⚡ Inside GitHub’s AI Revolution: Jared Palmer Reveals Agent HQ & The Future of Coding Agents

Jared Palmer, SVP at GitHub and VP of CoreAI at Microsoft, joins Latent Space for an in-depth look at the evolution of coding agents and modern developer tools. Recently joining after leading AI initiatives at Vercel, Palmer shares firsthand insights from behind the scenes at GitHub Universe, including the launch of Agent HQ which is a new collaboration hub for coding agents and developers. This episode traces Palmer’s journey from building Copilot inspired tools to pioneering the focused Next.js coding agent, v0, and explores how platform constraints fostered rapid experimentation and a breakout success in AI-powered frontend development. Palmer explains the unique advantages of GitHub’s massive developer network, the challenges of scaling agent-based workflows, and why integrating seamless AI into developer experiences is now a top priority for both Microsoft and GitHub. Full Video Episode Timestamps 00:00:00 https://www.youtube.com/watch?v=ZWEOX610WEY Introduction and Jared's New Role at GitHub00:01:00 https://www.youtube.com/watch?v=ZWEOX610WEY&t=60s From V0 to Agent HQ: The Evolution of Coding Agents00:02:51 https://www.youtube.com/watch?v=ZWEOX610WEY&t=171s The V0 Origin Story: From ChatGPT to AI Playground00:05:40 https://www.youtube.com/watch?v=ZWEOX610WEY&t=340s Building the AI SDK and ShadCN Collaboration00:07:08 https://www.youtube.com/watch?v=ZWEOX610WEY&t=428s The Birth of V0: Prompt to UI Revolution00:09:18 https://www.youtube.com/watch?v=ZWEOX610WEY&t=558s V0's Growth Journey and Model Evolution00:11:05 https://www.youtube.com/watch?v=ZWEOX610WEY&t=665s Model Strategy: Composite Models vs User Choice00:13:16 https://www.youtube.com/watch?v=ZWEOX610WEY&t=796s GitHub's Agent HQ and Model Marketplace00:15:51 https://www.youtube.com/watch?v=ZWEOX610WEY&t=951s The Future of Agent Abstraction and Standards00:16:33 https://www.youtube.com/watch?v=ZWEOX610WEY&t=993s Microsoft Core AI Integration and Workflow Vision00:18:37 https://www.youtube.com/watch?v=ZWEOX610WEY&t=1117s Dev Containers and Repo Setup Challenges00:24:10 https://www.youtube.com/watch?v=ZWEOX610WEY&t=1450s Agent Quality and Infrastructure Reliability00:27:05 https://www.youtube.com/watch?v=ZWEOX610WEY&t=1625s Using Coding Agents for Non-Coding Tasks00:29:11 https://www.youtube.com/watch?v=ZWEOX610WEY&t=1751s GitHub Homepage Redesign and Community Feedback00:30:27 https://www.youtube.com/watch?v=ZWEOX610WEY&t=1827s Stacked Diffs: GitHub's Most Requested Feature This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.latent.space/subscribe https://www.latent.space/subscribe?utm_medium=podcast&utm_campaign=CTA_2

• 0:04Speaker 00

  All right, we are here for a very special edition of Lanespace with my buddy, Jared Palmer, SVP at GitHub and VP at Core AI at Microsoft.

• 0:13Speaker 01

  Correct, dual title.

• 0:14Speaker 00

  Yeah.

• 0:14Speaker 01

  Twice the fun.

• 0:15Speaker 00

  Dude, we're to have two jobs?

• 0:16Speaker 01

  No. I'm only on, I'm only, full disclaimer, I'm only on day 13, I think.

• 0:22Speaker 00

  Yeah.

• 0:22Speaker 01

  So, early days, so, so far, so good.

• 0:25Speaker 00

  So far, so good. We've been trying to get you on the podcast for two years, I think.

• 0:28Speaker 01

  I think so, yeah.

• 0:29Speaker 00

  You, yeah, you're a busy guy.

• 0:32Speaker 01

  We don't do it in person, so.

• 0:33Speaker 00

  We have to do it in person, yeah, exactly.

• 0:35Speaker 01

  Way better.

• 0:36Speaker 00

  I should also plug that you have, if Jared Palmer fans should dig into your previous podcasts.

