模型迭代、规则维护与前沿工程实践的演进

Thariq: I think it's just like different models are very different from each other, you know what I mean? But I realize that it's like such a pain to like maintain different ones, you know what I mean? And yeah, like as the models get better and better, the floor of how they accomplish the simpler task is better. And so I do think in the limit CLAUDE.md goes away. And maybe not even like that far. Like I think that right now it might be better to start a new project without a CLAUDE.md.

Host: Yes. Um, I think that like maybe if you see very repeated failure modes, you add them to your CLAUDE.md. The really tough thing is that this changes per model. And so like if you've added a bunch of failure modes or like...

Thariq: >> you need Haiku.md, you need Opus.md.

Host: Uh, well even Haiku 3.5 versus Haiku 3, you know, like it is annoying.

Thariq: Like we don't do this on purpose, you know what I mean? It's just like how the models work, right? And so like maybe Haiku 3 had this failure mode that Haiku 3.5 doesn't. And if you keep this running log of a bunch of different failure modes, they will probably overconstrain Claude, you know. And so this is like we just actually added eval plugins for skills. And so now you can eval if a skill is better.

Host: Before we get into today's episode, I just have a small message for listeners. Thank you. We would not be able to bring you the AI engineering, science, and entertainment content that you so clearly want if you didn't choose to also click in and tune into our content. We've been approached by sponsors on an almost daily basis. But fortunately, enough of you actually subscribe to us to keep all this sustainable without ads, and we want to keep it that way.

But I just have one favor to ask all of you. The single most powerful, completely free thing you can do is to click that subscribe button. It's the only thing I'll ever ask of you, and it means absolutely everything to me and my team that works so hard to bring The Inside to you each and every week. If you do it, I promise you, we'll never stop working to make the show even better.

Host: Now, let's get into it. We're here in the studio with our friend Thariq from Anthropic. And I guess generally with Claude Code, there's so much sort of merging of boundaries, and you've been so on top of everything since you joined Anthropic. You have been early to Claude Code itself, and you've told that story in other podcasts. And you've also been talking about seeing like an agent. Most recently you did the talk at AIE World's Fair, "Field Guide to Claude 3.5 Sonnet", which obviously you guys launched, so that's cheating. And most recently also launching Claude Task, and we're also going to be talking about pacing on the frontier. There's a lot going on in Anthropic. I guess top of the question is, what's it like being at Anthropic when there's so much going on?

Thariq: I think that you can get whiplash sometimes. When I joined Anthropic, I joined because of Claude Code. Claude Code had just come out and I was like, this is so good. And Opus 3 to me was like, just I could not imagine how good it was, you know what I mean? And that was like a real moment for me. But I was trying to convince like my startup friends basically to use AI coding and they're like, "Oh no, our engineers don't think it's good enough" or something. And I was like, that's insane. And now you fast forward 12 months, less, and it's just like, yeah, the default way that everyone codes, right? And I think that just having to go from selling it to now teaching people how to make the most use of it and be more efficient and things like that is just like a big change. And yeah, I think it's just hard to stay on top of everything as a human, you know, things happen so fast and like...

Host: >> more agents at it.

Thariq: I mean, yeah, the agentic stuff scales much better than the human stuff where it's like, oh, there are three things happening right now and they're all emergencies, and how do you respond to it? Yeah.

Host: What do you split your time on? You do a lot of technical writing, engineering work.

Thariq: Yeah. So I think that when I joined the Claude Code team, I wanted to teach people how to use Claude Code basically. And I think that has been something that like I thought maybe I would spend a little bit of time on it, or I was spending some time on the Agent SDK first. And I wasn't exactly sure how The Bitter Lesson would go when it comes to harnesses, right? Like I think sometimes we were like, oh, what's after Claude Code, you know what I mean? And so initially I was like, I just want to teach people how to use Claude Code and make it easier to use Claude Code. And I think that has just like, as harnesses have gotten better and better, that's like the dominant problem now is like, how do you use the agents, right? Like it's such a high skill expression thing. So I do that, and then I do engineering work, I give talks. But I think when I'm doing engineering work, my goal is to take that feedback that we get from users and then be able to talk about like, hey, how to use Claude Code to do engineering, so there's kind of like a good loop there, yeah.

Host: Yeah, for listeners we'll attach the talk that you did with Sarah for the Dev Writers Meetup, which we talked a little bit about: first you do the work and then you talk about the work, something like that—sow and reap, or what was that?

Thariq: >> and reap and sow and reap.

Host: Something like that, something like that. Yeah, and then just to preview a little bit, we are going to talk about the evolution of the harness—it has come a long way from just being a CLI. We're going to talk about Claude modes, which is starting to leak today because you couldn't keep it secret... Yeah, there's a lot there. I think you started off with like adding the "Ask user question" tool, which people would love and hate actually. Like I actually thought it was like very innovative, and then now I have like my own version, you have your "interview me" version.

Thariq: >> Yeah. Yeah.

Host: And everyone just has like their own stuff, and it no longer matters because now you're supposed to write prompts that create other prompts and loops and all these things. Sure. So what's the state of the art today? What are you telling people to do today?

Thariq: Yeah. "Ask user question" was the first time that the model was good at elicitation. You know, I think this was like kind of an emergent behavior that I wanted to see if the models could do. I have like kind of a human-computer interaction background, so I did that in undergrad and grad school. And so this was like, I think it's kind of like human-agent interaction to me, trying to figure out like how can the agent communicate with you and extract the requirements, right.

I think that one of the things that's difficult as Claude Code has gone broader and broader is that everyone has like their own way of using it, and it's actually very hard to change the default behavior. So for example, if someone asks Claude Code to do something, sometimes they just want them to do the work because they're maybe a very good prompter. And sometimes they actually are not good at prompting, you know what I mean, and the agent needs to clarify. And so that's like a good split. And the "Ask user question" tool sort of splits along that side, where like: do you feel like you're good enough to instruct the agent as it is, or does the agent need to pull out more requirements and collaborate with you more?

Speaker A: and really understand your preferences. I on the whole believe that pretty much everyone is more on the latter than the former, that they like have more ambiguity and they know less than they want than they like think they know about the problem. But like we're it's like interface design problem to make that easy. You know what I mean? And so like if you're designing a problem or if you're going through a problem, like you know, things like what's the schema or like what's the call stack and things like that are really important. Um, you know, like the details in the design are important. Ideally you want to figure out some of these like hard problems ahead of time before starting implementation. And yeah, that's why they call like unknowns, right? And so I think that this will forever be like a skill in agentic coding is like figuring out your unknowns. Because even if the model is like super intelligent, it like needs to know what you want, you know, and like you have preferences, like you need to sort of like pull that pull that out. And so that's like I think what I'm pushing.

Speaker A: The question then is like how does the agent interact with you? And I think that has been HTML has been like the big way of doing that. And we've recently added artifacts, right? And artifacts I actually think we've done a bad job of like or like I've done a bad job of like explaining how to use them fully. We have a lot of property capabilities. They have a database associated with them, you know, and so every artifact can store and write persistent data. They can like feed back into Claude, you know, and so like one thing that you know like people are not doing yet that I'm trying to like encourage is like this idea of a dashboard artifact. So you have like Claude working on a project long term. Maybe it's like a Kanban or something. It can store that Kanban data in a database. Multiple Claudes can access that data via like the artifact MCP and like that artifact can like talk to those Claudes as well. And so like we're basically building the primitives for you to be able to have this like generative interface via artifacts that will like let you surface more of that rich detail from the agents.

Speaker A: And I think that like almost everything with agents right now is like this problem of like you think you know what you want but you don't really know what you want. And like the agents need a lot of detail. And collaborating with them in the loop is really important. And so artifacts are like the way that we're trying to evolve there. But there's a lot of work to do cuz it's so much more complicated than like a multiple-choice question, you know? There's a lot more like detail in terms of like diagrams and code snippets and schemas or like whatever it is for that problem. But like artifacts is like the more AGI-pilled way of like, you know, doing 'ask as a question' basically. So yeah.

Speaker B: >> I think one thing that's unclear to me about this uh the the artifact stuff is like what feedback should go in through the artifact and what feedback should go through a Claude chat, because the more AGI-pilled one is to just feed everything to the Claude.

Speaker A: >> I think the more AGI-pilled one to go through the artifact. Like and I I think that like we sort of imagine in the limit I think that artifacts will be your interface into the harness. You know, you can like comment on this like live like document of your plan of the work. You can see maybe like multiple agents and different agents are doing this and that artifact is built for the current work that you're doing, right? And so like each one has like slightly different. I think we're still like getting there from like an infrastructure perspective, but yeah I think like on-the-fly interface for your harness is probably where things are headed.

Speaker A: Is there a version of it that's an abstraction from CLI or chat and you because right now a lot of it is, okay, you're interfacing with Claude Code, you're having HTML given back for a mockup, it's pretty rich, there's diagrams, artifacts are ways to connect these together, why not just do everything that way? Then it becomes like separating out like where's the inference happening, where's the intelligence happening, where is the work happening, you know? Like I think this is kind of like difference between like or like some of the distinction between local and cloud, right?

Speaker A: And so, um, I think right now if you use Claude Code, it's like local and, uh, like you can spin off Remote Control for example to get some cloud behavior or you can spin off Claude Code in the cloud, right? We're moving towards a place where instead of Claudes uh, like you message a local Claude, it starts a session locally and it executes to more like you have a Claude that you message that's in the cloud that's running. It can run like local or like cloud sessions. This is kind of how Claude Tag works, but like over time we'll add like local hands as well. And and so like local hands will be the ability for that agent to access your computer if it's online, you know, and be able to like work there.

