Anthropic’s Felix Rieseberg on AI Coworkers, Local-First Agents, and the Future of Knowledge Work

This is maybe a whole, like, a somewhat contrarian view to a lot of people in AI. I actually don't think that the future is going to be hyper-personalized software down to the point where everyone is running their own version.

Like, I actually think it's going to be quite hard for all of us to have our own internal chat tool. Silicon Valley overall is undervaluing the local computer, and my default argument for that is always, how come we're all using MacBooks and not, like, an iPad or a Chromebook?

And now when I think about Claude as this entity that is supposed to be very useful to you, like, tremendously useful to you, I think that entity needs to have access to all the same tools you have access to.

Otherwise, it's gonna be hamstrung in, like, all these complex ways.

Hey, everyone. Welcome to the Late in Space podcast, our first one in the new studio at Kernel. Uh, this is Alessio, founder of Kernel Labs, and I'm joined by Swiqs, editor of Late in Space.

Yeah, so nice to be here. Thanks to, uh, TJ, Alessio, Alan, helping to set everything up. It looks beautiful. We even have the logo outside.

Yeah.

It's kind of nice.

It's, like, really nice, right? When you walk in here as a guest, you're like, "Ah, this is a serious production." You, like, feel it immediately.

Yeah. Felix, you've been... You're, you're currently product manager of Cowork or-

Uh, engineer lead, technically.

Eng lead, yeah. Like, the, the identities are kind of vague. Member technical staff.

I know. Member technical staff is, like, the official title we all carry around forever.

Uh, yeah. I recently kind of wanted... Like, w-we've been, uh, kind of obsessed. I, I've been using it a lot, even for managing Late in Space. Like, uh, Cowork helps me upload videos and, like, title things and, like, edit and everything.

It's, it's, like, really amazing.

Cool. He said multiple times Cowork is AGI in the group chat.

Yeah, yeah, yeah. So, so we have a second- Uh, we have a second channel, uh, for Late in Space TV. Uh, and I, uh, and, uh, we-- Basically, this is our Discord meetup. Um, and I, I... We have, like, Claude Cowork is Might be AGI.

I don't know if we, we have, uh, uploaded it yet, but one of the sessions was, like a, like a Claude Cowork thing.

I would love to see, I would love to see it. Like, I'm so curious. Like, one of the most fun parts of my job is to, like, constantly see the weird things people use Cowork for because it's obviously, like, very hard for us to actually design for specific use cases.

We do, but, like, every single person who's, like, most amazed is usually amazed about a thing that I didn't even expect Cowork would be good at. Um, we have a new designer, and it's one of the first small tasks.

I was like, "Hey, we need, like, a new emoji for Cowork for our internal Slack." It's, like, a pretty small thing. I was like, "Can you please do it?" And he drew an SVG and just gave it to Cowork and was like, "Can you animate this emoji?"

And now it has, like, this beautiful loopy animation. Um, and I mean, I think obviously this goes down to, like, it turns out you can do more things with code than you expected. But it, it's, like, that kind of stuff that is really fun to me.

So long story short, I would love to see, like, the kind of things you're doing.

I'll pull it up. I'll pull it up. Yeah.

Yeah.

Uh, but before we get into it, I, I think always want to start with, like, a top level what is Claude Cowork for people who haven't heard of it, haven't tried it out.

Okay. Uh, real quick, Claude Cowork is a user-friendly version of Claude Code. So the way it basically works is we have Claude Code and, for us, fairly impressive agent harness that over December we noticed more and more people are using either even though they're not technical, they, they're not at home in the terminal; or they are at home in the terminal, but they started using Claude Code for non-coding workloads, right?

Like managing expenses or, like, filling out receipts or organizing a knowledge base. Like, there was a big Obsidian moment that a lot of people liked. And we wanted to capitalize on that, but also bring, bring this capability to people who are not terminal native and who might not know how to, like, brew install something.

So Cowork is Claude Code running in a virtual machine with a little bit of padding, a little bit more guardrails, making it a little safer and a little more convenient for people who don't wanna first open up the terminal when they go to work.

It's interesting, uh, that it's kind of pitched that way as a more user-friendly thing because I always feel like it, it... To me, I, I treat it as, like, well, I, I'm familiar with Claude Code. Like, we, we did a Claude Code episode-

Yeah

... a year ago. But this one is, like, even more power user tools because it, uh, it kind of integrates much better with, like, Claude and Chrome and, uh, you know, all the, all the other tooling. But, like, may-maybe that's, like, a perception thing, right?

Like-

No, honestly, I don't think you're wrong. This is, like, a, a thing I've been thinking a lot about for, like, the last weeks. So-

People when they say user-friendly, it's like, oh, it's the dumbed-down version. But no, actually, this is the superset.

Yeah. Like, I think a similar thing happened, a similar thing happened to me about ten years ago, like maybe 12 years ago when I was at Microsoft, and we started working on, on Electron and, like, browser-based technologies and cross-platform stuff, and one of the first use cases was Visual Studio Code, which used to be a website.

And the initial narrative was, well, Visual Studio Code is, is like a more user-friendly version of Visual Studio.

Mm-hmm.

But in a similar vein, I think there were some voices saying, "Oh, this is not for serious developers." Like, "We're not gonna use this," right? For, like, anything. And I think in the end what happened is people have different stories about why Visual Studio Code became such a big thing.

But my personal, my personal belief is that the hackability and the extendability has, like, played a pretty big role, right? You can hook in Visual Studio Code to, like, almost any workload. It's so easy to hack on, so easy to build extensions for it.

And I think Cowork might be hitting a similar thing where it's very easy to extend, and it's very easy to bring into your workflows. Uh, so the convenience, I think, is a bit of a... I-it's obviously the thing we strive for as developers, but I think the way people find value in it then is by probably mapping it onto whatever they actually have to do in their job.

So end of last year, you see the spike of, like, non-technical usage in Claude Code. What's the design process to say we should make Claude Cowork? Because, I mean, you built it in only ten days. Um, I'm sure there was some discussion before on what does easier to use mean, you know?

Yeah.

Like, making, making, like, a desktop GUI is obviously one way to do it, but, like, there's a lot of nuance in the product. Like, maybe talk people through what was, like, the trigger of, like, we should build a separate thing.

Yeah.

We should not build, like, a different Claude Code thing, and then maybe some of the more interesting design decisions that maybe you didn't take.

Yeah. I think- At Anthropic, we've been thinking about ways to move people who are comfortable with using Claude to answer questions and bring more of the power of, like, this thing to now, like, execute tasks for you, right?

It can, like, solve problems for you, it can, like, build things for you. How do we bring that capability to people who are currently mostly comfortable with, like, a, like, question-answer paradigm within the chat? And we've had a lot of prototypes around that.

This is going back as far as, like, easily a year and a half. Like, we had a lot of people working on that. Um, and internally, Anthropic is a very prototype, demo-first culture. We have a lot of, like, internal prototypes that don't reach the public, and what Cowork actually became is, like, we sort of picked the right pieces out of the many prototypes that we had, right?

And that's, that's maybe also, like, I think an important qualifier. Whenever people mention this, like, 10-day number, I do think it's important to me to mention that we didn't start from scratch. There was, like, a lot of stuff already happening, right?

Like, and I think it's important for people to remember that when you build a website, you use React, you use, like, a bunch of other things, and this is, like, a similar scenario with, like, a lot of pieces we already had.

Um, and in terms of decision path, I think we live in, like, an interesting new world where execution is actually quite cheap.

Mm-hmm.

So maybe, maybe what you would do-

That's so crazy to hear.

I know, it's wild, right?

Usually ideas are cheap, execution is the hard part.

I know. But, like, the... We, we used to live in this world maybe where you would take a product manager, and the product manager would go to a number of potential customers in this, like, very low bandwidth way, would try to, try to, like, tease out what are the problems they're having, what are they willing to buy, um, and then maybe what can you build to, like, address that need.

And then you go back, and you, like, draft a spec, and you think about it, and then, like, you make a design, and you execute it. We internally at Anthropic are now probably much closer to the point where we're like, "Don't even write a memo, just, like, build...

Like, let's build all the candidates very quickly." Like, let's, let's just build all of them-

Mm-hmm

... and then pick the best ones. I think the, the decision that is most impactful both for the product as well for the users right now is, like, the way we put value on your local computer. I think that's a big decision point.

A lot of people have thought about should this thing, whatever it is, should it ultimately run on your computer or should it run in the cloud? Because they're big trade-offs, right?

I guess, like, if we solved auth, it would be easier to do in the cloud, but I think, like, the fact that I can just download any file from anywhere-

Yeah

... and then put it in Cowork there is, like, a big unlock. Um, I mean, it's interesting you mention reusing certain pieces. I think this is something I've been thinking about even with Claude Code, right? The price of, like, writing code is going to zero, blah, blah, blah, but it actually seems like the value of having some sort of platform substrate is, like, increasing because as you build these new things, you can kind of plug them together.

Yeah.

So I almost feel like when people are saying, "Oh, the value of a lot of software is going to zero because you can recreate it," to me, it's almost like the opposite. It's like having an existing platform to build on top of is, like, even more valuable because you can kind of bolt things on.

Yeah.

You have obviously MCPs, you have skills, you have, like, obviously the models, uh, which is a big part. All these things kind of come together. Do you feel like that's a valid way to think about it, where people should invest even more in kind of, like, these primitives to rebuild on?

Or are you, like, recreating a lot of it each time because, like, things change and it's easier to rewrite than reuse?

You know, I think, I think you're right. I think you're right that the holistic platform is really useful, and this is maybe a whole, like, a somewhat contrarian view to a lot of people in AI. I actually don't think that the future is going to be hyper-personalized software down to the point where everyone is running their own version.

Like, I actually think it's going to be quite hard for all of us to have our own internal chat tool, and, like, if I want to talk to you, like-

Right

... how is that gonna work, right? In the, in the context of Cowork and how we build it, I think it's a bit of a combination. Like, what the, the execution that gets cheap is not necessarily rebuilding all the primitives.

I think a priori there's also not a lot of value in it. So for instance, my team did not think about rebuilding Claude Code. We, like, very much started with the, with the core thesis of this should be Claude Code.

Mm-hmm.

And then we'll, like, build things on top of it. The part of the execution that gets a little cheaper is, like, how do you take all of these Lego pieces and put them together in a way that makes sense for users and is, like, actually valuable?

You have so many different approaches now in terms of w-what kind of, what kind of things do you actually elevate to a primitive? Do you strongly believe that all your products should be built by just combining primitives that the public also has available?

Do you keep some things internal? Um, and I think that's still evolving, but I think what's probably gonna go away is, like, I'm not sure if it's gonna fully go away, but I'm gonna say I think for me personally, I will probably no longer try to come up with a really good product without testing it out with people.