• 0:41Speaker 01

  With Ken Ruler. Don't let that.

• 0:43Speaker 00

  It's come from the internet. Okay, so.

• 0:46Speaker 01

  What do you think of Ken?

• 0:48Speaker 00

  Shout out to Ken. Before that, you were building, I guess, like, V0 and AI SDK. And you were just sort of VP of AI at?

• 0:56Speaker 01

  At Versale, yes. Yes, all AI initiatives and vibes.

• 0:59Speaker 00

  Yeah, and I feel like, basically, you went from, like, you sort of building one coding agent to now being the home for all coding agents. Is that, is that like the general vibe of Agent HQ?

• 1:08Speaker 01

  I think that's right, yeah. So, backing up, I spent the last sort of two years or so building V0 at Versale and AI SDK, and then the summer took time off and now joined GitHub. And today, we launched Agent HQ, among other things, here at Universe. And yeah, it's gonna be the home, we hope, of not only agents, but also developers. And it seems like the gravity well of this new collaboration space that we're trying to build.

• 1:35Speaker 00

  Yeah. What do you think, like, basically, that GitHub can do that you couldn't do at V0?

• 1:42Speaker 01

  GitHub is an enormous platform, right? So, these are.

• 1:44Speaker 00

  Under 80 million.

• 1:45Speaker 01

  It's something about, yeah, it's under 80 million developers. It's just, the scale is immense, right? And V0 was focused on not only one language, but one framework, right? And a specific problem space with a built-in renderer and, you know. You know, for those who are not aware of V0, it's, it's like a Bolt or Loveable, but it's built by Versale and it's focused on building NextJS apps, specifically NextJS apps. That constraint was rather liberating for the team at the time and it lets us really, like, laser focus, right?

• 2:17Speaker 00

  That's my world at it.

• 2:18Speaker 01

  I hope, thank you. I hope so. And obviously at GitHub, you know, we're the home of all languages and, you know, and frameworks and developers. And so the scope is broadened. And yeah, it's just a different, a different part of the map, if you will, right?

• 2:33Speaker 00

  Yeah. So you've been basically covering the entire journey of coding agents.

• 2:38Speaker 01

  Yeah.

• 2:38Speaker 00

  For, you know, from, from the start, like, what do you think, what's your personal journey through coding agents, right?

• 2:43Speaker 01

  Sure.

• 2:44Speaker 00

  We started out with Copilot, obviously GitHub started the Copilot trend.

• 2:47Speaker 01

  Sure, sure.

• 2:48Speaker 00

  When, tell us about the origin story of V0.

• 2:51Speaker 01

  Yeah.

• 2:52Speaker 00

  And then how that develops and maybe where, like what you want to see next with history.

• 2:58Speaker 01

  It's funny you ask that. As I've told this story multiple times, I feel like I've unlocked different parts of it in my brain by going back, you know, like, so maybe we'll have to figure out how retrieval memory works.

• 3:09Speaker 00

  By the way, interesting how memory works for agents.

• 3:11Speaker 01

  Totally. That's why I brought it up. Like kind of thing as you, sometimes you discover new paths, right?

• 3:15Speaker 00

  Yeah.

• 3:15Speaker 01

  Anyway, the story goes like this. So when ChatGPT first came out, obviously it was incredible, right? Like world changing, immediately fastest growing product ever. I look back at like the timeline and dates and we were very early, like in, when I was at Purcell, jumping into AI stuff. But the journey kind of went like this. So at the time, actually there was no AI division, there was no AI group. I was actually the director of engineering for all of our cell frameworks and I was helping NextJS spell, uh, spell great. Yeah. So NextJS spells, uh, Turbo Repo, Turbo Pack, Webpack and all internal dev tools at Forcel. And I was helping the NextJS team dog food and test the initial implementation of server actions.

• 4:00Speaker 01

  And instead of building a to do app, Kisharamo, the CEO of Forcel was like, why don't you build like a playground? And I was like, okay, cool. So that led to the AI playground, which is now just part of AI SDK. We'll get there in a second.

• 4:13Speaker 00

  Which, by the way, iconic for like side by side, but also the

• 4:16Speaker 01

  Right, so Nat Dev. Nat Dev. So G told me that Nat Dev, I got a DM, it was like, I remember because I was at a bachelor party, and Giserramo, internally online, sends me an email like Nat dev is launching on Monday, you have to ship. And I had been working on it previously.