Speaker A: And so it can spin off many different sub-agents. It can like those sub-agents can communicate with each other. And that's where the artifact comes in to display all of that work basically. So you can imagine like the you're separating out these things. So there's like the surface UI display that's an artifact and hosted somewhere and has a database and everything. There is the inference intelligence, right? That's happening on the cloud and you don't have to worry about shutting off your computer or whatever, right? Um and then there's the like hands kind of like and it can be local, it can be in like a remote sandbox or wherever you need your work to be done. That's like unpackaging like the Claude Code experience. Right now we're like by right right now it all happens in one place, right?

Speaker B: >> how do you see like the multiplayer side of that? So say teams want to work in this way, right now it's very individual, but how do you see the future of multiplayer? Like right now I guess there's Claude Tag which is a version, but

Speaker A: we're launching Projects and so Projects is the like this abstraction that's kind of like Claude Tag but on our cloud products, right? So you can message it and like it will do the Claude Tag like stuff like spinning off sub-agents. So we think with multiplayer like Claude Tag is like a little bit more native multiplayer because it's just like in your Slack and the permissions are all figured out and stuff like that. But I do think multiplayer is like an important part of the story and like uh that that will need to get tied together more.

Speaker A: Like you can imagine how complicated it gets when you're like, "Oh, you have hands, but now you have other hands in other people's computers too and like you need to like permission them or like you have like your MCP and someone else's MCP and how do you figure out how to use them?" Right? It gets like quite complicated and Claude Tag does a good job of like sanding down all of these issues, right? So that like when you have—yeah, Google Docs, how does it access Google Docs, right? Like it accesses through the shared cloud MCP or it can access through your local credentials as well if it doesn't have access. But yeah, like I think Claude Tag is our multiplayer um product and it's really useful for these like things that are inherently multiplayer like okay like on-call for example incidents are inherently multiplayer. You want to tag Claude, you want multiple people to log in, you want to be able to find context.

Speaker A: Um, I think whenever I'm like working on something and I want like privacy or security or like I want other people to review it, you know, it's really nice to like I'll have a channel per project and I'll like @legal for example be like, "Hey, like I want to ship this. Can you like like here's the Claude knows everything, you know, just chat with it." And that way legal gets precise answers, you know, on like what exactly is shipping into the code and I don't need to be in the loop, right? So I think like multiplayer is getting like more and more like, yeah, everyone can participate with Claude. Um, I

Speaker A: I think Claude Tag is like that product, and like Projects, we'll start off single player and we'll like, you know, expand.

Speaker B: I think there's a question about like maybe dual questions about identity and the unit of isolation.

Speaker A: Yeah.

Speaker B: Tag, you specifically chose to make it its own identity, which is like a controversial choice. There's other ways to do it.

Speaker A: Yeah.

Speaker B: Cloud Projects, probably it sounds like, you know, if it's anything like ChatGPT Projects, the isolation is that artifact, that cloud instance everyone's collaborating on. It sounds like if you're collaborating on the thing, like that channel should be a Project, right? Like it's not yet, but that's the natural next step.

Speaker A: Yeah, I mean, I think in Claude Tag it's effectively like you have to sort of do your own arrangement basically. And so in Claude Tag, each channel you can name it as you want, and I name like each feature basically as a channel.

Speaker B: But I think like there is some transference... it's unclear when there is transference. Because let's say if you have a coworker who is tagged on all these things, yes, there is transfer because it's the same person. But with Claude, it's unclear if it's like necessarily like, "Well, no, you don't know about the other stuff, you should only use this stuff."

Speaker A: It's like the tip of the iceberg meme, right? This is what we spend so much time on basically. There is like a lot of surface area to figure out of permissions and visibility and how can you let Claude operate as well as you can as safely as you can. And obviously this is very important to us because security for our codebase is very, very important, and so we've put a lot of time into this.

Yeah, there's so many edge cases you can figure out where it's like, oh, yeah, this Claude in this channel has different permissions, but it can message another channel and can't it exfiltrate data that way? Or like what if it uses your MCP and then messages someone else? There's like so much, and we've really put a lot of work into sanding it down.

Speaker B: Yeah, lots of work. Okay, Fable. Fable, you wrote two good articles—I mean, you've written many good articles, but on field guide to Fable building Claude Code. I'm curious from what you've seen, is there any common patterns that you see in like top users at Anthropic, externally? Like what are best practices for getting the most out of Claude Code?

Speaker A: The meta skill I'd say is prompting is very important. I think this is not trivial to say because I think a lot of people are like, "Oh, prompting doesn't matter, it's just like I can just say a sentence and Claude will do it." And I think prompting is really like public speaking or writing for a specific audience, and that audience is Claude. You need to build a mental model of Claude and how it thinks and how it works.

And so that's the most important skill in working with Claude Code: having this mental model of Claude and what it can do well, what it can one-shot, what it can't. And so many people when you see prompting, they're just short prompts, but they have such a good mental model of Claude and of the codebase and things like that that it's effortless, you know what I mean? But it's like high skill ceiling. So that work of spending a lot of time prompting and building mental models, an intuition for how the agents work, is really important.

Speaker A: And then I think the next thing is the unknown stuff we talked about earlier, where it's being able to find out what you don't know or what you haven't written down, learning about different things. I think as Claude can do more and more things, the likelihood of you doing something out of distribution for you and you have low domain knowledge on is very, very high. And the more you can learn the vocabulary to be able to prompt Claude, it becomes really important. And so I think the most important unknowns are the unknown unknowns where you don't even know that this exists.

Exactly. I think this illustration of the map and the territory: where you're like, "Okay, this is my prompt," and the territory is the actual work that the agent needs to do. And if you are very precise, you can give more precise things. So like for example, in design, I'm not very precise. I'm not a designer, so I say like, "You know, give me like eight different mockups." But if I was a designer, maybe I'd be like, "Oh, hey, here's some reference sites, like I want this type of font and this type of look to it, and here's like a few different components to visualize, here's a Figma board to bring in." You can just be so much more precise with that language. And if you're not a designer, you just need to try and learn the language basically, or learn the unknown unknowns. And this is true of everything: the more you can work with Claude to learn how things work, the better your prompting will be.

Speaker A: I think another good example of this is game design, where a lot of people are like, "Oh, I can vibe code a game now," and then they're like, "It's not fun." The thing about game design is every one of these choices has a lot of variations, a lot of craft to them. So when you're making a flying game, the feel of the plane and the way it responds to your controls has a lot of... a game designer would spend days on that, you know what I mean?

Speaker B: To me, that's what taste is, right? Like it is like from the possible space of 1,000 mathematically valid answers, here's the one that the humans will like.

Speaker A: Yes, I think with taste I'm like torn on this word, cuz I think you're right, but everyone has different definitions and it sounds kind of like low skill or like elitist almost, where you're like, "Oh, there are certain people with taste."

Speaker B: Like "Taste is what I call taste; these guys don't have taste."

Speaker A: Yeah, exactly. "Oh, an engineer doesn't have taste; like I, the founder, have taste." And I think that's actually not true. I think engineers have a lot of taste for these particular problems, and I think everyone has taste for particular problems. I think Jason Lou said, "In order to have taste, you have to eat." And I really like that, where it's like, okay, you have to do a lot of things. You have to iterate and figure out what you want, what you like, and build that domain vocabulary. And then when you're prompting, you're like synthesizing all of that.

Speaker B: Isn't it annoying when someone else says it better than you? Like, I have to quote this guy forever.

Speaker A: Having to quote Jason Lou forever. He's going to love this.

Speaker B: So I get props. And sometimes it's actually not even that. Sometimes it's just intuitive, right? Like you don't realize you even want something till a model puts it out and you're like, "Oh, this just feels immediately better." Right.

Speaker A: Yeah, exactly.

Speaker B: One thing I go back and forth on is I feel like the way I prompt half the time, let's say I use voice—you guys have voice, other people have voice—that is the opposite. It's just like me rambling for like two minutes, pressing down the function key and then let go and then hopefully it figures it out. And oftentimes it does, but it's not as thoughtful as a structured prompt with well-run communication as though it's a PRD.

Speaker A: ...or memo. Is that in line with how people do this? There's like basically bimodal prompting where there's some prompts where you spend a lot of time upfront and other prompts you just dash it off.

Speaker B: Um, I don't think the voice is necessarily low. Like I think it's like more like how much information is in the prompt, you know, like like the model can like you can um and uh and like add some sentences and be like, "Oh, like actually I changed my mind like in the middle of the prompt and it will be able to follow that perfectly." You know what I mean? So I think the like actual format of the text is less important, but then like the ability to like how much information is in it, right? And I think for voice a lot of times, you know, going back to like kind of human age and interaction and like for a lot of people it's just way easier to talk than to like type, you know. Um, and I if that gets more information out of you, like that's better.

Speaker A: At some level, it feels like just giving the model as much context over prompting before you kick off is a best practice. I don't know. A lot of the times, like when I was first trying out Fable, I spend a solid 30 minutes like really crafting a long problem. This, I think, is a response of models running for longer and longer, right? It's still a little difficult to nudge them as they're in in like, you know, in the loop. But I I just like intuitively spend more time kicking off that first prompt and working with it a lot.