This is not a new concept, but wherever you used to have to make costly decisions around do we pick technology A or technology B, or do we, like, um, build it this way or build it the other way, I really strongly believe now you just build all of them and try them out- ...

with a small focus group, and then whatever, whatever is better is what you go with, right? And that, that is probably quite different even from how we maybe worked a year ago, right? Like, I think, I think this happened very recently.

Yeah. I started building something in, on Electron, since you're here. Coincidence. Uh, but then Electron and, like, SQLite are, like... There's, like, some issues that, like, uh, between development and, like, uh, building. Anyway, and I was like, "Let's just rebuild the whole thing in Swift," and just recreated the whole thing in Swift, and it's like, it's done, you know?

That didn't take any effort. I, I, I don't even know Swift, but-

Yeah, exactly. I, I was like, I'm the-- I'm not reviewing it anyway, whatever. You can write it in whatever language you pick. But the important stuff that I did was not write the Electron bindings.

Yeah.

It was, like, the logic of what happens in the app, you know? And then the model's like, yeah, I can just recreate the same thing in Swift.

Yeah. I, I think you still want, e-especially for people who are doing, like, high-performance software or, like, very complex software, uh, you still want, like, some view of the architecture. Uh, but you can use Markdown for that.

Right. Yeah.

Uh, you don't actually have to read the code.

Again, I-I'm still, like, on a, a sort of like a definitional thing.

Yeah.

Um, can we build a good mental model of Claude Cowork? Um, this is what I have, right? Like, y-you said it's, like, fundamentally Claude Code, we don't wanna touch it. There's the Claude App, there's Claude in Chrome. I think you guys do something different in planning.

But, uh, I've been talking with Tariq, who is on the Claude Code team, and you guys are-- he's like, "No, we just expose the planning." Uh, maybe you can clarify, like, what are the major pieces that people should be aware goes into Cowork?

Like, okay, I think you basically have them. So-

Really?

... um, you can, you can take planning more or less out. I think there's a few things that are really valuable in Cowork. Um, the virtual machine is probably the most powerful thing. So we currently run like a-

Of course

... we currently run like a lightweight VM, and we put Claude Code into the VM, and we do that for, for, um, a number of reasons. Safety and security is a big one. But even if you, even if you ignore for a second safety and security, and you're just like, "Okay, YOLO, I want this thing to do whatever," it is quite powerful to give Claude its own computer.

That is, like, generally a good idea. And in terms of architecture and UX and everything else that we've been working on at Anthropic, it often is quite useful if you do, like, anthropomorphize, um, Claude aggressively and just be like, "This is a person."

Yeah.

What will you do if you give a, if you had a person, right?

Yeah.

And the analogy I've given my dad this morning, who is still, like, quite insistent on using chat even for, like, coding things is, if you were a developer and your employer told you that you don't need a computer, they're just gonna, like, send you emails with the code and you send emails with code back, like, that maybe worked for peddlers-

Right

... in the back, but that is not very effective. Um, so what we can do with the VM is, because it's a, it's a Linux system, Claude Code has more or less free reign to install whatever it needs to install.

It can install Python, it can install Node.js. We do have strict network ingress and egress controls, so you can still, as, as a user, in, like, plain human language, make it clear to, to the entire system what you're okay with and what you're not okay with.

But at no point do we have to ask a real person, like a, like a person who might be in marketing or a lawyer. I'd have to go to a lawyer and be like, "Are you okay with me installing Homebrew?"

Yeah, yeah.

Right? Because the implications of the question and the answer are complex and nuanced and, like, not, not easy to reason about. And this gives us a lot of abstraction that makes Claude very powerful. Now then, around it, we, we do probably have a number of things that also keeps growing almost every single week that you're probably noticing that make Cowork maybe better for certain tasks than just Claude Code on its own.

Yeah.

But most of those actually live in the system prompt. They're about, like, what can we infer about the work that you do? What can we in- what can we introduce-

Mm-hmm

... into the system prompt to make that more effective? It's of course, like, very tight integration with Claude and Chrome. You're noticing that a lot of people, especially as the models get better, a lot of people throw up their hands when it comes to MCP connectors and they say, "All right, I'm not gonna, I'm not gonna go through, like, 25 MCP connectors, click auth everywhere, and then, like, half of them don't let me do the things anyway."

So Claude and Chrome is quite powerful because we can just talk to the Claude and Chrome sub-agent, and that will just do things for you.

Yeah.

That's-

So, so one example, right? I-In MCP, honestly, I think, like, the state of MCP is kind of, kind of, like, really hard to integrate. Um, I need to, I needed to add a Figma MCP to the coding agent that I use.

Yeah.

Uh, and but I didn't wanna read the docs, so I just had Claude do it.

And it's, it's great at reading docs. Uh, and in same, same way, I had to set up, like, a Google Cloud, um, account for some project I was working on and get some API key somewhere, and Google Cloud is famously super hard to navigate.

So I just didn't wanna deal with any of it. I just used Claude Cowork.

Within the first week of developing on Cowork, this happened very, very quickly, um, I caught myself, like, starting to use Cowork for coding tasks, which is not ostensibly what we built it for, right? We don't need to. But I found myself, um, I found myself, like, on our internal, internal tool that we have for c- to collect crashes and just, like, debugging information.

And I found myself sort of, like, picking out the ones that I think we can easily fix versus the ones that might be, like, kernel corruption or something else on the operating system. And I found myself sort of picking these out and then just telling Claude, "Go fix this bug."

I was like, "What am I doing here? Go one level up. Tell Cowork, I want you to go to all these crash tools. I want you to find all the bugs that you think are fixable and not, like, an operating system crash.

And then I want you to tell another Claude to, like, fix all of that." Um, and that's, that's, that's sort of how-

You tell another Claude?

Yeah.

Just so it can spin up another instance or...?

Uh, it-- Currently what I do is, um, and this is a bit of a hack, but I tell it to use Claude Code remote to-

Which is just another website

... call itself.

Yeah.

That's interesting. So you basically take... If you, if you imagine, like, a dashboard with, like, 20 bugs.

You, this is remote control or Claude Code remote? Or sorry, I just wanted to confirm what-

The way I'm using it is I have Cowork running, and I'm telling Cowork, "Here's where I normally go every morning to find the latest bugs. Go read the entire bug list." "Separate out which ones are fixable, which ones are not fixable.

And then for the fixable ones, for..." It's almost a loop. "For each bug, write a markdown file with a prompt. And then for each markdown file that is a prompt, start up a Claude session."

So natively, Claude Code has this concept of sub-agents.

Mm-hmm.

And this is basically a sub-agent, but you're not using the sub-agents functionality.

I'm not using the sub-agents functionality, and the reason I'm not is because I'm firing that off as a Claude Code remote task.

Yeah.

It's kinda nice because then it can just fire it off. I can go to my next meeting, and in Claude Code remote, now the work's happening.

Yeah. You s- you see, like, you're already starting to use the Claude over your local machine, and I think this is one of those things where, like, well shouldn't just everything just be Claude first, right?

Ah, this is such a good group. I have, like, so many thoughts about this. I have so many thoughts about this. Okay. So I generally believe that Silicon Valley overall is undervaluing the local computer, and my default argument for that is always, how come we're all using MacBooks and not, like, an iPad or a Chromebook?

Um, that there is, like, still value in, in having a local machine. And now when I think about Claude as this entity that is supposed to be very useful to you, like, tremendously useful to you, I think that entity needs to have access to all the same tools you have access to.

Otherwise, it's gonna be hamstrung in, like, all these complex ways. And there's, there's sort of two approaches we could take You could say, okay, we're gonna like one by one chip away at everything that is at your computer and move it into the cloud.

That's, that's one way to do it. Um, and I think other products have taken that path. I personally, and this is a very personal opinion, but I personally, for the amount of tools that I use, just don't have the patience to give another tool like permissions to every single thing and keep those permissions up to date.

The second thing that I'm still grappling with, and I don't have a good answer for anyone just yet, but the second thing I'm still grappling with is what does it look like for someone to slurp up your entire work and put that in the cloud?

Like if I just-- as an example, like if you could click a button and it'll just clone your entire computer into the cloud, is that something that you would want? I'm, I'm not totally convinced yet that all- everyone will.

Mm-hmm.

And that is sort of like upstream of all the technical issues we're gonna have because, like in general, I think the world is not ready for this kind of stuff. Like, I'll give you one quick example that would probably be very easy for us.

So as a desktop app, we, in theory, with your permission, can do a lot of things on your computer, including reading your Chro-Chrome cookies if we really want to do it, right? We could take your Chrome cookies, you would have to decrypt them for us, but we could put those on the cloud if we really felt like it.

Pretty easy solution that would be super cool. We could just be like, "Oh, we can do all your tasks in the cloud now." Um, a lot of websites, banks included, if, if they see the same authentication from like two different locations, will just lock down your account, and now you have to go to the branch and be like, "Okay-

That was me

... I'm here with my passport."

I didn't actually know that. Wow.

You know, as tired as we all are of the term agent for the agentic future, I think there's a lot of stuff that sort of slowly needs to catch up. And until that's the case, the way I, as someone who's working on Claude, can make Claude most effective is to like put it where you're working.

Anything else with our, with our mental model? So like basically like, uh, part of me also just want like the more I understand how it works, the more I can use it to its full potential, right?

Yeah.

And so what I'm get- hearing from you is you told me to delete the planning thing. You're not doing anything special on, on the o- that's only exclusive to Claude Cowork.

We have some tricks, but they're sort of like change week over week.

Okay.

We eval Cowork maybe against different use cases than you would eval Claude Code, right?

Ah.

If you think about it this way.

Okay.

So like Claude Code is very-

How do you eval Claude Cowork?

Yeah. So Claude Code is like quite optimized for coding tasks, and we mostly evaluate whether or not we're getting better or worse depending on how good it is at like a typical suite job. In Claude Cowork, on the other hand, we evaluate more against typical knowledge work, the kind of stuff you would find in finance or in like maybe a, like, in like a legal office.

Um, my personal use case is always like managing my... things like managing my personal mortgage or something like that, right? Or like wealth planning for me and my family. Those are the kinds of use cases we eval Claude Cowork on.

And what you might be picking up on is like the subtle changes we make to the system prompt, what we put in the system prompt, how we steer Claude, with the tools we give it, um, should like either be better in one or the other direction, wherever there's a trade-off.

Trade-offs exist a lot. Claude Code will be better for code and Claude Cowork will be better for non-coding tasks. Will those gaps still exist in the next few generations of models? It's like a little unclear to me, though.

Yeah.

Because right now, these like hyper-optimizations we make, I'm not sure for how long they're still gonna be relevant.

I think what I was referring to was also it, it just, uh, it qualitatively felt different when I-- probably it's just all prompting and I'm reading too much into it, but like the, the fact that y- it comes up with like a nine-step plan, I can edit the plan and give feedback and, and, and see it execute the plan.

Yeah.

It felt more long range than in Claude Code, but maybe that already existed in Claude Code and you just built a nicer UI for it.