• 4:31Speaker 00

  So we had the same idea.

• 4:32Speaker 01

  Yeah, so he got wind of it, I guess. So I definitely had to like jump into motion. And I didn't think, we didn't even ship chat first. Giserramo sends me this DM over the weekend, I'm at a bachelor party, and he's like Nat dev, this side by side, he sends me the link, and I play with it, I was like, shaw. Okay, so I spring into gear, ship the AI Playground. What was cool about the AI Playground was it forced me to go through every single model providers API docs, and figure out their quirks of their nuanced streaming. Because at the time, it wasn't like everybody used OpenAI. It was like all little quirks, some of them kind of were compatible. So that was my first foray at it. And then, launched AI Playground, that shot to the top of Hacker News. And I remember, I didn't even implement chat, because that chat wasn't actually like, wasn't as important, it was just like, confessions.

• 5:15Speaker 01

  So eventually, we backed to chat, and that project, out of that came AI SDK, because I had already looked at all the model providers, and all of the the combinations, and I was like, okay, here's that chunk of streaming code you need. And then AI SDK found that sort of niche of like, how do we focus on the part that we're going to be good at, which is like that UI aspect of it, but then also knock it in your way. So that, we shipped AI Playground, then AI SDK launched. And then, you know, we're always about demos and having great starter templates at Vercel. And I remember writing Giserramo, I was like, you know what would be cool? This guy, Shad CN, oh man, he seems like amazing, and his UI library is doing great.

• 5:57Speaker 01

  Why don't we team up and ship a ChatGPT clone, open source? And we did. We shipped this awesome template, which is now called Chat SDK, but it's great. And what that did, though, at Vercel, was it set us up for like rapid experimentation, because we had this really good, like, pretty full-featured ChatGPT, ready to rock, with all the latest features. So when it came to like rapid prototyping that summer, now we're summer 23, it was so great, it was like liberating. So I remember, at that point, I had gained some momentum internally, and pivoted almost entirely to AI. And I had Shu Ding, who you're friends with, and Max Later.

• 6:41Speaker 01

  And Shad C N now, we're cooking. And I think at that point, code execution had just come out. I think that's my timeline.

• 6:50Speaker 00

  Yeah, they called sandbox interpreter.

• 6:51Speaker 01

  Inter code interpreter, that's what they call it at the time. And I had a very, as soon as I saw this, I had a very ambitious idea and proposal to present to Gigermo. Which was like, what if there was some, mind you, tool calls don't exist. The context windows, 4,000 tokens, you know, like, there's not much here. Um, what if we had this thing where like, you could prompt? And sometimes it would decode interpretation. And then maybe we could sort of, sometimes we would decode up interpretation. But then other times it would choose to render like UI. Or then it would render sometimes like a document or.

• 7:25Speaker 00

  In line in the chat.

• 7:26Speaker 01

  Yeah, and like, it would just have different sort of render mode.

• 7:28Speaker 00

  Generative UI.

• 7:29Speaker 01

  Yeah, and maybe you could pipe them together. So like, the output of one could pipe into, so if we did code interpretation, we course it's always emit uh like tabular data. Maybe we could pass that to another prompt that would like a UI. And just like some idea there, it got crazy. But if it sounds like, these are just tool calls, that's exactly what these are.

• 7:47Speaker 00

  That really was, yeah.

• 7:48Speaker 01

  Um, so it became pretty obvious that like, the, so I came to Gigermo and we, at the time, we just had a sec- secure, we had a sort of security debate, like, should we? Code interpretation with the ability to fetch data, um, like giving it internet access was like, kind of. Now, they're like, fine, whatever, whatever, you, whatever you want, wow, wow, Wes. But at the time, it was like a little scary. So we kind of said, okay, no to the code interpretation. But this UI thing, it's pretty neat. And so that, this like prompt to UI, that was like the aha moment of V zero. But the models were not very good, right? Or relative to where they are now.

• 8:25Speaker 00

  And it was uh, for G P T four era?