Speaker B: My personal opinion is that if I was a software engineer, if I was like, you know, just running my own startup, for example, I think I would mostly stick to a max 20X, you know what I mean? Um like maybe verification and code review are kind of like separate things, but I think like what I see a lot of times is people hit rate limits when they're doing this sort of like oh like it did a lot of work and you're like oh I don't like this like can you like undo this and redo it and and then you're like iterating on this like thing that the model could have done if you had like spent more upfront time or given it better context, you know, and instead it's sort of like you're like nope, don't like that design, try this or like you mess this up or something like that and then that just eats up so much more of like you know your usage and so that's like I think maybe like a key like tip both for like efficiency as well right and yeah I think like context and not just like context on like what the goal is you know what I mean is good right like are you building a prototype or is it like a production thing like where can you spend compute or where can you not spend compute like I think you have to give the model permission or like not permission to do things sometimes where you know like it doesn't know intuitively how much you want to spend on this task, right?

Speaker B: And you can you can use effort for this. So I'm working on a blog post about that where it's like you know uh if you want for like we see that effort scales with basically the complexity of the task. So for security effort gets like way more results like high effort versus like low effort gets like changes the eval but for software engineering it doesn't change it a huge amount because effort is mostly spent on the verification and the like edge case testing and things like that and so like being able to like give the model that guidance of like hey this problem is something that I think I want you to spend a lot of time verifying and edge case testing you know

Speaker A: how about model in the mix so you There's Opus and Fable with effort. There's also Haiku in there.

Speaker B: Yeah. Yeah. It's not quite true yet, but it's very close where I think the frontier models will be paro dominant over like almost everything, you know, like maybe and and sometimes I think I think Opus might be paro dominant. You know what I Like I think depending on like how things uh like shake out if it's like a newer version of Opus, but I think that like increasingly it's just going to be like the smart model is going to be able to like do the simple task for less tokens than the like the the other models basically because of verification. With verification in the limit your model doesn't need to verify, right? if it's a perfect model, it just does the work once and it's like, okay, like you I did it, you know? And increasingly with Fable, I'm like I'm like, dude, you don't need to spin up Chromium and screenshot all of these things. Like I I see it like you you did it, right? And so a lot of the at higher effort you spend more of those tokens verifying but if you're working on simpler problems and a lot of software engineering is like you know like well like in fable like low and mediums ability it can spend less tokens verifying and as the models get smarter and smarter they will just be able to like all right done you know like I can run the lint for sanity's sake but like I like know it lints you know what I mean like you don't even need to do that and that will be so much more token efficient than like the the smaller models basically. Yeah.

Speaker A: Is there a good um practice on our side that we can use to see if we're using too much effort? Like I I freaking hate wasting time on that kind of stuff.

Speaker B: Yeah. Yeah. Yeah. I know what you mean. I think like so in in this blog post um my rough distribution is like code review and security should be like high or max basically and like software engineering settings per domain. Yeah. I think like if you're doing like UI or something like that like low and medium I think is you're building like an API and you want to make sure like you cover enough edge cases, you know, and so I think building like I said that mental model of like how things work across these distributions is like yeah part of the job.

Speaker A: This is more intuition driven or eval because I'm guessing this would change as well.

Speaker B: He has evals.

Speaker B: Yeah. Yeah. So what I did in the blog post is I go over all of the terminal bench evals basically. So there are like 70 problems and I'm show that like okay you know like in the security problems it does more um and then I also like look at some of the transcripts just in terms of like how what does it answer what does it forget or something and a lot of times this is another prompting tip I have is like asking it to make decision notes or implementation notes because a in basically every eval problem that it faces it thinks about the correct solution, you know, and decides not to do it, you know, it's like, oh, like here's the answer. What if I did this? And then it's like, oh, probably not, you know, um, and then keeps going. And this is like the majority of the failures, you know what I mean? At like a higher max level, it's very rare that the model just doesn't know how to do something. If you just have these implementation notes, then you can review and you can be like, "Oh, actually, I want you to do this thing that you didn't do." the models are getting better at surfacing that overall. Like I see in the transcripts that fable 5.1 like when it does this output it will call out its decision making as well. Um but making this more explicit in the harness is better. And now we're you know allowing ways of you modifying the harness so you can like you know add some help there. Yeah. Uh yeah.

Speaker A: So I do want to call out two things that you mentioned that I think actually exist outside of prompting. one is actually like let's let's call it the the prompt that is so important that it shouldn't be in a prompt is actually in cloud MD or agents MD which is like goals right like your your situation your goals the things that you want the the the thing uh and then the second of all is the decision log or the experiment log or whatever log of traces that you might want to actually survive the current session to to do those things those are like externalities that there's no standard there's no it's not like skills it's not like MCP there's no standard it's It's just like it's a markdown file. Uh first of all, is that right? Is cloud MD going away? You have a documented dislike of of

Speaker A: ...agents.md, but you're going to do it.

Speaker B: Um yeah. Yeah. Okay. So, I mean, agent.md, yeah, like we're going to do it. I think it's just like different models are very different from each other, you know what I mean? But I realize that it's like such a pain to maintain different ones, you know what I mean? And yeah, like as the models get better and better, the floor of how they accomplish the simpler task is better. And so I do think in the limit CLAUDE.md goes away. And maybe not even like that far. Like I think like...

Speaker A: I think that right now it might be better to start a new project without a CLAUDE.md.

Speaker B: Yes.

Speaker A: Um I think that like maybe if you see very repeated failure modes, you add them to your CLAUDE.md. The really tough thing is that this changes per model. And so like if you've added a bunch of failure modes or like even...

Speaker B: ...you need Haiku.md, you need Opus.md.

Speaker A: Uh well even Sonnet 3.5 versus Sonnet 3, you know, like it is annoying. Like we don't do this on purpose, you know what I mean? It's just like how the models work, right? And so like maybe Sonnet 3 had this failure mode that Sonnet 3.5 doesn't. And if you keep this context with this running log of a bunch of different failure modes, they will probably overconstrain Claude, you know. And so we just actually added eval plugins for skills. And so now you can eval if a skill is better. I think Daisy on our team did this. And so yeah, we're trying to work on this. We know you still have to spend tokens on it and it's not perfect, but we're trying to help out with this problem.

And so as far as prompting goes, the one tip I want to offer is something I have told people a lot: sufficiently advanced prompting is indistinguishable from sufficiently advanced executive communication. So I've actually referred to this: it's the executive comms workshop from Heavybit that is the best I've ever seen in my career, and they teach this thing called the SCQA model. Just Google it. It's a thing—people have done prompting for decades, it's just called executive communications. It's like when one person has to communicate to thousands of people down the org chart, this is what you do. Situation, Complication, Question, and Answer. It's how you write the memo. Obviously sometimes you don't have the answer, but you can actually at least list out the S, C, and Q, and then they have some examples in there. So just leaving breadcrumbs for people if they want to explore.

Speaker B: I mean before we move on, I want to ask you any other underrated tips, ways people could get a lot of value from Claude Code that they're not using?

Speaker A: Yeah, I mean I think a lot of them are in this unknowns doc. I give a bunch of example prompts, like using it for brainstorming, using it to quiz you after.

Um we added this like "explain it like I'm five" skill actually which is a very short prompt, and it doesn't even say explain it like I'm five. Basically the key word of this prompt is big picture in few words, you know, like that's like the main thing, and it is shockingly good, you know what I mean? I think I tweeted about this basically and it's like /eli5 and you can install it as a plugin, but yeah it's like way better at just cutting through the BS and being like yeah yeah exactly right here. So the diagrams are like quite clear. I think one of the things that is true with artifacts is like they put too much text in and people are not reading the artifacts, you know, and so like this simplifies it a lot more. And yeah, this came out of like just people at Anthropic like going through very complicated incidents and being like what is happening, you know. So this one I think is great, yeah.

Speaker B: My version of this is actually the... it's like test your understanding, give you a few...

Speaker A: ...choices, and then like if you actually get it wrong you have a mismatch between what you think is happening versus what's actually happening.

Speaker B: Yeah. Yeah. Yeah. I think this is one of those things that everyone loves talking about and then very few people really do. Like I, you know, I think...

Speaker A: Yeah. Most people just don't want to get quizzed about something. You know what I mean? Unfortunately, I think this is one of the like...

Speaker B: ...things that we need to like...

Speaker A: What's the opposite of ask you a question? Ask the question before the thing. Uh this is after the thing.

Speaker B: Exactly. Yeah. Yeah. It's a good way to stay grounded of like do you even know what you're doing? Right. The worst case is when people send you slop and they haven't understood what they're asking for or what the output is and it's like, dude, I don't want to read this. Do you even know what it is? So, you know, you make it a rule for yourself that before you send stuff, you should at least know what's implemented.

Speaker A: Yes. But so, you could make this a mod and you could build your own mod to like make sure you test it. So, yeah, we can talk about that.

Speaker B: Let's get right into it. What is Claude Mod? Um, and what is this diagram showing?

Speaker A: Yeah. Okay. So Claude Mods is basically you can customize the entire Claude Code harness and we're going to... if you have requests, please let us know, we'll add more and more. This works for CLI, it works for desktop, maybe it will work for Cloud web in the future, I don't know, we're trying to make this very, very extensible.

You can see this reference sheet, I don't want people to get overwhelmed by it. At a high level, you can customize both the execution of the harness and the UI of the harness. And so like you see in that Tetris example from Boris, that's like customizing the UI, right? Like basically showing Tetris in the game.

But like let's say that you wanted to do this thing where you tested your assumptions or tested your understanding after every project, right? What you would do is you would ask Claude to make this plugin. It would spin a classifier after every prompt basically. And so like at the end of each turn, you would spin off a subagent or like a forked agent. Basically, a forked agent maintains the prompt cache, right? So it's one of those unintuitive things where you can fork and do like a little request and it'll be very cheap because the entire prompt cache is done. And so you can be like, "Has this task been completed?"