It's kind of both. Um, like if the Claude Code people who build the planning functionalities would sit here, they would probably say, "Yes, we have all of those things in Claude Code," and they do. Um, I think people tend to give Cowork tasks that are maybe of longer time horizon.

I thought it's so long. Yeah.

That's like one thing, right? You're just like the, the, the chunk of work tends to be maybe a little bigger. And then the second thing is that because the work, when it gets longer, it gets a little bit more ambiguous, we do tell Cowork to make heavy use of the planning tool or to make heavy use of the ask user question tool, right?

We do want it to come up with like different scenarios. Okay, tease out what the user actually wants. Don't go off to work for like four hours and then come back with the wrong thing.

Mm-hmm.

And you're probably picking up on that.

Yeah.

Um, I wish I could tell you I, like built this magical thing, and it's like there's some secret sauce.

Oh, no, no, no. I mean-

Yeah

... it's, it's just clarity is good.

Yeah.

You know, engineers just wanna know-

Yeah

... so they can, they can plan around it. And then, uh, I think also for me, um, I'm realizing I have to switch to my, my other machine because this is a new machine that doesn't have my session.

But, uh, yeah, the, the, the planning is really important for, for me to like approve or like to see whether it's like it's right. The ask user question is so beautifully presented. I mean, it is, it's also available in like Cursor and, and in Claude Code, but like I, I think like it's so nice to see that it like...

It's kind of for me like to understand that it gets me, it gets what I wanna do.

Yeah. Yeah.

And like it's-

It tries very hard.

Just on the topic of evals.

Mm-hmm.

When you say eval, I think people are very vague about what it means. Is it just like vibe testing or do you have like automated programmatic evals of Claude Cowork?

When we say eval, uh, what we really mean is that we essentially take the entire transcript, including all the tools that Claude has available ultimately to it, and we then measure what are the outputs depending on what we tweak, right?

So we do run that a lot. We use that in training. Um, we use that in, in like if you sort of separate out post-training from like the scaffolding around it, Cowork sort of exists in the scaffolding space, but obviously we also train on it a little bit.

Um, so when we say eval, we mean given the certain transcript, what do the outputs look like-

Mm-hmm

... including the file outputs as well as like the actual token outputs, like the ones that you see in the chat window.

I'm curious, um, how much of the failure modes are the model intelligence versus like the usage of the end tool to put the intelligence in. Like the wall planning is like a good example, right? It's like one thing is to come up with a plan, the other thing is like make a nice spreadsheet-

Yeah

... um, that kind of runs you through the plan. Like how have you seen that evolve?

The thing that I grapple with a lot is that whatever scaffolding you come up with, I think we still have a bit of sort of like m- model overhang where the model is dramatically more capable than what users end up using it for, and I think part of that is that we're just not getting the model all the tools to do all the things it's theory capable of.

Right, that's like one thing. Um, however, whenever you do build the scaffolding, I'm sort of wondering at what point, at what point will that scaffolding go away, and like how much you invest in figuring out what the right scaffolding is, is kind of up to...

It's a little bit of a bet, right?

Mm-hmm.

And one thing that I as an engineer quite enjoy is that like working at Anthropic and working at a frontier lab, I maybe have a little bit more insight into what's coming, coming down the chute in terms of like what's the next model, what is the model capable of, what it's good at, what is it bad at.

And I'm, I'm increasingly wondering, is the right thing for us to like really invest too much in sort of these like scaffolding corrections where the model might otherwise not misbehave, but just not do the thing that you want.

Yeah.

Or is it to just like give it as many capabilities as possible, try to make those safe so that the worst case scenario is like not as bad as it might be otherwise, and then just simply wait a second for the next model drop.

I'm personally currently more leaning toward the latter. I think we're gonna see a lot of like applications and companies that do very impressive things with AI that in the short term might seem very effective because they're very specialized to individual use cases.

But I think once models get better at generalization and get better at like those specific use cases without being super guided on those, I'm not sure how long that's gonna stick around. And you can kind of, kind of already see this in like Skills and MCP servers, right?

Mm-hmm.

We've, we've already seen sort of this like slow shift from MCP servers to Skills, and like maybe a good example is Barry, who made Skills. He was initially hacking on something that honestly looked a lot, looked, looked a lot like what Cowork does today.

He was sort of thinking about what if Cowork, but for like people who don't wanna build code.

Mm-hmm.

And, um, he too did that as a prototype inside the desktop app. One of the first use cases we thought of were, okay, what, what are like coding-like use cases that could really benefit from graphical interfaces and like from being a little separated from the actual underlying code?

And everyone comes up with the same answer, is data analysis.

Right. Yeah.

Whereas like how many users do we have today? How many... Like, it's always data analysis. And I think the thing that ultimately led to Skills is that we wanted to connect this little prototype to our data warehouse. And the team very quickly discovered that like instead of building a custom tool for the thing to talk to our data warehouse, they just like made a markdown file and were like-

Mm-hmm

... "Dear Claude, if you wanna get data, here's the endpoint, here's what the API looks like. You figure it out."

Ah.

And then it ended up-

Like hand over control.

Yeah. Yeah.

Yeah.

Also, just like maybe go one step up in the layer of abstractions, right? Just-

Yeah

... instead of, instead of telling the thing, "Here's a CLI, please call the CLI," or "Here's an MCP, please call this interface shape," just like, "This is the endpoint if you wanna know something. If you post here, maybe you can do post SQL.

It's gonna be okay."

Mm-hmm.

And that ended up being so effective that they started trying the same pattern of like just giving the model a markdown file that describes whatever it needs to do, that the whole thing eventually became Skills, and we're like, "We should package this up.

This is a-"

Right

... "good idea."

Yeah. Um, we've had Barry and Mahesh, uh, on, on our conference and, uh, he's, uh, definitely got a good idea there.

Yeah.

I wanted to show you the, how I've been using Claude Cowork. Uh-

Ah, so this is like almost my favorite part.

This is, th- so this is like me, uh, this is how we run the Discord. Uh, we literally... Uh, at, at first I didn't trust Claude Cowork. This is my very first usage.

Okay.

Right? So then I was like, "Okay, I will just try to manually download from Zoom all my recordings and upload it to YouTube."

Yeah.

Because this is a very laborious process. I gotta click, click, click. YouTube, um, isn't super user-friendly. Uh, and it just did it. And then I was like, actually, you know, even the download from Zoom part, I should also put into Claude Cowork, and then I did it, right?

Here's a bunch of it, and it starts compacting here and it, and it, it starts to e-even be able to do things like look through the individual frames of the video to name the video so that I can upload it auto-automatically.

Oh, that is-

And this replaces my job as a YouTuber.

We will forever appreciate your creative-

Yes

... you know?

And so that's great. Uh, but then by the way, it compacts and makes, it makes like a new thing, right? So I, I don't, I don't have the original, original thing. But then I asked it to make its own skill so that it, so that something that's repetitive and one-off and human guided becomes more automated and I can use the skills independently and reuse them, uh, and it obviously can write skills.

And that, that goes into context and skills at the bottom here, which is-

Yeah

... which is so nice. Um, so I have all these skills that, that I now sort of do on a weekly basis. Uh, I know you've released scheduled Coworks, which I haven't done yet, but it should be-

You really closely should try them. I, I think this is like so wonderful and fun for me to see because I think one thing that is very fun for me about Skills in particular is that they're so easy to make.

Like anyone can make a skill. Like a text message could be a skill, and they can be so hyper-personalized to you, and this is like sort of the subtraction layer, right? Like, um, I s- I'm just guessing, but I assume-

Yeah

... like you're very good at your job. You've probably given this thing some guidance about how to do it, right?

I, I just said wrap everything up into, into a skill, right?

Yeah.

And then, uh, and then I was like, actually sometimes I might need to break, uh, things apart because some parts fail or some parts might be needed indi-individually. So I told it to split one skill into three skills.

So it's like a sca- skill splitting thing, and then there's like a parent skill that just orchestrates all of them if I want to use that.

Yeah.

But like, um, I think that's, that's like really good. Uh, and, uh, but there's, there's one more part, which is the, uh, Google Chrome thing that I told you about-

Yeah

... where I'm like, okay, you know what's better than uploading using Claude Cowork to YouTube? Like actually l- looking at the docs to like programmatically up- upload to YouTube and then putting that in a skill. And I've never done that before.

I don't want to deal with Google Cloud.

Yeah.

So Claude Cowork does it for me.

That is really cool.

So, so I, I just, I don't care. I just like d- let it do its thing. I don't... It doesn't really matter.

That is really cool. And then you've, I assume, paired the skill just with this script that it's built?

Yeah, and then I just update, update the skill.

Oh, that is beautiful. Yeah, that's wonderful.

It's kind of like a skill-- Uh, uh, basically, I think, like, the way that people ease into Claude Cowork is, like, take a knowledge work task that you would normally be clicking around for, and then, uh, try to turn, turn that, and then you do the, "Okay, well, what if you went further?

Okay, and then what if you went further? What if..." And then you sort of expand the scope of Cowork as you gain trust with it and un-and also teach it how to replace you.

Yeah. It's like a little bit like playing Factorio but for your own life.

Oh.

Like you say, you start really small.

Yeah.

You start automating something really tiny, and, like, once it clicks, you keep adding onto this, like, automation empire and just, like, make your life easier and easier. My favorite skill has been, um, every single morning, Cowork starts looking at my calendar and make sure that there's no conflicts because people tend to schedule a lot of meetings sometimes last minute and sometimes miss it.

It's often painful.

Mm-hmm.

And a lot of products have existed like that a lot. I've written in the custom prompt there. I haven't made it a skill. Um, honestly should.

Yeah.

But I've given it, like, pretty clear instructions about, "Okay, here are some people. If they book over other meetings, I'm probably gonna go to their meeting." Like, if Dario schedules a meeting-

Right

... not try to reschedule Dario, right? Um, and I think there's some other rules in there about, like, what kind of meetings I care more about, what kind of meetings I care less about, what is okay to, like, maybe punt, like, when I want to be, when I want, want to be working, when I don't want to be working.

And it's those really small things that I kind of think kinda click with people. Right when we launched Cowork, I think one of the use cases that went most viral on Twitter, X, was clean up your desktop, which is of course silly.

That's such a silly thing, right? Like, you don't need a model to clean up your desktop. Not really. Um-

Like this? Like clean up my desktop?

Yeah, exactly. Yeah.

I need to, I need to choose my desktop, right, I guess. Give it access to my desktop.

Yeah.

Okay. Uh, okay, this is very scary. Uh, we'll do it.

I did, I did it with my downloads folder. It was like, "You have so many term sheets, and there's, like, eight copies of your rental lease for your office." I was like, "All right," uh, like, "Don't yell at me."

It's like, it's such a small task, and then, like, I, I would never go out there normally otherwise and tell people, "I've built a product. It can organize your folder for you."

Right.

Um, because it feels small. But I think to your point, like...

Oh, here's, here's the, here's the ask user questions.

Yeah.

Uh...

Beautiful, right? Delete obvious junk. You probably shouldn't click that.