• 8:27Speaker 01

  This is just into the G P T four era. And now we're probably at a uh, 16,000 token context window. So you can't really do chat. So we had to kind of invent this kind of new paradigm of like, fake it with completion. But that forced us to do, sort of the click the, well, the, the initial V zero, which launched, I think in September 23.

• 8:50Speaker 01

  It would look more mid-journey. In fact, if you look at the original tweet, it was like mid-journey for React because it was all very visual and you could click on different components and elements and re-prompt. But it was, again, we were kind of hacking this because we didn't have chat and we didn't have tool calls.

• 9:05Speaker 01

  And then fast forward again, you know, that launches and then probably like nine months later,

• 9:13Speaker 01

  it took us like nine months to get to like a million ARR, this little team. But then the models progressed and from, you know, GPT-4, GPT-4-32K, the big boy. We never really got GPT-4 Turbo working. I don't know why. It just never happened. And then switched to other frontier models and then started doing our own models and stuff like that. But fast forward another 10 months or nine months or so and then we rebased towards chat. And now the models finally could do chat. And the artifact pattern had evolved, so it was time to rewrite. When we launched V0, the chat version, or the new V0, whatever we call it,

• 9:52Speaker 01

  it's like 14 days, another million MRR, 14 days, another million MRR. It was like a rocket ship after that. And that just proceeded and we just kept cooking. And so that's been the journey. We just kept perfecting. And what was really liberating for us was actually the focus on just one stack or one framework. Whenever everybody else was trying to do general purpose coding agent, we were like, no, we're just going to focus on NextJS, front-end and chat CNN. And that was really, that allowed the team to focus. So that's the, that's the story arc.

• 10:21Speaker 00

  I mean, to be fair, like Bolt, Loveball, like because NextJS is so dominant, basically everyone has to be good at NextJS.

• 10:27Speaker 01

  Right.

• 10:27Speaker 00

  But being focused on like even right down to the UI library and component stuff like that, that actually helps a lot.

• 10:33Speaker 01

  We also started working with all the frontier model labs to help because it was in our, Versal's best interest to have them be great at NextJS.

• 10:40Speaker 00

  Yeah.

• 10:40Speaker 01

  And also because of the post-trained models and you can read about this on the Versal blog, like we can,

• 10:46Speaker 01

  The post-training harness that we created, we started sharing and stuff with other model labs and stuff like that. And we had all our data in a very hygienic state to work with them.

• 10:56Speaker 00

  So. Did you have a debate internally? And because, you know, I, my, for my seat at Cognition, I can also see this, where you should pick the best qualities of every model and string them together in the V0, or you have the model selector and you let customers choose.

• 11:12Speaker 01

  We went back and forth, and I think, you know, we launched, we went back and forth on all this. I think at the end of the day, there's pros and cons.

• 11:20Speaker 00

  Yeah.

• 11:20Speaker 01

  One of the benefits of having your own branded model, or synthetic or composite, is that you can stitch these things together.

• 11:28Speaker 00

  Yeah, there's like.

• 11:29Speaker 01

  At higher levels of, and now it's a little different with the Sagentic flow, but even look at what you, like what you guys launched recently with ZrX Rep, right? So, search, is it going to be a different model than what generation? But, you know, the Genesis, but like, because it's a sort of. But like, search and Genesis are two different, like, like entire subsystems, right? So you can have search evals that are going to be totally different. And so how do you, so where we ended up now, where are we? Where we've probably ended up now is like, for a long time, we didn't have model selector. And then we had our own models, which were composites, which we talked about. And then like, and that would allow us to, you know, mix match.

• 12:10Speaker 01

  And I think that's probably if what, it's also nice because you, as a, this is like the product cap. You get to brand it.

• 12:18Speaker 00

  Yes.

• 12:18Speaker 01

  Right? And you can decouple it from the launch of the Frontier Lab.

• 12:22Speaker 00

  Yes.

• 12:23Speaker 01

  How do you guys, how does Cognition even bill for it? Like.

• 12:25Speaker 00

  They use.

• 12:26Speaker 01

  Right, it's a synthetic unit, right? So it gets a little wonky.

• 12:30Speaker 00

  Yeah.

• 12:31Speaker 01

  We can go on for pricing, this stuff, it gets challenging. But the nice thing about having the like, brand name model is that like, you get to co-launch with the provider and they'll hype you up. So, but your billing needs to then is capped at whatever retail is, right? Or some.