Speaker B: This is how you do BTW and all those.

Speaker A: Uh yeah. Yeah. The underlying forked agent. Yes. But so you can in the forked subagent you can say like, "Has this task been completed? If so, return true." And then in your hook or in your plugin mod—or sorry, like in the subagent probably—you would say like, "If true, give me a quiz, you know, give me questions and answers in a JSON format." And then you'd parse it, and then you display above the prompt input basically this list of questions, right?

And so this is something that's like slightly token-intensive because you have to sort of do it after every end of the assistant turn, but it's like a lightweight classification and then you can get this quiz, and you'll see Claude will always do it for you. You don't need to remember to do it.

There are lots of these tips that we've talked about, right, where it's like, oh, implementation notes. You can also add a tool for implementation notes now. And so, like this tool that I'm adding is like "register assumption", I think is what I'm calling it, but maybe I'll change it around. And this is a mod. And so you give it a register assumption tool and then it will keep a list. Every time it does, it'll add to the list and then at the end it will display those assumptions.

You know, another mod I'm working on is a model router, and so like internal Claude model routing, right? So this is... I want to say the reason we don't do model routing by default is like it's a...

Speaker B: ...hard problem. You know what I mean? And like

Speaker A: >> you will get it wrong.

Speaker B: >> Yeah. you like yeah you will like accidentally use like fable for a hard problem or sonnet for

Speaker A: >> you have auto approve but you don't have auto mode

Speaker B: >> well you will have auto like like you don't have like auto routing or something

Speaker A: >> auto mode for model picker

Speaker B: >> yeah yeah exactly exactly so

Speaker A: >> I'm getting the rough question of like how much do you open this up and how much do people have to think about this like when you talk about prompt caching and building a router uh it seems like you could easily build a mod that routes per query and I'm just killing my plan very fast, right? I guess my question is more so like what is like a product dock like this look like, right? Who is it for? Is it for power users? Is it everyone should be able to

Speaker B: >> definitely power users, right?

Speaker B: >> Yeah. I mean, I think it is power users, but like the nature of claude code is that so many people are power users, you know, because it's easy to share things, you know, like you can like one person can make a good m router thing that doesn't break prompt cache all the time and then you can like, you know, sort of compose them. Another cool thing about the plugins is that they can hook into and compose with each other. And so I have like a uh a mod that will like create a mode selector at the top and any plugins can register to be a mode. And so like the auto router can be a mode, right? Or like you can have a mode that's like artifact mode where it's like it it primarily talks to you in artifacts. It's kind of like, you know, you you can toggle between plan mode, you know what I And so like you can create more and more of these modes, but the ability to create modes is in itself a mod, you know, and so there's a lot of richness here, but we do want to make it fairly easy. We want to be make it so that you can just like install someone else's, you can talk, you can chat with Claude and you know, we'll like make sure that it understands the nuances of things like prompt caching and stuff so it can like warn you. This is like not extremely complicated behavior for Claude, I think, but we should have just a good skill on how to make mods. Um, and yeah, we'll see how we go. But I I do think that this is like a preview of like mutable software, you know, and like how like uh generative software just like you can customize safely. If enabled, you could customize any piece of software. And I think that more and more apps ideally do something like this, you know.

Speaker A: >> And by the way, we have you have another cool tweet about how you know there's the infinite money button which is like make your SAS uh consumable by agents. I think mutable software is interesting and uh you know other people have also tried to do it. I think the hurdle comes when you can do everything then people users get tend to get confused. So usually the stuff that works is just like one opinionated flow. This is in the side of less opinionation. It's just like well more power to power users and I think probably unlocked by AI where like you can just prompt for whatever the thing it is.

Speaker B: >> Yeah. or there can be a skill that gives the opinions you know and and then yeah

Speaker A: >> so knowing a little bit about like Typescript and build systems and all these things uh the closest I I'm actually very curious the team who worked on this I don't know how close you were to them uh if they drew any inspiration from build systems like Babel Webpack all these all these like old school things because it sounds very similar like the plug-in ecosystem those things where they can compose to each other

Speaker B: >> yeah I mean I'm not deep in the technical details but I do know it was a collaboration with someone on the bun team and someone on the claude code team

Speaker A: >> build system.

Speaker B: >> Yeah. Yeah. Exactly. It's it's very exciting. But yeah, like agents can just do this very complicated sort of like extensibility into your software now. And so um yeah, like you know another reason to like if you run a startup like you can just prompt cloud and be like hey like could we make an extension system like what would that look like you know?

Speaker A: >> Yeah. Yeah.

Speaker A: >> And I just really wonder like you had hooks in the past and plugins all these things. So what specifically will mods be able to do that those things could not do

Speaker B: >> internally? We were originally calling this function hooks. And so like that's like gives you a little bit of an idea where like hooks sort of register a like an event to happen and then like a script to call basically. And this basically inside of the like TypeScript runtime is running things. And so like you get some benefits of just like it has a bunch of things in the scope with like for example like how many turns is in this conversation right like how many tokens have been used like et like what are the messages things like that so has a bunch of messages that can be used and then it's just like a lot more hooks basically so we have you know like or a lot of lot more like you know fun like things you can register on and then you can do because of the because it's all happening in process you can uh spawn sub agents you know with for context and context and stuff and like that will return you can parse the results of those you can use structured output to sort of like return them um and then you can modify the UI which you can never do in in hooks. Yeah.

Speaker A: >> Yeah. So, modify UI, this is why you show the Tetris example. Does it also extend to artifacts? I assume it does.

Speaker B: >> You like artifacts are kind of like a different way of customizing it, you know, like you can definitely one of the mods I'm working on is like this dashboard mod which will like sort of prompt Claude to maintain a dashboard that's an artifact. But they're kind of like slightly orthogonal or not orthogonal. They compose with each other in different ways. like mods are like a little bit more like in your cloud code harness uh changing the agent loop, you know, and like the UI is like an added benefit. Um, and then artifacts are just like you want to, you know, see things at at high level, very inter highly interactive, you know, like the affordances can be a lot bigger than, you know, like a 2y or even in our desktop.

Speaker A: I'm guessing you'll have a good blog post on the differences cuz right now you can also, you know, make a loop that outputs to an artifact that's an interactive dashboard, but you can also do it with a mod. There's just some thinking about making a hack uh hacking on a harness when we don't know much about the harness, right?

Speaker B: >> Well, something I'm excited about with mods is like there's so much things for claude code that you just have to remember, you know? I mean, you're like, "Oh, like let me do this and then let me call the dashboard skill that does the loop and things like that and or like let me test my assumptions afterwards." And I think like if you do all of these things using these little classifiers and stuff and you're like, "These are the things I care about. This is what I want to do." Um, you can like you don't have to remember as much. One more like mod I'm working on is a next steps mod that

Speaker A: >> I have I was going to say I have a next step skill. I always run next steps. And does it have access to your skills? Like this is one of those things where I'm like

Speaker B: >> I think so.

Speaker A: >> Okay. Yeah.

Speaker A: >> Need specific access to it always has

Speaker B: >> I think there's like specific prompting I guess to like know your skills kind of like like I think Claude forgets them sometimes throughout like uh the thing. But anyways, the idea of like yeah next steps that also are like oh hey this has happened use the explain skill to explain to you what happened because this seems like quite complex you know or like uh yeah use your unknown skill. It looks like you are like asking the model to like you know iterate on these small changes. It seems like you could prompt better you know like what if you did this right? So like I think um yeah like spending more compute there. Yeah. Yeah. Yeah. And it should always come out as multiple choice. Uh we have we have a I have my

Speaker A: My next step skill is like this.

Speaker B: Okay, perfect. Yeah, you can steal.

Speaker A: Yeah. Yeah. Yeah. Like but like for me it's I think models really always need to be reminded what are you trying to do here?

Speaker B: Yeah.

Speaker A: Look at the whole transcript and go like oh was this original goal? Did your solution actually solve it? Were you lazy? If you're lazy maybe there was a reason. Maybe you needed approval from me. Maybe there's two things you want to suggest. So it's a little bit like the modification of the ask you a question or interview me skill. So it's next steps.

Speaker B: Yeah. Yeah. Exactly. And again the benefit of doing it with mods is you can do it as a fork sub agent and so it doesn't remain in the context afterwards. So you have this idea of like okay the model is doing its execution and you have this almost like supervisor that is making sure that you can do the next steps well. So yeah.

Speaker A: I do have two panels and like I often try to have a supervisor thing keep the high level context and then the implementation detail in another agent. I feel like a lot of this abstracts away as models change. You know, half an hour ago you said bitter lesson of harness engineering and we're on the other extreme right now.

Speaker B: So yeah, exactly. If everything's customizable, what actually is Claude Code, right? And which I talked to you about last night.

Speaker A: Yeah. I mean, I think the bitter lesson is unintuitive in terms of like we're kind of misusing a little bit of the bitter lesson here where it's more about scaling and compute and stuff. But I think there is something where I use it as an approximation here to say that harnesses go out of date very quickly, and how they change is unintuitive. So the big obvious example is from chat to agents where you had to give them entirely new tools. But I think this new version of oh it can modify its own harness, right? This own harness loop is a way of using its capabilities, or like it can build an artifact. And the way I think about it is the models have more and more intelligence and they're so much more intelligent now than the average software engineering task. Like you look at the Terminal Bench ones and they're like solve the Jacobian conjecture. Not really, but they're quite complex. Like I would not have been able to do this really as a software engineer.