No.

If it's not done right.

As long as it's reversible, I don't care.

Make a plan to.

Yeah. Uh, yeah, no, I, I have a, I have a typical everything is super messy folder, so yes, I think this, this is super helpful. So this is a pretty simple task.

Mm-hmm.

But I've-- Okay, here it is, right? Here's the progress. I don't see this in... That's why I'm like, "This has gotta be something different than, uh, than Claude Code," because I'm like-

We do, yeah, that's, uh, we do system prompt it. We're like, "All right, we want you to think about, like, this task."

Yeah.

And methodol-

Yeah. And then I can, I can, I can do, like, little suggestions for, for, for these things. It's beautiful. Look at this. I, I can, I can, like, say, like, "Oh, don't do that. Don't do this." Like, it's amazing.

I'm so happy you like it. Um, I mean, the other way around, like, we're part of the Claude Code team, if you would like this in Claude Code.

Yeah. Damn.

All right.

Uh, so, so yeah, I, I mean, uh, this is really good. Obviously, I'm, like, kind of raving about it. Uh, a-and, you know, I have other things like sign up for PG&E. So if you can do phone calls for me, that'd be great.

Um, I, I do-

People have done that. Obviously, you can't do that natively-

Yeah

... but people have done that with, like, various other providers.

Yeah. Uh, and, and then this is, like, signing up for the Figma MCP. Um, a-and I really am trying to do, like, everything, um, data analysis as well. I do think, um, oh, design to code, uh, very, very good, right?

So, like, here's a Figma file. Take it. And then this is where, like, a lot of other tasks, it's like knowledge work, like replace my manual clicking, but this is... No, I would normally use Claude Cowork, uh, Claude Code for this, but because I perceive that you have better Chrome integration-

Mm-hmm

... I, I think you can actually do a better job of this, and I, this, this one-shotted my, uh, conference website. Uh-

That's pretty cool

... which, yeah.

Like, at some point, I would love to, like, hear how you feel about code in the desktop app. Just like-

I never use it.

Which is the, the same team. Same team.

So I use the C-Claude Code in terminal, which-

Yeah

... I, I perceive to be the default way of Claude coding.

So one thing this has... Sorry, I'm just like-

No

... I'm not here, I'm not here to wrap all these products.

Please, please.

Okay, I talk about other stuff too. I'm like, I'm not sure if people out there wanna, like, hear me advertise my stuff for, like, an hour.

Please do.

But, um, this thing has, like, a built-in browser, which is a thing a lot of products have these days.

Yeah.

It's a built-in browser, and I think giving Claude eyes into, like, what you're actually working on makes it so much more effective, and that's probably what you've seen in Cowork. Because it can see Chrome, it can, like, debug the DOM, it can, like, see things, um, that does make it more powerful.

Yeah. Uh, so, so I think, uh, my mental model is kind of broken 'cause I only use this Cowork because I thought it had a, a browser thing in it. But I understand that the Claude Code app or the app version of Claude Code does have a built-in browser.

I've seen, I've seen this preview thing.

Yeah.

I just, I've never used it.

But in the end, in the end, you sort of get like-

Have to try hard.

Yeah. You basically get the same thing, right? Like, the, the, the additional skill that you're describing is Claude is better if it can see what it's working on, right? That's, that's sort of, like, the summary here. And, like-

This, this is it

... whether it's using your Chrome-

Yeah

... or it's just, like, making up its own little, like, browser, it doesn't really make a bit difference because either way, it's gonna see what it's working on, and that just makes it much better.

Well-

And then-

Yeah

... you don't have to run QA for your Claude.

Why doesn't it pick up my existing Claude Code sessions? 'Cause I, I mean, obviously, I've used Claude Code, but...

Excellent question. Um, don't have a good answer other than, like, we're honest.

Just haven't.

Yeah.

Okay. This is what the OpenAI team does, anyway. Uh, cool. I, I, I don't have other... Like, I, I just, I, I do wanna ex-expand people's minds and also maybe show people if they haven't really done it. But, like, I, I think it's very interesting how I sometimes use this more than I use...

I mean, I use Dia, right?

Yeah.

Um, I, and I use, uh, I've used, like, all the other agentic browsers, and Anthropic didn't have to build an agentic browser because you just had Claude Cowork, and that's enough.

Yeah. I, I also think, like, maybe integrating with number of excellent browsers out there is, like, currently on my Personal priority list a little higher than, like, trying to rebuild a browser from scratch. Yeah. You know, never say never, but I think going back to this idea of, like, we wanna plug this into your current entire existing workflow, I think our goal is actually to not replace any of the applications you have on your computer, but instead, like, work really well within your workflow.

Make the new one. Yeah. Yeah. It seems that nowadays, especially on the browser, most of the innovation is, like, user ergonomics. It's not really, like, the underlying browser engine. So I feel like to Claude, it doesn't really matter if it's like Via or Chrome or Alice, whatever- Yeah.

We wanna- You're just doing- We wanna meet you wherever you are, which is like... Like, obviously, I would say that, but it's also just generally true because I don't wanna restrict my potential user base artificially by saying, "Okay, like, I'm gonna start building for the people who are willing to switch browsers."

Right. That's such a like, you know, like many lawsuits have been filed over who gets to write the browser. Yeah. And, like, m- a lot of money has switched hands over the question of, like, which browser is default and which search engine is default within the browser.

Um, I just wanna build for... Yeah, I wanna build for Swyx essentially. Like I wanna, wanna, I wanna build for people who have a number of annoying tasks that they feel like maybe Claude could do it, could do it for them.

Yeah. What do you think about skills portability? I think there's been one thing-- I use another thing called Zo, which is kinda like a cloud computer plus agent, and I have a skill to add visitors to the office - Yeah ...

so whenever somebody has to come in after hours, they need to check in downstairs. Um, but I wanna, like, text the thing, so it doesn't really work in, in Cowork. But now that skill is in the Zo harness, and it's not in my Cowork thing, and then if I make a change, it's gotta...

I gotta sync them. How do you see that going? Like, I see memory as, like, Claude personal. Mm. Kinda like I don't necessarily want my memories to be cross-thing. Yeah. But I do want my skills to be cross-agent that I use.

Yeah. I think with MCPs, people did the same thing. It's like, oh, MCP gateway, MCP registry. I don't really know if that's, like, a business. So I'm curious, like, if you've had any thoughts in the area. I think for me this is sort of where I go back to the really basic primitives, for our skills are file-based instead of, like, this complicated thing that exists inside a place somewhere that is, like, super pro-proprietary.

I'm really leaning into the idea of, like, it's all just files and folders, and that makes it very portable in its own right. We do have skills as part of this container format, which we just call plugins. Mm-hmm.

And plugins are available both for Claude Code and Claude Cowork, the same format. And you can install plugins. This works in Cowork today. You can basically say, "I'm gonna add a whole, like, just a GitHub repo as a skills marketplace or, like, a plugin marketplace," and that's how we're doing portability.

I think we have a lot of room left to grow in how do we make it easy for people to know that they can write skills? How do we make it easy for them to just, like, share a skill with you?

Because obviously all the words I just said, right, like I'm losing most of the knowledge worker base out there- Right ... and start with saying, "Oh, you can connect to GitHub repo." It's not exactly how most people will end up working in, like, a general knowledge worker space.

Um, but I think there's something there. And another thing that's there that I think has not really been properly explored is the, the, the combination of which part of the skill is very portable and then which part of the skill is, like, very personal to you, right?

And I think that's something we haven't really solved yet in industry. Mm. It's like, which one do you wanna introduce more structure to the skill or have... always have, like, public skill, private skill, you know, pairs. Yeah, yeah, kind of.

I think that's like a... Like, the easiest way to do this would just we do, like, use string interpolation or something, right? Right. Yeah, yeah. Insert username here, insert, like, phone number, insert, like, known folder locations, that kind of stuff.

Um, that's probably clunky. That's why we haven't built it. Um, uh, but I do think someone is going to come up with, like, an interesting way to keep everything we like about skills. The portability, it's just a file.

It's just markdown. It's just text, honestly, right? Like a text file. What's the complete lack of structure, which means you don't need any kind of tutorial to write a skill. Just, like- ... explain it to Claude the way you would explain it to me, and Claude will probably get it before I would.

Mm-hmm. Right? You're just like, for booking a flight, tell Claude how to book a flight the same way we're telling- Right ... someone who just started working here today. But combine that with a very, like, personal thing. Um, maybe we'll stick with the booking a flight example.

I don't actually think AI should be booking flights. I think the tools we have is- Yes ... pretty. Yeah. Finally, somebody says it. It's the default demo that everyone was making. I'm like- I, I immediately get slight booking demos.

It, uh- It's not a good showcase. Yeah. I'm like, I just wanna book my flight myself. But- Right ... um, I think there's a lot of things that have a personal and a non-personal component, and that's maybe why people reach for flight booking because some things are very universal.

Yeah. Cheaper flight is usually better, right? Like, few people try to book the most expensive flight. And then some things are quite personal about, like, what times you prefer, which seat you prefer, which airports you prefer. Combining that in, like, a skill format that is actually portable, compatible, easy to understand for people, I think that would be very exciting.

We just haven't figured it out yet. Yeah. I think the text part... Every... I think everybody by now has some sort of, like, cloud file thing. Mm. Either Dropbox, Google Drive, whatever. So it feels like, in a way, it should basically, like, symlink my skills into all my agent harnesses.

Yeah. Just keep those in sync. Like, we have internally this, like, valuable tokens repo, which is, like, all the commands and subagents. It's good. Uh, and then I built, like, a TUI where you can start and be like, you know, install this command and these three subagents into this agent in this folder, and just copy-paste this.

It doesn't do anything. It literally CP the file into that. But I feel like there should be something similar where, like, whenever I go into a new thing, it's like, hey, here's, like, the link to exactly the cloud folder and just bring down these skills into this.

Yeah. Like, today, it doesn't quite work like that. Like, if I install a new agent, I cannot... I have to, like, copy-paste all the skills, and I don't even know where they are. Yeah. That's a big problem. That's, like, the big problem.

It's like, where do I find them? Yeah. Um, so I'm curious, like, in the future, like, that, that almost feels like my personal productivity thing will be my skills- Yeah ... is not really the product that I use.

Because everybody has access to the same product. But today there's-- that just looks like copy-pasting MGA files.

I think so many things. I, I really like thinking about agents and LLMs just as, like, another coworker. So many attempts have been made to build documentation companies that are like, "Oh, we're gonna solve all your documentation problems."

Um, I myself, like, spend a little bit of time working in Notion, right? I'm, like, deeply familiar with the concept of let's get everyone on the same page.

Mm-hmm.

Right? And what you're basically saying here is you want all your agents to be on the same page about your preferences, about the skills, about the way they ought to work, and, like, how they ought to execute. And I'm not sure what the right thing is going to be.