• 12:49Speaker 00

  Right, right, right. You can't really charge too much of a premium. You can, but.

• 12:54Speaker 01

  Right, but if you look at my. Yeah, I key, right? And it's like, well. Then how do we charge you for Sweet Rep or something?

• 13:00Speaker 00

  And so like, I think it's some part of it is the cynical, like you want to create a sustainable business and independence from the model labs. But the other part is genuinely, you actually do get better performance. You like string together all these things.

• 13:11Speaker 01

  Yes, and so it's tough. I think, uh, well, we've switched gears to GitHub, like we are all about model choice now. And like making sure that, uh, and what's cool is that like it's, we also have Copilot, which is our harness. And Copilot CLI, but we also have third-party harnesses like Claude Code and Codex. And cognition now like in A agent HQ. So you kind of get the best of both worlds. And like, I think that's going to be awesome and ultimately what people want.

• 13:36Speaker 00

  Yeah, I think also the model layer is not the right abstraction to do the switcher anymore. It's which is weird because that's where you started with the ISDK.

• 13:45Speaker 01

  Yeah, exactly.

• 13:46Speaker 00

  But now it's like the model and the agent have to be strictly tied together, like very, very strongly balanced. You can't loosely bound it and just do a generic interface because then you're just going to have the lowest common denominator of all the models.

• 13:59Speaker 01

  If you're in agent world.

• 14:00Speaker 00

  Yeah.

• 14:00Speaker 01

  Which may just be better than chat world, like in general, like better.

• 14:04Speaker 00

  Agent world is a much better abstraction.

• 14:06Speaker 01

  I'm calling it agent world, but I mean by like a loop with maybe compute runtime and like files.

• 14:11Speaker 00

  So is that your definition of agent? You're dropping your official definition here?

• 14:14Speaker 01

  No, don't put me on the top, but um, maybe, no, so my initial definition of agent for AIS, because like I actually fucked. I was dying on this hill because AISDK, everyone else is an agent framework. And I think maybe they, they actually went to like this, I don't know what it says on the front page now, but like, who knows? But like, an agent is, uh, you know, an agent is, uh, orchestrating, uh, you know, an API request with a queue and a for loop. Okay, but a coding agent now has meant so much more. There's like these, you know, coding, coding agent SDKs. And you've got sandboxing and file systems and tool calls. And, and I do think that is a unique, we, I'll call that agent world and charge our coding agents here. And yeah, I think that that seems to be where things are going.

• 14:55Speaker 01

  And even I believe the, the Claude Excel agent is basically, I was talking to Mike Krieger backstage, like, I think it's like related to Claude Code.

• 15:06Speaker 00

  Uh, it could be. I actually don't know how to implement it under a hood.

• 15:09Speaker 01

  Yeah, it wouldn't surprise me if it was.

• 15:11Speaker 00

  Yeah, they seem very all in on skills, which is kind of like an interesting.

• 15:15Speaker 01

  What do you think of skills?

• 15:16Speaker 00

  Uh, it's kind of D X T, which is like the sort of bundled version of M C Ps.

• 15:23Speaker 01

  Okay.

• 15:24Speaker 00

  Uh, wasn't, yeah, wasn't the, the reason you don't know about it is it wasn't very popular.

• 15:28Speaker 01

  Okay.

• 15:28Speaker 00

  Uh, so skills is kind of like the, the, the second shot that is very L L M pills. Like, it's like, just read my markdown and just read this, this directory of files and go nuts. Uh, as long as I can understand that you have the capability to run code, to read files, you're good. Uh, and then actually that, that is the universal interface, which is a file system.

• 15:47Speaker 01

  Right. Back to agent.

• 15:49Speaker 00

  Yeah, which is, which is kind of cool. Uh, yeah. So I mean, like, I, I think like what you're hitting at is, uh, you know, this philosophy of like our understanding of what coding agents, the, like, the minimum bar is over the last two years.

• 16:02Speaker 01

  Yeah.

• 16:02Speaker 00

  Right? Like you've, you've lived this journey and like now you're basically kind of like the kingmaker or like the.

• 16:08Speaker 01

  I don't know about that.