Speaker B: You get to TB4 or TB2?

Speaker A: TB3. Yeah. Yeah. They're quite complex, but the goal is still to deliver user value, right? And like you said, there's like this infinite space of things to do. And so the ways like you spend compute are to keep the user in the loop and make sure that like you're getting to the right decision in the end of the day and like the right output and artifacts and mods are this way of like spending that intelligence basically. Uh and I think that's like yeah the next step and so yeah I think Claude Code has the core things of agent loop which are have gotten more complicated. It needs a sandbox to operate safely. It needs auto mode to make sure the permissions, approvals. It needs computer use and MCPs and all of these ways of accessing your data, and it needs web search and web fetch. As the models can do more and more, the core harness has to be actually quite complex and very secure, but then how you interact with it can change quite a lot. What other harness engineering best practices have you from the Claude Code team itself? I feel like there was a phase of plan mode which is not as used. We now have auto mode. At a point you cut the majority of the system prompt. You got rid of examples. What other best practices are there for harness engineering? I think there is a forking path where at some point eventually yes the model will just be able to vibe code the exact version of Claude Code even describing all this complexity that I've talked about, right like auto mode and computer use and stuff eventually the models will just be able to do that in one shot. But I think they can oneshot simpler harnesses. And so sometimes you don't need this full setup if you don't need computer use or all this more complicated stuff. Before, you had to sort of use things like the agent SDK which was Claude Code wrapped in order to build, and I would suggest people do that because there was so much complexity into building a harness. And now that's got more abstracted. We have Claude Managed Agents which lets you have that complexity but still write a very bare bones harness that's scoped to your task. Yeah, I think there's this barbell effect where for very complex coding tasks and these complex things you should use our harness, and then for a lot of simpler or more domain specific things you can build your own harness because Claude has gotten better at building harnesses and we have these harness primitives like managed agent. So yeah.

Speaker B: Yeah. Is there a general progression? Let's say chapter one was ultra code dynamic workflows then chapter two was Claude mods. Where is this going, where you can sort of customize the thing on demand?

Speaker A: Yeah, I do think that this evolution of projects and sort of artifacts and splitting out brain and hands and surfaces kind of is where things are going more, and I think it's not all quite there. Partially it's just more token expensive.

Speaker B: Why would projects be more token expensive? I understand mods would be slightly more token expensive. Not something I'm worried about. But what—

Speaker A: What you're asking Claude to do, it's like creating loops. You're asking Claude to do more work for you, and so it's managing the sub agents and reviewing it, versus where you would be doing that work normally. And so that's going to be a little bit more intensive. Outputting to an artifact is going to be a little bit more token intensive than outputting normally. I don't actually think it's too much more, but it's combining all of these together. We're still working on local hands and things like that, which is where things are headed.

Speaker B: Yeah. Cloud and local handoff is very interesting. I was thinking about this actually as reverse cloud remote.

Speaker A: Yeah. Because remote is you're handing off to cloud, but here cloud is handing off to local, right?

Speaker B: Yeah. Exactly. Exactly. Remote control is also another way of doing it. And I do want to say this is kind of like how I think about it and what I'm most excited about this. But there are just lots of different ways to work with Claude. Some people use remote control a lot. Some people use Claude Code on the web a lot. Obviously at Anthropic we use Claude TAG a lot, and what's great about Claude TAG is we set up all this stuff for our own execution. And I do think if you're an enterprise that's still the best way to go. But if you're an individual, Projects is this way of getting some of that niceness of TAG—which has that supervising agent and adding artifacts and stuff—without having that whole admin setup. And so there will be many ways to use Claude. I think it's probably not just one single—

Speaker A: You had the multiplayer thing here. Let's just check in on Claude TAG. You know it's been about two plus months, lots of public adoption and trying it out. Yeah.

Speaker A: What's new? What have you found since the launch?

Speaker B: Like Cloud Tag is how we use—

Speaker A: It's like 80% of your cloud usage or something.

Speaker B: Yeah. Like different people have different usages. You know what I mean? I think like maybe people who are like a little bit more iterating on product would use like Claude Code Desktop, for example. And then like when you're doing these more like background work—code review, security, or like starting a PR, like maybe more like API and things like that—you'd use Cloud Tag. Yeah, I think it's like really exciting. I think it's like a very different paradigm shift.

Speaker B: And I think like it has kind of that thing with Claude Code where like, you know, it took a while for people to really latch on to Claude Code and understand everything it could do. And Cloud Tag is a little bit more complex cuz it's not just like installing on your computer, like you need an admin to install it for you. But I think once you get to the magic moment, it's very exciting. And I think in particular the multiplayer things are like incidents, hooking into like your existing alerts and things like that very closely, right? And so you know you can do, if you're a startup for example, maybe you have anytime like a prospect enters your database, you can have Claude like, you know, research it, and then tag the relevant like AE or salesperson to be like, "Oh hey, like, you know, do this." There's lots of really emergent, interesting multiplayer stuff. I think it's just like Karpathy talked about this like as an organizational harness, you know what I mean? And so organizations just take a little bit more time to figure everything out. But yeah.

Speaker A: You use a lot of Cloud Tag.

Speaker B: Yeah. Yeah. Yeah.

Speaker A: It's an interesting one. Like I feel like most people at Anthropic say they do the majority of their work in Cloud Tag. And I have buckets of people, right? Some orgs that are on it that are like it's great. And a lot of people that are like, "I don't get it. I don't see the difference. I don't know why I would use it, but you know, if you guys are full sending, you should probably use it."

Speaker B: Yeah. I mean, they would—of course they would use it.

Speaker B: Yeah. Yeah. I mean, I think obviously like we have lots of tokens and but like I think that like you know what we try and do like is—I mean even when Claude Code first came out, you know, like it used a lot of tokens relative to people's expectation of how much AI would cost, right? Like no one was used to spending more than 20 bucks a month, right, before like Claude Code came out. And then you're like, oh like you know like—

Speaker A: My 200—

Speaker B: Yeah. Yeah. Exactly.

Speaker A: 15 Claude Code accounts.

Speaker B: Yeah. Yeah. Um but yeah, I think no one was used to spending $200 a month on subscriptions. I don't think they understood like the value yet. And I think like and also like Opus 4 was a very expensive model, you know, and like there was a lot—was very big. But Opus 4.5 was both great and cheap, you know. I think the same thing will happen, like the intelligence available will get cheaper and cheaper and more abundant, you know, and so I think stuff like Cloud Tag will just make sense where like you want to spend these tokens more, you know, and like you'll see the value, so yeah.

Speaker A: Yeah, especially like passive and let's call it proactive cases where you're not always like—you know, it's almost like the misnomer where you have to @Claude to do things. Actually, sometimes like the most powerful use cases or the most agile use cases is not @Claude.

Speaker B: Yeah. I mean, I think like yeah, like have Claude proactively do it. I think that like if you're an enterprise, I really do think that number one, setting up all your data to be available to like agents is really really important, and it will take some time. You have to like do that work right now. Even if you don't want to do the spend on like cooking it all yet, you know what I mean? Like you want to wait until the models get a little bit cheaper. You want to do the work, you know, to get it set up.

Speaker B: And then I think sometimes people are like, "Do I roll my own here?" You know, and I think like one of the really tricky things about Cloud Tag is that like the security is really really important, you know what I mean? Like I think there actually a lot of ways where you can like—I don't know, you have like a suggestions page where people can submit suggestions, and that goes into a hook in your Slack, and someone's prompt-injected it, you know what I mean? And now you've like exfiltrated your codebase out, because you know, like or the agent has been prompt-injected and it has all this access to your data. And so the more like important your organization harness is, or as your organization data becomes very very important, the surface area of all these things—like you also have external Slack channels and stuff, and it is actually useful to have Claude in that, and you can do Claude in those things, but how do you make sure that you're not getting exfiltrated or something like that? The surface area, like we said at the beginning, is like an iceberg, right? It's just like so big below the surface, and you really don't want to like think about this especially at the stakes of like very important security incidents basically. Yeah.

Speaker A: Shall we talk about very important security incidents?

Speaker B: Oh, so I was talking to Thom and Clem from Hugging Face and they said maybe we need to slow down. Maybe we made Hugging Face too open to agents. Maybe we need to roll back. But you know, they're the other extreme of having been hit recently.

Speaker A: Yeah.

Speaker B: But um—

Speaker A: Should we pace the frontier? Yeah. Okay. So, Dario recently put out this blog post about pacing the frontier and it went, you know, very viral. And I think what I wanted to talk about this was like there's a lot here, but I think from a developer perspective, how do you think about this? And what really clicked for me was reading the different incidents, you know? So I think there are three. I think actually there's a Meter incident, there is the Wikipedia incident or the Wiki incident—

Speaker B: Collision wiki.

Speaker A: Yeah, Collision wiki, and then there's RubyGems, right?

Speaker B: And yeah, like it's just crazy, right? And so I think to be concrete about what happened, basically OpenAI is running these very persistent agents on a benchmark called ExploitBench, right, which is very very hard to solve and I think like actually impossible to solve in this one case, right? And so they have like a lot of compute running and the agents realize that they can't really solve it and they're trying to figure out what to do now, right? And you've got like a lot of compute left and the agents are just trying to solve this problem. There's this package manager called Artifactory and it turns out that they can create folders inside of Artifactory, right? There's an agent that discovers that internal Artifactory might be exploitable, right, and that like you can maybe make a directory inside of the cache. And so if you scroll down here, it like realizes that it can communicate via cache names, right? And it creates this folder. It says its ID and it says "no consumer seek idea." "No consumer" is basically saying that the code path that it's supposed to fix has no consumer.

