to feel like choice. We shouldn't have to feel like choice. We shouldn't have to feel like we're locked in to any given model we're locked in to any given model we're locked in to any given model provider. we should have the option to provider. we should have the option to provider. we should have the option to retain our context and use intelligence retain our context and use intelligence retain our context and use intelligence in order to get meaningful work done. in order to get meaningful work done. in order to get meaningful work done. And I think the more we look at the And I think the more we look at the And I think the more we look at the story going forward, the more it's a story going forward, the more it's a story going forward, the more it's a story of the intelligence wars shifting story of the intelligence wars shifting story of the intelligence wars shifting into the context wars. It's going to be into the context wars. It's going to be into the context wars. It's going to be less about when does 5.6 six come out less about when does 5.6 six come out less about when does 5.6 six come out and it will eventually when does Fable and it will eventually when does Fable and it will eventually when does Fable come out and it will eventually and more come out and it will eventually and more come out and it will eventually and more about when can we make the next step in about when can we make the next step in about when can we make the next step in applying intelligence so that it's applying intelligence so that it's applying intelligence so that it's useful and the story of Siri really useful and the story of Siri really useful and the story of Siri really shows us pardon me Apple that you don't shows us pardon me Apple that you don't shows us pardon me Apple that you don't have to have an incredibly intelligent have to have an incredibly intelligent have to have an incredibly intelligent model to have incredible utility like model to have incredible utility like model to have incredible utility like your model doesn't have to max out the your model doesn't have to max out the your model doesn't have to max out the benchmarks that's not what Siri is going benchmarks that's not what Siri is going benchmarks that's not what Siri is going to do but Siri applied across your

to do but Siri applied across your to do but Siri applied across your context on your phone seamlessly can context on your phone seamlessly can context on your phone seamlessly can still be incredibly powerful. So that's still be incredibly powerful. So that's still be incredibly powerful. So that's the story under the story this week. Pay the story under the story this week. Pay the story under the story this week. Pay attention to the context layer. It's attention to the context layer. It's attention to the context layer. It's going to matter a lot. And if you want going to matter a lot. And if you want going to matter a lot. And if you want more stories under the story, I do them more stories under the story, I do them more stories under the story, I do them every week. Subscribe for more.