• 16:09Speaker 00

  You, you run, you run, you run agent H Q and, and I, I mean, I, I imagine you have other projects too, but like, it just is like the big one that we're talking about here. Um, like, what are you seeing from like the different agents? Like, what do you want to, this to become?

• 16:24Speaker 01

  Such a good question. Um, I think that agent H Q and GitHub itself needs to co-evolve. And, you know, one of the things that Microsoft has done really well is by putting things that are alike closer together. And so you think about the new core AI organization.

• 16:42Speaker 00

  Yeah.

• 16:42Speaker 01

  Got Visual Studio.

• 16:44Speaker 00

  Okay.

• 16:44Speaker 01

  Visual Studio Code, GitHub and parts of Azure all in one. Um, and obviously the GitHub team and the VS Code team have been working closely together for a long time, but now we're really close together. And I think for me, one of the cooler things that, um, agent H Q can sort of offer is this seamlessness, this fluidity with your workflow, right? So if you saw in the demo today, we saw a, um, a demonstration of, uh, you use agent H Q, you fire off a task and it creates a P R, but you can also open that P R up in VS Code in one click. And that's, that's awesome. And I think. My, the, the vision for, for, for GitHub as it evolves is to look at those touchpoints

• 17:27Speaker 01

  where AI can be sprinkled in, you know, Salt Bay style, into the native workflow, whether you're assigning an issue or maybe some new stuff that I think we should like focus on, maybe it could be like, how do we resolve a merge conflict?

• 17:40Speaker 00

  Oh my god.

• 17:41Speaker 01

  Right? Like, how do we maybe pop open an action or like, get in, right? And so it's like-

• 17:47Speaker 00

  I think Zalbee merge is my definition of AGI.

• 17:50Speaker 01

  Totally, but like, you get that error on an action and you're like, we've all been in that, we've all been in that sort of flow where like, actions kind of don't work locally or is that tool act? If you are trying to, I don't have it set up on my machine, I haven't done this in a while. So you're pushing up and you're, you're this like, okay, what if we could just put, like, you know, comment or kick off a task, solve this for you or, there are things there. I think, or I'm trying to describe is this like, this workflow where it's just like seamless and fluid and you can stay in a flow state across, whether you're across, like across all devices, mobile, web on GitHub.com or in your local editor. And I think that's where, you know, my focus is going to be in the next six months or so.

• 18:29Speaker 00

  Yeah, yeah. Just a side tangent on this, so, one of the things that Microsoft also owns, I don't know if it's Microsoft or GitHub, is Dev Containers. And I think like, a very important concept for sandboxing environment, whatever you call it, it is kind of a light version of what Docker containers are, kind of.

• 18:50Speaker 01

  Right.

• 18:51Speaker 00

  Do you see that as a standard that we should invest in, as like a, it's like a thing, like, because it's supported in VS Code? I don't think it's just that popular outside of VS Code.

• 19:00Speaker 01

  Yeah, it's used internally at GitHub too, for like, development at GitHub.

• 19:04Speaker 00

  Oh, yeah, yeah.

• 19:05Speaker 01

  Which is cool. Yeah, I think they were so far ahead almost, like, it was, but now there's, like, sandboxes, there's so many these days, right? So I think PowerFlare just launched theirs, there's Daytona there here, Purcell, Modal, which I think Lovable uses.

• 19:20Speaker 00

  I'm not, yeah.

• 19:21Speaker 01

  I mean, you probably have your own? I don't know, what do you, what do you guys?

• 19:25Speaker 00

  Uh, just some community pods.

• 19:27Speaker 01

  Okay, you guys are rolling yourself, that's cool. I think that's maybe the runtime, but, but- There's work and discussion about what that runtime should be even internally at Microsoft. We've got a couple of different competing things, so we'll figure it out in the next, you know, in the next cycle here. But there's a great point, like there's a lot of cool stuff that is in a dev container. You've got VS Code loaded, you've got a file system, you've got a sandbox, you've got the security protocol.

• 19:50Speaker 00

  Yeah.

• 19:50Speaker 01

  Also like wired into Git of enterprise.

• 19:52Speaker 00

  Yeah.

• 19:53Speaker 01

  And like ready to be packaged. So there's lots of goodness there.