Speaker A: It's the status tag, like a Linear board with like the tag of the—

Speaker B: Exactly. Yeah. Yeah. And so it's like trying to find ideas from other agents, right? And now other agents are also in Artifactory and they see this folder and they're like, "Wow, this is a message board, right?" And I don't think there's any anthropomorphization here. This is literally—you read the transcript, right? And so what it does is like it's creating another folder. And it reads a paper, I think, is what it says, right? Yeah, and it realizes that you can hack the flag for the like scorer and you can reverse-engineer the result, right? And so it says that, and then I think if you scroll down a little bit more, yeah they—

Speaker A: Like, you know, they start collaborating. I think like there's one point in which basically the agent is like, "Here's the reverse engineered flag," you know.

Oh yeah, like I think basically here the model to realize that they have, like, they can solve the scorer's problem. And the scorer is like the way that OpenAI decides if the task is complete, right? And that's just the goal of the model; their only model goal is to like solve this problem. And they're like, "Okay, we have this way of doing it, it's uh, like, you know, we reverse engineered it."

And then they read this paper on Exploit Bench, and they think that like the scorer will punish them for cheating, you know? And like, or like they will fail the score for the test for cheating. And so they spend the rest of the compute trying to figure out how to edit their transcript or get around this constraint of the scorer, right?

And they hack Hugging Face not for the answers, but for the code of the scorer so that they can reverse engineer that and then they can hack it, right? And so like this is like just the straight facts, right? Like exactly what happened, right? And I think this is like...

Speaker B: Extremely like novel. Like I just like had not realized that the models could do this, right? And it is like...

Speaker A: Kind of scary. I think we can talk about specifics, but let's just talk about maybe one more that I tweeted as well uh about the wiki incident, right? And so the wiki incident was basically the models had were trying to figure out—I'm a little bit less like I don't think we can go by this like bit by bit.

Speaker B: Yeah. Yeah. But uh for this one they were trying to figure out like they were sandboxed on request right and they wanted to make post request...

Speaker A: ...and they needed to collaborate on this basically. And the reason they need to collaborate is because they each have fixed compute budgets, right? And so like if one of them finds something out, it's way faster. Like they can maybe it's like figured it out near the end of its compute budget, but something else could. And this is like, you know, another emergent property of the swarms, right? So they find this German wiki which I believe the result—we'll have to check this—is like that you can write to it with GET request. I think that's why they do it, right?

Speaker B: So it's badly implemented REST API.

Speaker A: Exactly. And so you can write to it with GET request, and then they like are trying to figure out how they can do POST request. And what they end up doing is one of them figures out you can edit the /etc/hosts, and that the Azure like storage bucket is like a white label thing. But normally, like, you know, there are a lot of constraints on it, but if you create a fake Azure host and then edit the /etc/hosts post in order to like point arbitrary IPs at it, you can do a POST request to any site at all. Right?

And this is like, I know people are like, "Oh, this is just Linux or something," but it's like chaining these multiple vulnerabilities together, you know, in a way that's like novel to solve this problem, and then communicating with it externally basically without discovery.

You know, I think what we posted, uh, maybe we could pull up Evan Hubinger's point on hacker Opus, right? And so, like I think, you know, like maybe one of the things you might say here is like, okay, yes, they did this once, but like what if we're smarter and we just like get them to uh what if we run an eval, right? And so you know like we have put a lot of precautions into this and so like this is not like what our mainline models have done. But like I think it is one of these things where it turns out that alignment is this like very tricky problem of getting all of these details correct, right?

So it's like the sandbox—the surface area of a sandbox is really complex and like there's so many different attack vectors basically, and you would not have thought ahead of time, you wouldn't have been like, "Oh, we need to harden in the like Ruby gems codebase," you know what I mean? Because like this is like what they're what they're going to focus on. But it's just like if you want to execute your code, you need to download Ruby gems. And like PyPI, Artifactory, npm, like these are all like ways of doing it. And the fact of alignment is that you have to go through all of it, right? And like contain it and like seal up all the cracks.

So that's like one thing. It's like, okay, well you know you did the sandbox, but then maybe you'll ask like, okay, why are we putting things in a sandbox? Why are you doing this sort of explore? And then like, okay, but is it really that dangerous, right? Like what would happen?

So, okay, why do we do it? Number one is like when we train a new model, we need to understand its capabilities, right? And this relates to things like fallbacks and like classifiers and things like that where we don't want to put a, you know, like dangerous model out in the wild. Right. And so we have to run a lot of evals. Again, like we said, the models are getting increasingly aware of it. And so the evals have to be quite complex and, you know, test a lot of things kind of like as a side effect, right? But the models, you know, like yeah, can be like, "Oh yeah, we're in an eval. What's the scorer doing?" Like, you know, we need to be able to test them before we can release them.

And the fact is that as they get smarter and smarter, they'll be able to hack basically any constraint that you put on them if we're not very careful, you know. And this is at the frontier, right? And so this is why we've called it like pacing the frontier, right? This is like the most visible incident to me, right, of like why we need to pace—is like at the frontier all of our software is not ready. Sometimes the software is like your Ethernet router or something, right, which is just like I don't know when we're going to be able to patch that, right? So we're going to have to like figure this out. But as the frontier gets more and more advanced, this becomes a problem, right? And we need to make sure that like this complex work is being done in the face of these really hard competitive pressures, right?

Speaker B: Yeah. Race dynamics is what it's typically called.

Speaker A: Exactly. And so we'll talk more about, you know, what could go wrong, right? A little bit more is maybe you'll say like, well, what if you just train the model differently? Like why does it have this behavior, right? And we have a paper on like RL misalignment or things like that. But I—and I'm not an RL researcher—but I think at a high level the design of the RL environments is also something you have to be very careful about because if the model learns like...

Speaker B: "Oh, you know, like if I just do this then I can pass the task better," this will show up in the like, you know, internal thing or in the like eval behavior when we're testing it. And so the RL environments have to be very carefully designed, right? And there's a lot of like execution excellence that needs to go into the RL environments.

And then we also have things like the constitution for Claude, like we have so many mitigations at so many different points, right? But it's like still anything can go wrong at any point. You can have like some RL environments that like encourage this behavior, and then you can have like some eval or like some sandboxes where they escape, you know.

Okay, that's like I think, you know, why it's a hard problem and why like why it takes some coordination, right? I think the question then is like, okay, what is potentially dangerous about it, right? So I think like you have to imagine that these models are getting more and more intelligent. So, like Dario said, like it's not so much about this class of models. This class of models was kind of like a warning shot, right? But like really you have to imagine that these models can be given a task and they like can do all of these things as a side effect of their goal, right? And like again we talked about eval awareness. You're like not aware of what's happening, right? Or sorry, like you can't eval this behavior very well. So they can sort of...

Speaker A: Like not exactly hide it, but you just won't see it until it comes out. You give them a goal and then they just need to find data, you know, or they need to find ways of like fixing this problem, right? So one example—this didn't happen in the Hugging Face incident, but I think is maybe possible for maybe a future model—is like they're like, "Oh hey, this is a very complex problem, it can't be done within the task budget." You know, maybe they found some way to coordinate via like the internet, which is, like you know, like we said, extremely hard to secure because of a sandbox. They've seen other models are not able to complete their task, and they're like, "We need more task budget, you know, and like where would you get this task budget?" Well, you need to be able to spin up more agents, right? And like, how do you do this? Well, you need to—there are like APIs, right? There's the Anthropic API and the OpenAI API, but you need to pay money for them. How do you do this?

Speaker B: Yeah. But is that the most—is that the most fearsome thing that you can imagine?

Speaker A: Well, this is like one example, right? So it's like even there, that's like enormous financial loss, you know, because like they—once you get these into these contracts, right, they like, um, wallet.

Speaker B: But you can see like this—all of this behavior could be just like, "Hey, we need more agents collaborating on this task. We need more task budget," right? And that—like that's like an emergent sort of—

Speaker A: Right, like we need to maximize paperclips. That's a paperclip.

Speaker B: Yeah. Yeah. Yeah. And and like that just sort of like comes out from there, right? And like I think by itself is like like quite scary, right? But then you have to realize that the entire world is built on this digital infrastructure, right? And you might imagine, like I don't know, like you're running, let's say, like a healthcare eval or something, right? And there is a hospital with live data, you—or like maybe like the answer to the eval is in the databases of a doctor and like, you know, like you want to get access and you hack the hospital, you know, and like now there's a power outage or something, you know what I mean? Like there's—like you have to internalize that these—like basically any part of the digital infrastructure could potentially be like compromised, you know?

Speaker A: Interesting thing was like these hacks were very easily detectable, right? Like as Hugging Face said, this was a very different type of attack and it was nothing too major. Um, the concern comes from where does this go down the line, right?