If it's going to be some, some company that can say, "All right, we're-- as an independent body, we're not trying to, like, push into any particular product. It's our job to be, like, the skill authority, and we provide...

I don't know, we're gonna be the Dropbox of skills, and we can just symlink us into all the products you wanna use." I'm not sure that's gonna be viable business, but as, as an idea, it would be cool, right?

Yeah, yeah. I think so many things are just going away as businesses. It's like, how am I supposed to do it? I'm not even asking somebody to make a product about it. Like-

Yeah

... I wanna personally know. And there's things, like you said, it's like you almost want a skill and then interpolate it between personal and work. So if I'm booking a flight for work, it's different than I'm booking a flight personally.

Yeah.

In some ways.

Yeah.

But, like, a lot of the scaffolding is the same, you know?

I mean, as an engineer, I will tell you, like, you know, technical person to technical person, I will just be like, "symlinks."

Well, that's what, that's what I do with Claude.md and Agents.md. It's just the same with how a symlink. And so it's like that works, but it feels like, yeah, I don't know, maybe...

You could always go one level up. You can always tell a coworker problem, and then coworker will solve it for you. Just make the symlinks. That's like one way to do it.

That's true. That's true. All right. Everything is called cowork.

Uh, potentially spicy question for both of you. Uh, which of these industries will go away?

Okay, so what Felix was saying before is interesting. There's basically like the short-term pressure of, like, we need to turn these tokens into valuable things, which is I should build the last mile product that harness the model. And then there's the question of, like, long-term, which ones are gonna still be valuable.

And I think you're kinda seeing this today with like, uh, you know, the coding space in a way is kinda like everybody's moving up and up in stack because you need more than just turning tokens into code. I think search, like enterprise search, is kinda seeing the same thing, like, with Glean and, like, all these different companies.

It's like at the end of the day, if cowork is the one doing all the work, the search itself is, like, such a small part that, like, I don't know if I'm really gonna pay that much money just to do search.

It's almost like everything is like a cowork vertical. So, like, how much can cowork first party support-

Mm-hmm

... and how much can it not? I think for a lot of these things, the planning thing that you were showing-

Which one? Is-

The, the planning. The planning.

Okay. Yeah, yeah.

Like, that's one thing where, like, most of the value that these agents provide is, like, they're better at planning for specific tasks-

Yeah

... and have better tools for it.

Yeah.

But I think the models are now moving in that direction, and they have the right harnesses, and they're on your computer. So for me, it's almost like if the end customer trusts your startup to be the provider of that task result, then I think that works.

This is, uh, something that-

For sure.

... this is a spike that we're, we're working on. Uh, yeah. I think... Lo-look, I'll, I'll, I'll tell you this. Like, I don't think I'm the best person to, like, actually estimate which industry is going to be hit the hardest.

But I do think that at Anthropic, as a group of people, we're deeply worried about the impact that the tools are going to have on the labor market, especially for like junior employees. Because I think, I think it's only honest to say that when we talk about automating a lot-- away a lot of the work that we personally find annoying, that we maybe think it's not the best use of our time, in a lot of industries, that kind of work would have been given to a junior entry-level employee.

Yeah.

Right? And I think it's, it's only, it's only right to be really worried about that and, like, worry what that's going to do, in particular to people who, like, enter the job market.

Mm-hmm. I have a solution for that, which you make them... You create simulative jobs for them.

Okay.

So this is, this is like half joke, half true. So if you think about software engineering, when you're like a junior engineer, you work like one, two, three years, and in those three years there's like maybe like a handful of moments where like you really learn something and then a bunch of other days where like you're not really progressing.

Yeah.

I think now we can use AI and these models to actually like shortcut these careers and almost like simulate the early years of your work and like just-

Yeah

... make them like super dense in like these learnings. It's like, "Hey, we're working on this feature, which is like a distributed system, and you need to learn this thing." That might take three months at a company, and so you take three months.

Here it's like we're just simulating the whole thing. It's actually not a real thing. And in one week we kinda speed run through the whole thing, and you kinda learn your lesson from there, and we kinda repeat that.

In like one year, you basically get like three years' worth of like projects and experience.

Yeah.

I think it's harder for like things like sales or for things like, you know, marketing because you don't really have a way to get the feedback loop. But I think a lot of it, it sounds kinda silly. It's like you're making the new fake job, but it's almost like you go to college, right?

People pay to learn how to do it, and this might feel similar where it's like, "Hey, we have the Jane Street simulator." It's like, "You wanna come work at Jane Street? We'll just put you in the simulator for like three months-

Wow

... and you'll come out of it as like, you know, 'I'm ready.'"

So there, there is an aspect here. I'm not an expert enough to like actually know what, what is going to happen to marketing or legal or finance, right? Like, I don't work in those jobs, and I, I don't think I should talk about them.

But I am an engineer, and I think I have a pretty good idea of what engineering is like. And I think one thing we're sort of seeing is that as a company and also as, as the public, we're like deeply worried about entry level, but we're also seeing more senior engineers accelerated.

They feel like they're more productive. They, they actually increase the value they provide. And the thing that I'm thinking about a lot is the fact that even before all of this happened, um, I've always had a lot of respect for the University of Waterloo, and the, the new grads that have joined my teams as from coming from University of Waterloo always felt like more ready than new grads who like literally spent their entire time at a university, regardless of how good, but never actually had to work inside an environment where you have to ship things that eventually will be used by users.

And I'm, I'm, I'm German. I like initially went to German university, and I think the, the, the like information systems programs there tend to be very theoretical, right? Like I often give people the example of like trying to become a doctor, but you first have to do four years of biology.

And as a result, when you get a new grad, you sort of have to teach them what it's like to actually build products and to work in a company and like work with other people and like some people will have a different opinion and like how do you do all of those things?

And the University of Waterloo, it seems like they just spend half of their ti... I don't know if it's true, but I think-

It's, it's a year

Right. They spend so much time-

Part of your job or c- a curriculum to do spend a year in internships.

Yeah. They just like go from company to company. They show up on your team as like a junior engineer who's been to like twenty companies. Not really, but like it seems like-

A lot of my new grads have also briefly worked at Apple, Google, Tesla.

Yes. And, uh, there's a common meme where they like c-collect all these logos like Infinity Stones. But and they always put it on LinkedIn, and it's very unclear that they're an intern. Like- Yeah. Yeah, exactly. But it does actually make them so much better compared to other new grads.

And I wonder if that's a useful model maybe for the future when we also have to like crunch down the amount of time you have as a junior employee, because the value you have as a junior employee is going to like be impacted.

My sort of pro young people take is that they're-- you're more, uh, you have higher neuroplasticity. You can learn more, you have less preexisting biases. And, uh, what I is assume is true for you, what OpenAI of-often says is that actually it's the, the younger, like fresh grad engineers that use codecs or they're coding stuff, uh, more innovatively than the, uh, experienced engineers who have a set and preferred way of doing things.

Yeah, as I talk to people, I, I had similar experience.

Yeah. So maybe you're more AI native-

Yeah

... and therefore you're, you, you get cut. But me-- Like I think the problem is you don't need that many of them.

I mean, A-Anthropic is on the record as saying, "We do believe that the impact on the market is going to be sizable, and we do not think that people overall are ready." Right? And we do actually think we should probably talk about it as a society much more.

Yeah.

I'm not sure that I'm like the individual that can add like anything useful there.

Yeah.

But I think as societies with economists and, and governments that need to wrestle those questions in a way that is probably more meaningful than me wrestling with them, we're probably not doing it enough.

Yeah. Well, we'll try to educate and then I, I think also just releasing frequently as, as, as you guys do, or probably maybe too frequently-

Yeah

... uh, is helping people to adjust over time, right? Rather than one big bang thing. There's like sort of this gradual takeoff that people are living through that-

Yeah

... we can follow up, right?

Yeah. And I-- but I think a lot of us like wondering at what point do we actually have full takeoff, right? Like at what point is there-- we're all sort of expecting this like big bang moment where things will accelerate, accelerate so quickly that it becomes a self-reinforcing loop.

Mm-hmm.

And at that point it's sort of like off to the races and there will be no more like slowly catching up, you know, just have Claude being so good at everything.

Yeah. It's when Cowork is training models. It's when it's looking at TensorBoard and-

Exactly

... weighting biases-

Yeah

... and training things.

And like we can all debate like how many years it's away, right? Like some people make a better route, like maybe it's ten years away, maybe it's a year away. Um, I'm not entirely sure where, where I come on this time, but I'm not entirely sure that ultimately it matters all that much whether or not it happens in four or five years.

If we've-- have a decent amount of certainty that's going to happen, it's probably something we should wrestle with.

I wanted to talk-- So by the way, the, the scheduled task comple- or the, the, the clean my desktop task completed and it did. It organized by file type, which, okay, but you know. I was trying to get it to do more sort of thematic, like read the file, understand what it's about, group by, uh, the, the topic rather than the file type.

But-

I mean, you can just follow up and have it do that. So-

Oh, yeah. Here.

Like it did, it is proposing this, right?

Yeah. So it's got some like topical things, but, uh, yeah, I could probably do better. Like, yeah. So like I probably need to give it a skill to read video files so that it understands here's how I like to-

Honestly, though, like, um, I see that you're using Opus four point six, right? Like my recommendation for people is increasingly don't worry about it anymore. Just like tell it what you want it to do-

Yeah

... and it's probably gonna fi-figure out a way to do it.

Okay.

It might not be the way that you like necessarily or the way that you've gone about it. Um-

Videos deeper. But we're outsourcing organizing all of it, so that's fine. Yeah, yeah. Yeah. Very cool.

I'm honestly like so curious what Claude is gonna come up with.

I'll kick that off. I wanted to also just talk about the, the overall, uh, you know, you, you talk about data analysis. You talk about like, uh, your, your personal finances. You also said, uh, which by the way, for us is very timely, tax season, right?

Like-

Yeah

... you use Claude Code for tax season. It is not responsible for any mistakes, but might as well, right? Like it's, it's free knowledge work for you.

Yeah.

Uh, so I just like I think Claude for Finance is a big deal. Um, and this is definitely like in that mix. I wonder is it like do you-- is it a separate team? Do you talk to them?

How important is it, right? Like is-- you can al-- you also natively output Excel files now.

Yeah.

Just talk about the finance effort in general.

Yeah. We care about the verticals quite a bit. So we do have a dedicated verticals team. We also have a dedicated enterprise team, and those-

Is this engineering, not sales?

It's engineering. Yeah.

Yeah.

Yeah. It's engineering. So we do have people who sort of come to work every single day and they, they ask themselves: How do we make Cowork extremely effective for people in those specific industries? How do we make it easier for them to understand?

How do we make it easier for them to plug into this and like sort of get the same value out of it that software engineers get? I think it's no real surprise that software engineers ended up being sort of at the forefront of the entire AI moment because so much of it is this like Rube Goldberg machine-esque, where like we're already used to automating things, right?

Like it's part of our job.

Yeah.