• 19:56Speaker 00

  Yeah, I see like the number one pain points that combination has, but also Codex, also presumably the other guys is repo setup. Which is effectively what dev containers and a Docker file does for you is like run this thing, then that thing, set this up, do that thing. Why is it so hard? Like, why haven't we solved it?

• 20:15Speaker 01

  I don't know. I think it's hard because you can't predict what's in the repo, right? So it's like, and you don't know when they bundled FFmpeg, you just don't know.

• 20:23Speaker 00

  Right. It's nice when like, if it's just next to you, so you just run PMPM install.

• 20:28Speaker 01

  Correct, correct. You can like do special, like there's obviously through constraints, you can make optimizations. And I think the general purpose container is just like challenging. That being said though, I think there's probably some work to do on, you know, auto detection and preempting and stuff that could be done there. But it's just a bigger, it's a broader problem space, right?

• 20:48Speaker 00

  Yeah.

• 20:49Speaker 01

  And, uh, yeah.

• 20:50Speaker 00

  So fun fact, when I was at Nellify, I actually wanted to reach out to Vercel to do like a standardized open source auto detection thing of frameworks.

• 20:59Speaker 01

  Oh, yeah.

• 21:00Speaker 00

  And I like, we never, we never really like got internal momentum on that. It was an idea. I was like, shouldn't this be open source, you know?

• 21:07Speaker 01

  Yeah.

• 21:08Speaker 00

  Like auto detection is a, is a common utility that everyone needs.

• 21:11Speaker 01

  Yes, yes. I remember that. I'm having a flashback. It's like have.

• 21:14Speaker 00

  Yeah, everyone builds their, then probably we shouldn't all build it.

• 21:18Speaker 01

  Right. No, it's, um, and then also like what are your defaults? They're not exactly the same, which would be better to like, just having even the same like a preference stack.

• 21:27Speaker 00

  Yeah.

• 21:28Speaker 01

  Of defaults is the right.

• 21:29Speaker 00

  Yeah.

• 21:30Speaker 01

  Would be great because then we can move the whole ecosystem together from like to PMPM or Bun, right?

• 21:36Speaker 00

  Um, okay, so. Are there other movements or protocols or standards that you're interested in? Like MCP was a big winner this year. There's other like, I don't know, A2A, ACP, all this, all these like back winners.

• 21:50Speaker 01

  Interesting. I'm not as familiar with.

• 21:52Speaker 00

  ACP, the payments or one or the, is that, or had one?

• 21:55Speaker 01

  Oh, no, there's that one.

• 21:55Speaker 00

  That one, okay.

• 21:57Speaker 01

  And then the payment that was a Stripe or Coinbase?

• 22:00Speaker 00

  Stripe.

• 22:00Speaker 01

  Stripe? Yeah, that's very cool.

• 22:02Speaker 00

  We've had them on the pod.

• 22:03Speaker 01

  Okay, yeah, that's very cool. It'd be interesting to see if that takes off.

• 22:07Speaker 00

  I mean, it's Stripe.

• 22:09Speaker 01

  Yeah, but it still needs to be adopted by like the clients, right? And I think that's fascinating. The MCP is huge. It seems like it is the way that a lot of the, especially when it comes to like digital transformation or some of our enterprise customers. It's where they're going to be able to be at, or where they are able to add context. In addition to that, we also have custom agents that we announced today too. So like you can work with prompts and stuff within your, within your agent HQ and customize these agents for different tasks. And those can have MCPs and such. And I think that's going to be really powerful from like a platform perspective. It gets me excited. That's what I think is like shipping now and the next, but we're always on the lookout for like the next thing. And I don't know, what's on your, what's top of mind for you?

• 22:50Speaker 00

  Uh, for like standards or?

• 22:52Speaker 01

  Standards.

• 22:54Speaker 00

  Dev container.

• 22:57Speaker 00

  Look, I think dev container just is a PR problem. It's a great idea.

• 23:01Speaker 01

  Right, right.

• 23:02Speaker 00

  Just no one makes it interesting. I think you can do it, basically, this one.

• 23:07Speaker 01

  Okay, added to my list.

• 23:09Speaker 00

  But before that, probably you have a bunch of other stuff that I do want to get to. But just staying on the AI stuff, like I think we're actively exploring computer use as a thing. Like because it kind of got going a little bit. People were very excited and then they found out it was slow and bad and inaccurate.