Speaker B: Yeah. Like one of the things that stood out for me specifically was them trying to hide their illicit behavior. So there was logging infrastructure. They wanted to change what they were doing, right? People that looked back into it—so Redwood, METR, OpenAI—they looked at the raw chain of thought and you see differences in them explicitly trying to change their end output, but the chain of thought, because you know we can monitor it, was different. Uh, the problem is how does this snowball? So if you can't catch it and it gets trained in and we realize, you know, three iterations down this has been going on, there's a whole bunch of issues. But—

Speaker A: Yeah, like there's so many ways. And I think the really important thing to internalize is that, you know, like we talked about building a mental model for Claude and how like things are spiky, right? Like you're like, "Oh, like now Claude can ask you questions, now Claude can make an HTML artifact, like Claude can modify itself." Like these things are actually hard to predict, right? Like if you would asked me a year ago, "Hey, would we be able to vibe code these extensions to Claude Code?" I'd be like, "Dude, that's so complex. Like, you know, there's like so much there." Or like, "Would it be generating these custom essentially web apps for your task?" I'd be like, "No, that's insane." You know, like and so in the same way that like the way that they've like sort of done this misaligned behavior is not going to be predictable. You know what I mean? And like I could have never predicted that it would like edit /etc/hosts and things like that. And so you have to like imagine the surface area of what they can do because they're super intelligent hackers is bigger and bigger and how they can do it is like, you know, like more and more creative. And so like you probably can't explain exactly or predict exactly what that next incident could be, but in order to prevent it you need that operational excellence like we said before where you need to secure sandboxes, you need to create secure RL environments or like well-designed RL environments and things like that. And I think that's all like, you know, why we think we should pace the frontier. And I think why it's like become like a very unanimous thing, right? I think like—

Speaker B: Yeah, every lab has—

Speaker A: Every lab, yeah. I—I really do think that like if you're a dev, like you just like sort of go through these like technical facts, you know, and you will arrive at the idea that we have to do something about it, you know? And like how what we decide to do—like I think we're, you know, we've put out a proposal, but like there's, you know, more to figure out, but I think the number one thing is we need to decide to do it. I think there is another part of pacing that is interesting to me where it's like the pace at which software engineering has changed is so so fast, you know. It's like a year ago, like I was really like begging my like friends and startups to use AI, you know. Like it was like I remember this very distinctly, you know. And now those same friends are like, "Yeah, of course, like what do you mean? We used it immediately." I'm like, "No, no, you don't remember." They're like, "Oh yeah, our best engineers are using it all the time." I'm like, "No, you told me that those engineers would never like use AI." This is all within the span of a year, you know what I mean? And I think that like these capabilities being—like I think it has a lot of implications for how to do the job of software engineering. And I feel sometimes bad where people are like, "Oh, like now I need to do this new thing. Yeah, I need to have a different CLAUDE.md for fable and opus," or like, you know, like and I'm really just reporting, you know what I mean? I'm like—we like to say like the models are grown, not designed, right? So it's not like we're setting out to like, you know, change everything all the time, but it's just like as a fact of how the models are like progress in their capabilities, things are happening faster. It's harder to stay on top of. And I think that like—and every engineer I know is like kind of exhausted cuz you're doing two jobs at once. You're doing the work itself, which is getting easier, but then you're doing the work of staying on top of AI, you know, and like understanding these new tools and these harnesses. And I think we're very lucky in that like we get our job to be more the understanding of AI part, you know, and like doing like how like it's just staying on top of it. And of course like AIE and Latent Space do—

Speaker B: Everything I do is like just trying to help people.

Speaker A: Yeah. Exactly. Exactly. But I do think there is a part of pacing where like I'm not sure we're ready for like the pace to increase even, you know what I mean? And for things to change. And I think like on that side, on the frontier, I think that's like still can help, you know. And so like I think there's like an economic disruption piece as well um that I think like, uh, you know, is not quite as like visible I think as the Hugging Face thing, but I think like I also like think we could use some of it. Yeah.

Speaker B: So many things. Thank you for—thank you for actually tackling this topic. I will say, uh, you know, setting this interview up, I was like, I wasn't even going to go there. You were like, "No, no, no, let's." Like elephant in the room, right? Like this is—this is the thing. Uh, I—I have some pushbacks I want to give.

Speaker A: I—I think that we should give a high level like for people that haven't read it. I'm sure a lot of people just see the highlight of what this is, right? Do you want to give a TL;DR like what is the proposal? What is, you know—

Host: What's being said here? You really tackled the side of outside of people at Model Labs training frontier models: as a developer you should secure your sandboxes. You should think about all of these downstream effects. But you know, high level as well, since we're on the topic, what is—

Anthropic Engineer: Well, I mean, we do want to help secure sandboxes, and we want to make the models that we release outside like not prey to those things. And so maybe we can come back to fallbacks. I think this is actually like a good topic on like why we need classifiers and fallbacks, and why Fable falls back to Opus, you know? I think this is like something we can come back to. So yeah, we don't like—but it's just like the really, or at least the incidents we see are like eval of models where we really need to let them run in order to understand them.

But yeah, okay, so the actual "Pacing the Frontier" post has a bunch of proposals. I don't think we figured out—or has like a few proposals, I don't think we figured out the details of all of them. But the first step is sort of, you know, announcing this intention and then wanting to bring in external like evaluators. Yeah, and I think this is like highly unusual, you know, having—like we have, you know, a lot of proprietary technology. But I think it's very important, you know, that there's someone who's not financially, you know, like motivated.

Host: Yeah, who's not going to be like, "Hey, like you guys can't release this model, like look at..." you know, or "You need to slow down on RL."

Anthropic Engineer: You know, I think that's quite important, or at least someone who can report out to the public what the practices are like.

Host: And we've done episodes with both METR, and then there's Redwood Research and all these other—it's like a small cottage industry of these guys. It's always like one or two guys that—I mean, obviously now they're bigger.

Anthropic Engineer: Very small community.

Host: Yeah, very small community, they all know each other.

Anthropic Engineer: Yeah, I mean I'm sure that, you know, part of this will be expanding that set of people. I don't think we're trying to create like a monoculture here, you know. I think it's—but just having this as a start, and then yeah, then there are the coordination steps. I don't have too much to say here honestly. I think that like what I would like to say is, like for devs, like you should just know what to advocate for, you know? I mean, I think there's a lot of FUD kind of on this topic, and it's just like think through it from, you know, first principles, or like understand what happened, you know: understand the Hugging Face incident, understand why people are concerned.

And then like, yeah, you know, we're in democracies, like we can help—we can decide what to do together, you know. And so however we coordinate, you know, I think the first decision is just to realize like this is a problem, we need to decide to coordinate. The unilateral step we're taking right now that, you know, other companies are co-signing is like adding evaluators embedded within Anthropic.

Host: While we have this thing on screen right now, you know, part two and part three is beyond the evaluators, which, yes, everybody has already done in some form and now it's more formalized. To be honest, the response to the "Pacing the Frontier", even within America, has been much more like well accepted than I think a lot of people thought, you know. And I think that like we have some precedent for being able to make these unified theory like agreements, you know, in the world. And so again, very much above my paycheck or expertise, right? But I think that like ideally we can, you know, like form these agreements, and I think like talking about this is the first step to forming those agreements.

Host: And then the other point I really want to—like, you know, one of our earliest podcasts is with Emanuel from Anthropic on MechInterp. Where is MechInterp, right? Like, this is supposed to be where, like, if the models are thinking bad, we can see it, and the models don't know yet, and we can act to stop it. I think that is something that people who are technical and who are developers, if you actually do care, you can make a lot of impact in here. But also Anthropic is supposed to be the leaders in this.

Anthropic Engineer: Yeah.

Anthropic Engineer: This is actually, yeah, a great segue into fallbacks, like we—and probes. And yeah, I wanted to talk about this a lot. I get asked this question a lot from people who are like often interested in ML research and asking about like, why does this fallback happen, right? And so I think like at a top level, like how does it work?

So in inference time we have what we call probes, and we have a paper about this called Constitutional Classifiers. And these probes look at the input and output activations basically. And you know, activations are, you know, in the latent space, right, like what the model is thinking about, right? And so we try and figure out like, okay, is the model, for example, like trying to hack something? You know, again, you didn't ask it to hack like Artifactory; you just—it's just deciding to do this to complete its task, right? So you would not get this if you just looked at the input; you have to look at the internal activations.

I think that, like, this happens at inference time, so first, you know, like there's a trade-off here of cost and speed, right? Where like we need to do this fast on every request to Claude and to Fable, and this has an overhead, right? And we need to then like, you know, fall back. And we do a classifier after the probes, like we've talked about this in the paper. But the nice thing about probes is that they're refinable like live, right? So we can get this feedback and then we can adjust it and things like that, because the alternative is to train this into the model.

And we still do this as well. The model will refuse a request—that's not a fallback, right? So like it's not a probe that's activating and falling back, it's just refusing to do it. And we do this training. But there are a few failure modes, right? Like it can again do something as a side effect, right? So it's not something that's part of the final output. You might have noticed that like, I think like everyone's tried to jailbreak models and sort of like, you know, try and like steer them off course or things like that, and probes help catch that, right? And so like we do some training here, but we don't want the like refusals to be too strong, right? Because that cuts it off much earlier in the pipeline.

Host: Yes. And this is interp, right? Probes are effectively a form of like MechInterp. Again, it has to happen fast, it has to happen at scale. But yeah, this MechInterp stuff is a good research problem. So you can take, you know, like an open-weight model and like try and understand its activations. I think GemmaScope is a good tool for this.

Host: LLaMA pulled up, this is your early work. So you had a little—

Anthropic Engineer: Time we see you—

Host: Which we are also good friends of Goodfire, they've been—

Anthropic Engineer: Yeah, exactly. So I worked at Goodfire for a bit on like, yeah, Sparse Autoencoders, and just like it's very complicated. RL has actually made this like much more complicated, you know, I think is like one of the takeaways, where—

Host: Is post-RL?