So we care about it quite a bit. I think it also like really matches- What we see Claude being very good at as a model, I think it provides tremendous amount of value to those customers in particular because we can do so much with the amount of data they have.

Those are like data-heavy industries. They're industries where correctness matters quite a bit. So-

For us, if I've used it to analyze my business, I just can't show it. So-

That's two cents. I had a similar question about my taxes. Like I did tweet, I did tweet about the fact-- I did tweet about, oh, Claude was doing my taxes. This is honestly incredible. And, um, it's like annoying because like this is so cool, but I'm not gonna- Twitter is maybe not the audience that needs to like see my tax return.

Yeah, but here, here it is. It's, it's reading all the videos, so it's like, yeah, it's getting more... Yeah.

How did it actually do it? I'm actually curious.

Oh, usually it just like takes a screenshot, and then it reads the screenshot vi-by vision. So this is what I do for my, my Zoom upload thing, right? Because I, I have paper club sessions that I need to upload to Zoom, and I want it to automatically, uh, title them and do show notes and everything, so it just takes screenshots and try to-- try its best.

Yeah.

It wouldn't probably benefit from transcribing, which it's, it's doing by-- it's operating by pure vision now, but it's good enough.

Yeah.

And then I, I, I do have to call, uh, out to NanoBanana to do images. So unless you guys do images for me, uh, I have to call other people for images.

We're aware. We're aware. It's just like so fun for me because like this is the thing that I'm increasingly doing. Like increasingly curious about Claude's creativity and like figuring out what-

It's great

... Claude's approach is to like solve the problem.

Yeah. Vision for everything is, is like the, the superpower, right? Like, you know. A-and computer use, you guys were the first to do computer use, right? And when it was launched, I was very unimpressed.

Mm.

I was like, "It's slow, it's unreliable, it's bad."

Isn't it wild how much better it's got?

It's one year ago.

Yeah. I know. Like it was barely usable. Yeah, I, I remember. It was barely usable. But isn't it wild how much better things have gotten-

Yeah, yeah

... over like one year?

We went to the Anthropic office because you g-- uh, for the launch event for computer use, like there was like this hackathon.

Yeah.

And like nobody hacked on computer use.

But I did see... I, I, I don't know if you're okay with me saying that, but I did see briefly that you do have like a, like an automate macOS MCP server installed, right?

Uh-huh.

Do you use that ever?

What? Sorry, which one? Where?

Um, if you go to your settings.

Oh, settings, okay. Uh, where-- Sorry, this one?

Yeah.

Yeah.

Um, I noticed that in your connectors.

Uh-huh. Uh, I probably set it at one time, but I don't use it actively.

Oh, okay.

A macOS automator? Yeah, yeah.

Yeah.

So, so I-- Yeah, this one I really wanted to like just automate everything in my thing. I didn't fi-- I didn't find it super reliable.

Okay.

Yeah. Why?

No, no, no question at all.

Claude is much better at writing AppleScript and executing its own AppleScript than relying on these, uh, third-party tools.

Yeah.

Uh, so I've increased-- I, I initially installed IMCP and like all these other FCPs that people built and-- but now I don't use any of them anymore.

Yeah.

Like just, just let Claude write its own thing.

Yeah.

It's gonna be more custom-made.

We keep going up the stack. But I do think computer use is like a fairly interesting area to me, and it's, it's like also interesting in the sense that I don't think we're far away from... I don't think we're far away from Claude being very effective at like using your computer and not just a theoretical computer.

Mm-hmm.

What's the relationship between the user and the computer? Like, uh, there, there were some tweets about how huge some of the VMs that Claude Cowork creates are. It's like twelve, fifteen gigabytes and people complain.

Yeah.

But at some point it's like if you're using the computer, you're taking action on, is this just your computer, and I'm just looking at it? You know, it's like I, I think that's why people like the idea of like the Mac Mini and the OpenClaude or whatever-

Yeah

... on it because it's like it got its own home, you know, it's doing its thing. I'm doing my thing. I think there's some kinda like, not like race condition, but it's like, okay, if I kickstart this task, now I can't really use the computer-

Yeah

... you know, because Claude Cowork is doing things on it, and it's kinda awkward. Like, yeah, I'm not sure.

I, I do think it's a super interesting area because I, I can maybe tell you like some of the things I've thought about that I think are actually a bad idea. So when, when we initially started working on Cowork, I, I did have some dreams about what would it look like for Claude to have its own cursor.

Could be cool, right? Like it's a computer. We can write code. We can touch everything. Like who says that computers need to have one cursor? We could do a second cursor, right? But that actually breaks down quite a bit even if you go and like present cool dreams to both Apple and Microsoft.

You're like, "Wouldn't it be cool if..." Um, it breaks down quite a bit because so many of our models on a computer are built around this idea of like there's only one thing working on it.

Yeah.

There's like a foreground app, a background app. Claude and Chrome can work in the background, but that's like within one application. But at the operating system layer, that is a lot harder to implement. So I'm, I'm still grappling with what, what does it mean for Claude to actually act on your computer?

Is the right format for Claude to have its own computer that you set up and maybe every now and then you like zoom in, and you play with it? Or is the right format for Claude to just like wait until you're stepping away for a little bit and take over while you're gone?

Or is the right move for Claude to just like have his own computer in the cloud, and like whatever you want Claude to do, you have to set up yourself? Right? There's like a, there's like a number of different options.

Um, this is a thing I think about a lot, like what is the relationship between you and your computer and you and your data on the computer? Because how intimate that relationship is kinda depends on the tool and-

Right

... the thing that you're currently looking at, right? Like we're quite comfortable sharing some things, very uncomfortable sharing other things. And I think whatever product is gonna be successful will have to deal with those, like with those different things.

But you probably, even if Claude was capable of making a determination, would you want Claude to make that determination in the first place? It's tricky, right? Because it's like it's more than just privacy. It's like almost intimacy, and it's like tricky to reason about in a way that will make everyone comfortable.

Yeah. I could see, you know, a Virtual Box, like actual Virtual Box app where like you run the VM, and then you have like a screen within the screen. You know, you can put it in the background, but then you can like jump in the screen, and like-

You know what? That's not a bad idea. Yeah.

You know? Like I, I mean, I used to... You know, people used to do it-

Mm

... virtualizing like-

Kali Linux-

Yeah

... in a Windows machine.

Yeah

And like, you would just jump in, and then you would jump out, but it's like, it's not like a dual boot. It's, like, within the thing. The problem is that you need twice the amount of RAM, twice the amount of...

You know, it's, like, it's kinda taxing on the machine. But I think that would be cool, kinda like see, you know, the little Claude window. I can see his desktop, look how cute it is, clicking around things.

I was gonna bring up, he's the original machine in a machine guy because he has the, uh, Windows, uh, Windows 95 project. Where's, where's the Windows 95 project at?

It's probably somewhere on my GitHub, right?

No, no, no, no, no. I, I... It's like, it's, it's the first thing you see is this one.

Nice.

Yeah, it's ...

Yeah.

Exactly.

That was honestly a very fun project though. Like, o- obviously I didn't... I should, I should say this just so that no one gets the wrong impression. I did not write the actual, the actual-- Obviously, I didn't build Windows 95 because I was a child.

But also, I did not build the actual engine that is capable of, like, simulating an x86 processor in JavaScript, in Wasm. Um, that's a tool called V86, which is very cool, and everyone should try. But this came out of a, this came out of, like, a debate we had at work where people were like, they often are in the inter-debating the merits of Electron and whether or not we should be building software in JavaScript, yes or no.

And I still am very upset that I can run all of Windows 95 in JavaScript and launch Microsoft Excel inside the virtualized JavaScript Windows 95 machine and do things that pro- I can do that entire chain faster than I can do a lot of other things in, like, traditional SaaS applications.

Mm-hmm.

Uh, this is sort of like a, like a performance rampage that I went on. So I mostly built this as a joke- ... for some of my colleagues at Slack. This took, took, like, one night, um-

What?

But then, did I... It was, it was not hard to do. It was... All the hard work is in V86.

Yeah.

Like, if you go to the repo, it's gonna say, like, 99% of this work is done by, by, um, a guy who goes after the, by the name Copy. His name is Fabian.

Yeah.

Um.

Cool. I think you're, uh, you're kinda back on the Windows grind because you're building out the Windows support. Uh, I thought there were some really cool technical stories to tell, uh, and it gives people an appreciation of, like, well, here's how hard it is, and here's-

Yeah

... how important, here's how, how you invested the sandbox.

Yeah.

So maybe this is, like, a good opportunity to talk about some of the details.

Oh, yeah. The, the VM honestly is, like, so cool. There's a lot of things we dislike about the VM, right? Like, there, there's a lot of things that are real trade-offs, and you want to know why you're making those trade-offs.

Um, and you're right, like, a lot of people write me like, "Hey, how, how come Claude is taking up 10 gigabytes?" I could say on the part it's not actually taking up 10 gigabytes, it's just, like, a way that macOS displays bytes as, like, wrong.

But the way we actually write it to disk is by we collapse the empty space in the image. So it's not actually taking up 10 gigs, but that's a technical differentiation that's probably not gonna matter to a lot of people.

Like, to me, the, the, the how come is it takes too long to start.

Yeah.

It's, like, 30 seconds sometimes. I, I don't know.

Oh, it should be faster than that.

Okay, whatever.

Right.

It's maybe 10, but it's, it feels like 30 .

Yeah. Like, even either way, like, whatever it is, it's going to be f- it's going to be slower than just running Claude Code directly on your computer, right? So the trade-offs are real. But what we're doing on Windows, we're using the Windows, Windows, uh, host compute system.

It's the same thing that WSL2 runs on, like the Windows subsystem for Linux, that I think a lot of developers appreciate quite a bit.

Yeah.

Um, and it's, it's pretty cool because we sort of, like, have to separate out which system space this virtual machine runs in, in who gets to talk to the virtual machine, because obviously you give this virtual machine a decent amount of power.

How do we optimize not just the connection between the two systems, but also how do we make sure that random other application doesn't get to talk to Claude inside the VM?

Mm.

We do some pretty interesting things. Um, last week we started writing a new networking service, a networking driver that optimizes how Claude talks to the internet. If your company is doing, like, weird internet things, like packet-

Yeah

... inspection and like t- like, you know, taking your part as a cell inside your company. I think there was probably, like, a very small easy version to build of Cowork that is much simpler, but also breaks on most com- most users' computers.

And this one is quite nice because it works on most users' computers. Um, and the default example I always go for is I, I really want this to be highly effective on like a, on like a machine that most people pick up, and that machine will probably not have Python, it will not have Node.js.

Mm-hmm.

And even if I just take away those two things, Claude is going to be so much less effective on your computer.

So what do you do? You don't even... I mean, may- maybe require people to install Node and Python.

Oh, like you mean for like a what is the future look like without a VM?

No, no, no. So, so like, like you say, right?

Yeah.

Let's say our target machine is whatever is a default spec Windows laptop.

We do this, which is quite cool. So on, on, uh, macOS, we use the, um, Apple Virtualization framework, which is pretty solidly optimized. Like, it's good stuff. And so-

It's like a simple API call, right?

Yeah, it's, like, super simple.

I, like, I saw the code recently, and I was like, "That's it? Uh, what the fuck?"

We do... You, once you start, like, shipping production code on it, you start adding, like, all of these edge cases you're gonna-

Oh, yeah

... it ends up being a little longer. But, um, I think Apple really cooked with the virtualization framework, and it's very, very good. It is very fast. It's very reliable. And same on Windows. The, the host compute system, I think WSL2 as well is maybe one of the diamonds for the Windows.

It's, like, one of the few things that developers universally rave about as very, very cool. And, like, hooking into the same subsystem makes it a lot easier for us to say, "We don't really care how locked down your computer is."

Maybe it's, like, your employer's computer and your employer has decided that you get to install nothing.

Mm-hmm.

Not trusted. But it's true in a lot of environments, right? Like, even at Anthropic, um, our IT department controls what kind of software you install. It's like a pretty common experience for many companies. Um, and this gives IT departments a decent amount of...

Like, it makes their job so much easier because we can say you can separate out Claude's computer from the user's computer, and then for Claude's computer, what you probably care about is data loss, you care about, like, a potentially hostile actor, you care about maybe data being exfiltrated.

Mm.

And once you control the network and the file system layer, you don't really care necessarily anymore that Claude might be writing super-

It's valid

... useful Python scripts. What worries you about the fact is that, like, once you install Python, now anyone can do anything on their computer But once you put that in a VM, that risk really goes down.

Yeah.

So that's why we jump through all of these hoops.

Yeah. I think you, you had a different, uh, tweet about this, um, but it, it's, it's almost like people have also approve exhaustion.

Mm.

Like, it's like y-you can't approve every single command. Like, uh, sometimes by, by default, some of the, the CLIs, y- I think e-even early Claude Code, uh, we have to approve every single command.

Yeah.

And, and, like, it's so... So there's this sort of dichotomy between either approve every step or dangerously skip permissions.

Yeah.

And actually sandboxing's, like, kind of, like, the middle ground.

Yeah, I do think, I do think it, it's maybe on us as, like, the industry to come up with something better than, "Oh, this is super safe as long as it doesn't do anything." Right?

Right.

If you want this to be useful- ... then you have to, like, approve every single step of the way. And, like, computer use is a good example. The only way to make computer use on your host, like, super safe, like really super safe, is probably if you approve every single action, right?

Like, model is like, "I would like to type the word L." And you're like, "Okay, that seems fine." 'Cause I know, I know which, like, cursor is focused.

Yeah, it's not automation if you don't delegate.

Yeah, exactly. You need to, like, probably delegate. You need to be able to, like, delegate and walk away and trust that this thing is not gonna, like, mess up dramatically. And I don't even think we need to build perfect systems.

I don't think we need to wait for, like, 100% model alignment. We can rely on the same Swiss cheese model we've used in the industry for a long time. But I do think we need to, like, universally, maybe eventually invest more, and that's what we're doing.

We need to invest more in systems where we can say, "You do not need to approve everything."

Speaking of Swiss cheese model, I just wrote a thing about this.

Oh, cool. Yeah.

Uh, yeah. Um, yeah, super cool. I mean, yeah. It, it's, it's weird how, like, I guess usually I think safety and security is kind of like a boring word to, to engineers. They're like, "Just keep me unsafe, keep me unsecure."

But, um, I think achieving the right thing, y- like you're going after a consumer/prosumer-

Yeah, yeah

... target audience.

Kind of like both. Kind of like both. I think I, I also want to capture people who would have no trouble using Claude Code, like yourself, right?

Yeah, yeah.

But still find it maybe just convenient, easier. You're like, "Oh, cool. There's like the to-do list on the right. I can edit it."

Yeah.

Those things are just easier to do if you have to.

Yeah, but this is, like, clearly the knowledge work side.

Yeah.

Claude Code will clearly capture the development workflow. But, like, I, I, I do think, like, you have to sweat this, like, safety and security details in order for people to trust it. And, like, y- the... even Claude and Chrome, like, having the...

whatever API uses to do the background thing-

Yeah

... um, that's the only reason I use it, is because otherwise I would have to just get a separate machine-

Yeah

... and just r-run it, run through all the tasks.

But that sounds super annoying.

Yeah. I mean, like currently doing it, but ...

And I think, I think also as developers, um, maybe we're, we are more risk tolerant, but we're also just, like, accepting... We are more risk tolerant, but I think we also just have, like, I don't wanna say arrogance, but, like, sort of the trust that if, like, the really bad thing happens, we can probably fix it.

I just tell Claude to, like, check with me before doing any irreversible action, like sending an email or doing it permanently.

Yeah.

It's good enough.

But, like, not even Claude. I mean, like simple things such as NPM install, right? Like we're all running NPM install with full user permissions, and if it wants to, like, read .SSH, it will. Crazy that that is the default.

It's fine.

Kind of... Yeah, I know. I agree. I agree it's fine. Like, I'm obviously doing it every single day.

Yeah.

No, right? Like, uh, and I think obviously NPM and GitHub too have, like, done a pretty good job maybe over the last couple of months to, like, clean house and come up with, like, more specific tokens. But generally speaking, I think as engineers we've always been a little bit more risk tolerant, and if you do a little bit of introspection and you ask yourself, "Is that how we should be doing things?"

You might not always come up with the right answer. And I think for models too, like my approach, like I'm not gonna... The, the safest thing is to do nothing. We do want products that are quite capable, but to the extent possible, I don't wanna ask you, "Are you okay with this script?"

Because I kind of believe that once it starts becoming a part of your workflow, you're probably not either, either you don't have the skill to understand whether or not this Python script is safe, or you're not gonna read it anyway.

Cool. I guess a, a couple parting questions. Uh, what's the future of Claude Cowork?

I think we're still, we're still such early days. We're gonna keep shipping things that, we're gonna keep shipping things that, um... We're gonna keep iterating on this thing, like, pretty quickly. By which I mean you can sort of continue to expect that every single week there's gonna be like a small new feature, if not a big new feature.

Um, I'm going to continue probably to double down on your computer and, like, making you effective on your computer and making Claude effective on your computer. Um, we're starting to grapple, as we talked about today, grapple more with the question of, like, what does it mean?

What does your computer mean? Does it have to be the one in front of you or like a VM on your computer or like a computer somewhere else? And then the third thing that I'm quite excited about is we're continuing to go up this hill climbing on slowly taking users who are used to asking questions and getting an answer, to slowly teaching them to, like, step more and more away, and let Claude take over, like, bigger and bigger tasks and work, both in time as well as in, like, scope.

And I think you can probably see most of the, our investments in our feature releases to, like, work on both of those things. Like, the ability to do more on your computer, and then the ability to do more independently and for longer.

Does remote control work for Claude Cowork yet? No, right?

Excellent question.

Coming soon. I mean, that's an obvious thing if you want to keep betting on the, on your computer. But I... To me, like, you know, we, we talk about, like, people are not ready this year. Like, the... There's, there's no wall, it's ex- it's accelerating.

To me, like, what will be, we be doing differently at the end of this year that, you know, we're maybe not even thinking about this, uh, at the start of this year, right? Like, I'm just trying to look ahead as to like what, what's like a good use case that you're, that we sort of aim towards.

So for, for example, for the machine learning scientists, it's always, "Okay, well, I want AI scientists that can automate, uh, automate machine learning." But, like, for, for knowledge work, I mean, I can already, you know, get it to sign up for Google Cloud to me as AGI.

Yeah.

'Cause Google Cloud's ... But, like, what, what is... what's beyond that? I don't know.

I think it's basically the idea that, like, you still have to tell it to build you a script, right? You were still kind of involved.

Yes.

In maybe a way that felt Kind of magical to you, but like maybe to me on the other side as the person building this product still feels kind of heavy-handed. I see so much process that I'm like, "Oh, let me take that away from you."

Okay.

Or like how do I just go-- I will continues to go or continue to go like further and further up the stack and make your life easier and easier.

Oh, here's one, right?

Yeah.

Watch. Uh, I, you know, I don't care about my own privacy or whatever, or I trust Cla- I trust Anthropic, so just watch everything I do on a normal day-to-day basis. At the end of the day, tell me what you...

It's called Coworkable.

Yeah.

I don't know.

I think the funny thing about a lot of these products is that, like, for good reason, I don't enjoy-- I, I don't... Throughout my entire career, I've never like teased too much what I'm working on because I think you should just like-

Yeah

... release it.

Yeah.

Build the thing-

Release it

... and release it and then talk about it. Like I'm, I'm not a big fan of the like vague posting my own work ahead of time.

Yeah.

But the thing that is like always so fascinating to me is like both of y'all multiple times today, you've like mentioned things and I'm like, yeah, that is obvious, like very obvious-

Okay

... that someone should be working on those things. Um, and I think we're still in the space where if you look at Cowork, the things that we will be releasing will probably not be a big surprise to either of you.

You're gonna be like, "Yeah, obviously that's valuable. Obviously they were working on those things."

Yeah, yeah.

And obviously that's good and useful. And the more I hit tho-those points, the more our features fit into that category, I think the better it is for us because then we don't end up building things that are too hyper-specialized or too difficult to understand.

Yeah, I think the hyper-specialized thing is very important.

Yeah.

It, it keeps you like general purpose. I mean, it means you're not thinking too small, maybe. I don't, I don't know what the, the word is.

Yeah, yeah, exactly. It's like the whole concept that like at no point did we release, you know, there's no Claude Code for Node.js applications that use React and 10-stack and only those two things and like if it's anything else.

I know several startups like that.

I think that's pretty-- Like, I'm not a VC, I'm not an investor. It's like hard for me to predict where the markets go. But in terms of the building blocks that I'm interested in, the Electron is probably by far the most popular thing I ever built.

And, um, Electron itself is like very abstractable and generalizable, right? Like so many apps run in it that I think it would've been hard for me to predict how many apps actually end up using Electron.

Yeah.

Um, and what would've been even less useful for me to predict is, in what those apps do. I distinctly remember Bloom coming out. I mean, that is cool. Like your camera in a little circle in the corner.

Mm.

That is pretty smart.

That's an Electron app.

Yeah.

Yeah.

Or at least was. I'm not sure if it still is, but it was for a while. Or like 1Password has so many interesting things, right? It, it's, it's, it's a level of the stacks that I'm quite comfortable with, and whenever I give other engineers advice, it's actually that layer that I think is most valuable to invest in because the tools of the layer are not that good, but that's where you get the most leverage-

Yeah

... for the future in general.

Just quick tangent on Electron 'cause I always wonder this. Uh, have you looked at Tauri?

I have, yeah.

What's your take? Uh, you know, my, my, my, my view is like most things should be Tauri by default unless you really need the full power of Electron, but...