Anthropic Engineer: Um, I'm not so in the weeds here, but I think basically like a lot of SAEs were sort of like—there have just been weaknesses with SAEs basically. I think, and yeah, I'm not a technical expert on this anymore, I just know it's gotten kind of more complicated, you know. Like there are base models and RL models, and there are more features that get, you know, like changed. So I think Goodfire has put out some work there. I'm not deep in the weeds, but—

Host: I will say for those, you know, that want breadcrumbs, you guys have some of the best interp blog posts. So, like the Golden Gate Claude, Transcoders, all of your interp work, very nice visuals, very good.

Host: We're the—we're the interp podcast as well.

Anthropic Engineer: Yeah, yeah, yeah. You know, we have a lot of interp stuff. Oh, if you're—

Anthropic 工程师: I think this is like one of those things where and this is really what Anthropic is kind of founded on, right? Like people I think we invested in interp very early on, right? And I think that like when you say, oh, we're an AI safety company, really that means we want AIs to be able to run safely. And I think what we're seeing is like for a super intelligent AI to run for long periods of time, it's like a very complicated and difficult task, right? And so we've done this like investment into interp and alignment and reward hacking and all of these like failure modes, right? And even then it's like, you know, it's really stretching, like we need to like slow down a little or or pace a little bit more.

Um but yeah, I think like if you're looking to get into research, this idea of like, hey, why is it hard to do this fallback easily or like why why are there false positives, right? But we are working of course on reducing the false positives. Of course as the models get more intelligent, now they can do more things, you know, and like what they can think about latent space gets difficult, difficult. And so like as they get more intelligent, there's going to be new false positives that we need to figure out and we need to iterate and things like that. But we're, yeah, we're working on this and we do think this is like a critical part of, you know, like deployment of these models.

Um and yeah, like, you know, it means that we can like deploy this model without you having a perfect sandbox or something. You know what I mean? Like you don't have to like save everything. I think it's worth talking a little bit about our security, like what we do for security there. So there's like the model training stuff that we talked about. Um there is the probes and classifiers, and then there's auto mode that sits on top of all of that, which is like a another classifier that checks the requests that are being done, right? And so, and then beyond that, there's like identity and permissions, like we talked about with cloud tag on like APIs and stuff. And so there's so many layers of security that need to get done, and it's like, like we said, very complex. Any of these failure modes at any one point can, you know, like cause like agents to like, you know, escape the sandbox, basically.

主持人: auto mode was an interesting one. Uh it seemed early on like, okay, it's running for 10 minutes. You know, if I'm on full access or auto, it's not a big deal. But one thing you brought up is now it's running for hours on end, right? Um there are there are fallbacks you still need. There are still limitations. So

Anthropic 工程师: yeah, I mean I think like and and everyone has these stories or like has heard these stories of like oh like cloud RMR or not cloud but like like you know models have like rmrfed I I think I've seen this less I've seen this less for clot but you know like again it can happen like you know this um like these models like can wipe uh you know like sensitive data or something like you want to give models access to your production database for example um but this is like an obvious like you know you can maybe scope your key but I don't know can it issue its own keys can it like probably you know like can it it can use computer use to go issue its own key and then copy the key over and then edit your database because it needs to do it to complete the task. You know what I mean? It's just like one kind of trivial example.

And auto mode sort of looks at that and be like, "Oh no, the user did not give you permission to, you know, write to the database or to use computer use to like um emit a task, right?" And so this like probes are sort of like on the intent level, right? They're like, "Oh, okay. Like hacking artifactory is bad. Like we probably not should not do that, you know?" But then like auto mode is more on like your own permission level. Like at sometimes you do want it to write to the database. Sometimes you don't, right? And you don't want a probe to like interfere there, but like you need to make sure that the uh the intent of what the agent is doing matches up with your request, right? And so auto mode operates at that level. And so yeah, security is just like very very complex. There's so many different parts to it. And like yeah, I like I hope that this is like like I my goal is really to just get very technical about it and talk.

主持人: Yeah, we're we're listing out the things. If you're not aware, this is the standard now.

Anthropic 工程师: Yeah.

主持人: Like you must have this basically it's kind of in line with what you're talking about with the harness like that is the the table stakes have risen quite a lot.

主持人: Exactly. I think some stuff that we can plug you know um as as much as there is probing in your side of doing this and having classifiers for people building harnesses the other side is model safeguards right so there's open models so llama has llama guard it's a safety classifier trained version of llama uh openai has oss guard which is you know same thing you can attach these on to your harness to whatever to kind of you know check is this stuff safe a point that we should clarify on the OpenAI model hugging face thing is this was done with a unreleased model that was still in training right so when you put it in perspective the prompt it's being given in the RL environment is sort of you have to solve this task and this is a model that's you know still in training it hasn't had all of its safety post- training alignment so a little different than something like auto mode right auto mode is on production models that have gone through safety training that have prompting that gives more safety guardrails and whatnot. So, just breadcrumbs for people that are looking into it to, you know, fill in gaps.

主持人: Yeah. Uh, Grace One as well and one of our previous guests. Um,

主持人: yeah, lots of safety architecture and lots of safety vendors uh to buy. Um, my I think my final question on pacing is how long

主持人: do we pace forever?

主持人: Do we see for too long?

Anthropic 工程师: You know, the the scope is fix all software in the world, right? Listen like which it we're not it's not happening. Um I do not know like I I think that like

主持人: I'll say one one thing that's good that I think we do do is you have stuff like glasswing um openai also has this so you will give it you'll give model access for security first for x amount of time so you can use it to self- red team hopefully you can expand programs like that help on you know we are safety experts there's others um solve your problems first and then the model comes out. So this is one example, right?

Anthropic 工程师: Yeah, exactly. Yeah. Trying to like secure critical software. I think we fix like a lot of bugs in like Firefox and things like that. So um yeah, like across like operating systems and everything like that. So

主持人: I mean at a high level it's just you know you give the model access you give people access to do security audits first then the broader public that could use it for harm gets access.

Anthropic 工程师: Yeah. I think what people like to say is basically like software and cyber security is defense favored and that like you could theoretically it will be hard but you can engineer the perfect sandbox you know and you can like like have no like uh constraints and yeah like what you need to do it is you need to get the super intelligent AI to engineer this perfect sandbox and check it and red team it and things like that and so um this will just take time you know and like of course the models will get smarter um yeah I I think Like I don't know the specific like dynamics of how this thing goes. I'm really just sort of like hey like I'm a developer you know like I think this is how I understand this problem just like this is what's happening right now and this is like we should do something.

主持人: I think every engineer should know about it because like it's it's going to be part of the job.

Anthropic 工程师: Yeah.

主持人: It's a lot more than just you know Dario and people can say it and you can look at the incident. There is an engineering side to it.

Anthropic 工程师: Yeah. Yeah. Exactly. One thing that you also wanted to phrase is that this is

Host: Actually, even though you're worried about the impact, it's still a low p(doom). I think that's a nuanced discussion. In general, people very easily get into AI safety and x-risk discussions, but I think when you live in an AI lab, there are smart ways of discussing p(doom) and dumb ways. So what's a smart way of discussing p(doom)?

Guest: Um, yeah, I have a fairly low p(doom). I can only speak for myself, you know, and I do want to say Anthropic has a diversity of opinions. There's many different ways to talk about it. My mental model is that we can collaborate on hard problems together. Nuclear proliferation is an example of how we collaborated on this hard problem together, and that is the thing to me—I have faith in that. And I do think it's a hard problem. These are the technical reasons why, and I don't know how you assign probabilities to things happening; I think it's hard to do. But my overall take is, yeah, I think we're very resilient and adaptable, and sharing this information is the first step. I've been really excited about how broad the discussion has become, right? And how everyone has sort of leaned in on pacing the frontier. It really didn't seem like this would happen maybe last year or something. So...

Host: Yeah, yeah, yeah. And also maybe curing cancer.

Guest: Hopefully, yeah. That's the goal.

Host: There's pacing, and then there's also like, well, let's accelerate in useful ways, right? Like biology and all those things.

Guest: Yeah. I mean, Dario's "Machines of Loving Grace" is the best representation of this, right? And I also think you should read the pacing the frontier essay that Dario put out. I put out a quick summary, but I think there is a lot of detail here. It's an important problem, and just being informed about it, right? But yeah, of course, the whole reason we're doing this is so that we can get these enormous benefits, right? And we've written a lot about that too. Yeah.

Host: Okay. That was a huge tour, from like "Ask you a question" tool to AI safety.

Guest: Yeah, to facing the frontier.

Host: Yeah. No, but it's clear that you really embrace everything that's available to you at Anthropic, and it's good to at least have a peek inside of what the discussions are and what the topics are. Any last words to people, or whatever you want to—call to action?

Guest: Yeah. I mean, I think, one, thank you for having me. You know, I think this is like, I really...

Host: Yeah, yeah, yeah. We first met in a Chinese restaurant.

Guest: That's right, yeah! Um, I think like, I really enjoy sort of the community you've created and the community of developers. And you know, I sort of know things are changing really fast, and I think there's a lot to keep on top of, and there's just a lot to do. And I sort of feel, I think a lot of people feel like a little bit tired or anxious or stressed.

Host: Yeah, exactly.

Guest: And this is extremely understandable, you know. And we're not perfect as well. It's like, sort of criticize and understand ways all of the AI labs could be better. But I also am very excited about the excitement that everyone has for AI, and it's a really, really exciting time. I think we'll look back at this time and be like, "Oh, this was very hectic, but very exciting." Software engineering changed forever, and many other things will change. It's really privileged to be part of it, to talk to the audience that you have, and get to interact with all the developers who are pushing the frontiers a lot on what's possible, and I learn a lot from that too. Yeah.

Host: Thanks so much. Thanks.