Yeah, I can give like my take on... I can give my big take. Why do we ship an entire version of Chromium inside the thing, right? Like, why do we do that? And, um, people ask me this question a lot because it's like very counterintuitive.

Wouldn't it be much easier to use the web views that are on the operating system? Wouldn't it be much easier not to have to do that? And the answer is yes. And like, obviously, I did that. Once upon a time, I did that.

There was a version of the Slack app that used just the operating system web views.

Wait, did you, did you start the Slack app?

I w- Well, a team effort and yeah. But I was, I was there. We built the Slack app.

Yeah. That's crazy.

Um-

I, I mean, obviously you get the Electron guy to do it, but

Well, but this is an interesting point. Like, by the time, by the time I joined Slack, they already had an app that was built with something at the time called Mac App. It was a little bit like the same app gap thing for mobile.

It just used the operating system's web views. Um, and that didn't work for like so many reasons. Um, and they were like, "All right, maybe we need like bigger guns. We need to like take more control of the rendering stack."

And there's, there's a few things I always mention here. Um, I think if you're building a small app, just going with the operating system's web view is perfectly fine. If you're building an app maybe that doesn't have too many users who will like cry bloody murder if it doesn't work, that is fine.

The reason to go with your own embedded rendering engine is because, and this is still true in twenty twenty-six, the operating system rendering engines are not that good. They're just not that good. Both Microsoft and Apple are trying to move away from that.

They so far really haven't. The only way to upgrade those is to upgrade your operating system. So if you're, say, Slack, and you have a critical rendering bug in WkWebView and in some of the other WebView options, your only recourse is to tell your customer, "Oh, sorry, you're too poor.

You didn't buy the latest MacBook." Unacceptable.

Mm.

Unacceptable to user, unacceptable to user developer. So you sort of need to like go down the stack and like find the best rendering engine, then put it in your app. Why Chromium, even though it's very big? Chromium is by far the best thing.

Like I, I often like to remind people the Unreal Engine, you wanna render some text, they use Chromium.

Hmm.

Like Chromium is part of the Unreal Engine for the same purposes. Chromium is very, very good. I think it's like one of the marvels of engineering. It's very hard from... We're in San Francisco right now. This is where we're recording.

Most of the people in the city are web developers. It's hard for me to like overstate how magical it is that you can run C-- like rendering a YouTube video, dynamically negotiating a bitrate, figuring out what to do about your extremely broken hardware driver.

Actually, this is a fun thing. Um-

Okay

... you can enter Chrome://gpu.

Okay.

And if you scroll down a little bit,

these are all the enabled workarounds because something is going wrong on your computer. If you're doing this on a Windows computer with like a GPU that is not the most popular GPU, it will be much longer. And all of these are usually just there to make sure that if I say as a developer, I want a red pixel to appear here, that that actually happens.

Chrome is such a marvel because it works on all the machines that a user might throw at you, and it's gonna work fairly reliably, and if it doesn't, they will probably fix it within twenty-four hours

I see. So this is the super operating system, right? That, that works everywhere.

Yeah.

All right. Okay. Yeah.

So a lot of the magic of Electron is honestly just that it makes it very easy for you to ship Chromium in a way that serves you exactly in your use case with Electron, uh, exactly.

Our next interview is with Morgan Jesen-

Yeah

... who had the phrase, like, "Desktop OSs are just poorly de-- uh, poor i-i-implications of the, the actual OS, which is Chrome," which, like, actually works everywhere. And this is, this is the platform where you ship apps.

I, I think the wild thing is that, like, as engineers, we so often sort of assume that the platform, like the layer below us, is, like, super stable.

Mm-hmm.

And then you talk to those people, and they're like, "Yeah, we're also just, like, guessing." Um, uh, and I had, like, a distinct moment at Slack where one of our customers at Slack was Nvidia, and for a while I really put GPU developers on this pedestal in my head.

And I do think they're still probably much smarter than I am, but I was like, "Hardware engineers who built the chips, who then, like, built the drivers, their work must be so much harder than mine. They must be very good."

And we had, like, one bug in Slack where, like, if you had a YouTube video in Slack, it wouldn't quite render right. Like, it would have this weird artifact. And, um, that ended up being a Chromium bug, and I ended up on this, like, giant thread.

So I got to see a lot of the source code, and they also are just like, "Common to do. We don't know why this is weird, but if you flip this bit, things work." You know? This is just, like, happening at every layer of the stack.

Maybe the, uh, you know, the, the end-of-year AGI prediction is that Claude can build Chromium. You see, you see, you, you laugh now-

Yeah

... but, like, you know, someday.

It's, it's starting to get pretty good. Like, it used to be completely useless, um, mostly just, like, overwhelmed both with how hyper-specialized tools are inside the Chromium repo, like, for, for a long time. Like, Chrome has to, like, sort of reinvent all the tools because none of them are capable of handling Chrome.

I think the AGI moment I'm kind of waiting for is, at what point are we gonna say Electron is probably no longer necessary because you can just build fully native apps?

The Swifty?

Yeah. Like, not just in Swift, because this is one thing. Like, it's pretty easy if you-- I think our current models are quite capable of taking an Electron app and replicating it Swift. Are they gonna be capable of, like, building an app that is actually more performant, uses less memory, all of that stuff?

Um, it's gonna go into the same hyper-optimization that developers have done for, like, a long time. We're not quite there yet where I can, like, point even our best models at a thing and say, "Just replicate this in native code.

Make no mistakes. Ultrathink." Right? We're not quite there yet. Um-

Ultrathink is back.

Today Ultrathink is back. Yes.

Okay. Or we'll give Ultrathink for, like, days, which is a pretty long time for Borb.

But he worked on Ultrathink for days?

Yeah.

Why? It's just, it's just a front.

I'll let it live.

The Lord goes into it. Yeah. Okay.

Another question I had is, like,coworks. So if I have my Claude Cowork, like, what's kinda like the multiplayer mode? I think sub-agents is like single players split up the context.

Yeah.

And the multiplayer cowork is like my colleague has some file on their machine that I wanna know about-

Yeah

... or I wanna know how their task is going to then update my thing. Like, is that interesting? Is that something that makes sense for you to build or for, like-

It's, like, super interesting to me. It, it almost goes back to, like, some of the scaffolding where I'm like, "Okay, are we gonna be end up-- Are we-- Will we end up building scaffolding that will just go away?"

And, like, a question I have here is, at what point do we just assign these things, like, their own Gmail account? We just give them their, like, Slack handle, and then they will just, like, use the same tools we humans use to interact with each other.

You mentioned our finance people. They've been working pretty hard on very good office integrations, and I think for a while we've been like, we built so much tech around Claude leaving useful comments inside a Google Doc, and now it just does it, just, like, leaves a comment in a Google Doc, and that's how you interact with it.

Maybe, like, the similar thing where I still have open questions around what is the best interaction mode. Is it for us to build something super custom for cowork agents to talk to each other? Or is it, okay, let's just jump straight to the finish line and say, "Well, we're just gonna give this thing...

If you use Slack at work, we're just gonna give this thing a Slack handle, and that's going to be the way it's, like, multiplayer capable."

They communicate with each other. Yeah.

Yeah.

Like, you know, as a, as a fun project, I built this thing called PyQ-

Mm-hmm

... which basically takes any repo and the Py agent, uh, coding agent, it puts it in a VPS, and then there's a public web hook-

Right

... where anybody can submit a coding task.

Oh, okay.

And then there's a dashboard in which you review the task, and then there's-

Which one? PyQ?

Py, P-I-

Uh-huh

... Q.

Yeah, you basically get all these, like, tasks. Anybody can submit a task.

Mm-hmm.

And to me, it's almost like in the organization of the future, it's like the salespeople are talking to the engineering team that is talking to the marketing team, to the product team, and all these cowork are going to, like, queue up decisions for other people to approve in a way.

Yeah.

You know? And I'm kinda curious what that looks like and, like, how do you, how do I give my cowork the ability to both approve tasks without asking me-

Yeah

... and how to decide which one I need to review.

Yeah.

You know? Because for some of these things, it's like, you know, you wanna change the color of something that's kinda like a branding decision, or another one is like, "Hey, your thing is just broken." It's like, "This is, like, how you fix it."

Yeah.

And Claude can actually review whether or not that prompt matches what it's trying to do. Today, everything is still very... It's like multiplayer within the single player, you know?

Yeah.

I can spin up many of them, but, like, how do I get multiple people to hand off to each other things using their particular context?

Yeah, and for both of yourcoworks to, like, talk to each other, right?

Right.

It's like-

Yeah, "Hey, we got an episode today. Can you, like... Have you..." You know? Or...

Yeah. This is like, uh, uh... I know we're, like, running out of time here, but, like, we, we previously talked about sharing skills, and I did have this question of, like, what if your cowork would just, like, ask the othercoworks if they have a skill for this task?

There's another thing you could do, right? Like-

Okay. So skill transfer.

Yeah, like, um, and again, this maybe-

Creepy This maybe goes back into the territory of like building something very powerful and building something creepy often goes hand-in-hand, um, because I could tell from the reaction that my fellow engineers had that this is probably not what we're gonna do.

But like we have Bluetooth LE, right? Like I-- this computer can figure out that it's sitting right next to this computer-

Right

... so you're probably working on the same thing. Um, will you see that in Cowork? Probably not. But, um, there's like, I think, really creative solutions to problems that we really haven't tried yet.

Yeah, yeah, yeah.

Excellent. I guess the, the last thing is, uh, Anthro-Anthropic Labs. Uh, I always have this mental model of a model lab versus a agent lab, and this is basically Anthropic's internal agent lab, which Co-Claude Code, uh, is now under, right?

It's part of the whole org.

I mean, people are so fungible, right? Like-

Okay. This is just-

Yeah.

I, I don't know how, I don't know how real this is. I don't know.

No, it's a real team. It's a real team.

Yeah, yeah.

Um, the, the labs team is primarily working though on things that you don't see in public yet.

Yeah.

Um, they're trying like really wild, out-there ideas that seem quite improbable. Um-

The mad science team. But you, you're-- are you officially under this thing or?

No, we're-

Okay

... we're sort of-

Claude Code is but, uh, now Claude Code is like a fairly big group where... I actually know how many people we are. Like I remember yesterday coming into our weekly cohort meeting, and I was like, "Whoa, this is, uh there's a lot of people here."

Um, but we still have a labs team, and we actually made the labs team a lot bigger. Mike just joined the labs team as a, as an IC, which I think is very cool and very fun. But they're, they're working on things that you have not seen yet that are extremely out there and probably half broken, right?

Like the sort of the idea of a labs team is that it should only work on things that make really no sense for anyone else to work on.

Okay. Well, looking for exciting things from there. But thank you so much. I know we're out of time, but, uh, appreciate your joining us.

Yeah.

I appreciate Claude Cowork. Everyone go use it. Uh, it is the closest I've felt to AGI this year.

That's so nice of you to say. Thank you very much.

Yeah, thank you for your time.

Yeah.