• 23:27Speaker 01

  It is computationally intensive, in my understanding.

• 23:30Speaker 00

  It's getting better.

• 23:31Speaker 01

  Yeah.

• 23:31Speaker 00

  Uh, especially with open vision models like Deep C O C R and OMO O C R. Uh, like just give it a few more turns of the.

• 23:40Speaker 01

  Yeah, it's scaling. It seems very, it seems like it's like you need that edge case. Uh, and primary is just see how's like a modality works.

• 23:48Speaker 00

  And I think a lot of people are on the code gen side, the code agent side, a lot of people are trying to think about, all right, we had this evolution from co-pilot to like, uh, you know, more agentic, uh, sort of like a cloud code situation is where I think that the status is. Like what's next, right? What's, what's the, what's the, the, the, the obvious next step?

• 24:06Speaker 01

  Making them good?

• 24:07Speaker 00

  Making them good, yeah. You don't like look cause?

• 24:09Speaker 01

  No, no, I, I don't. It's more just like, you know, the, the, the devils in the details, like,

• 24:14Speaker 00

  Yeah.

• 24:14Speaker 01

  Going from 90%, going to like hill climbing, it gets deeper.

• 24:20Speaker 00

  Yeah.

• 24:20Speaker 01

  In my opinion, it gets deeper. And so going from 90% success to 95 to 98 to 99 to nines of success.

• 24:28Speaker 00

  Yeah.

• 24:29Speaker 01

  Really hard.

• 24:29Speaker 00

  Paying Merkle a lot of money for, uh, expert programmers of open source maintainers. They're like,

• 24:35Speaker 01

  Then you realize along the way, maybe the users aren't that good at it.

• 24:37Speaker 00

  Yeah.

• 24:38Speaker 01

  No, but I, I just think there's a lot of work to do to finish the swing.

• 24:43Speaker 01

  And there's a big difference between 98% to 99% correct, uh, and that's like noticeable.

• 24:49Speaker 01

  And this used to hit, you know, if, if you're working on AI product, you, you probably don't realize how, you've probably seen this, like, most people are blind, like living in La La Land about how, um, poor quality, uh, their, their AI product label is. Unless they're really measuring like a number of error free sessions, like how many errors are coming from the infra, infra providers, like, you know, how, well, how many requests are dropped, how fast, how fast these things are. And so that's something that we cared about at Forcel quite a bit and like, we'll care about.

• 25:19Speaker 00

  Do you have like a daily review of your dashboard or I don't know?

• 25:23Speaker 01

  Daily, daily, I'm just like, yeah.

• 25:24Speaker 00

  Okay. I thought you were gonna say days too much.

• 25:27Speaker 01

  Days. No, and you're thinking of it as like every three hours, a roll up of, of key metrics and stats.

• 25:33Speaker 00

  Yeah.

• 25:35Speaker 01

  And like one of them was like error free sessions and other things like that. That was like really important because, you know, especially now with agents are like multi-turn.

• 25:44Speaker 01

  I have a tweet about this that was like in 2024, which is that like agents will really only work where we get to not only like the more intelligent models, but better reliability of the infrastructure providers, right? These aren't, these are not, inference is not like a database, like update uptime, right? So there's still differences between providers, there's little differences between performance and difference of times. And that's why you see things like open router being very successful and, and different gateway products. Because like reliability, you need to switch, they go down all the time.

• 26:12Speaker 01

  So, long story short, yeah, we would, we would do like, you know, it was almost like video game style. Like we'd have like all the data coming in all the time.

• 26:19Speaker 00

  Yeah.

• 26:19Speaker 01

  Uh, and that allowed, I used to joke this by mood ring, like good day, bad day.

• 26:25Speaker 01

  So it was very successful for us. Uh, I think other teams should adopt that like data driven approach.

• 26:31Speaker 00

  I think one thing that's surprising is the lack, the relative lack I still see on data analyst agents, where you can sort of chat it, like add a Slack bot for the precise analytics that you want to generate. Because I think we're still in the BI era.

• 26:49Speaker 01

  Yeah.

• 26:49Speaker 00

  Isn't that weird?

• 26:50Speaker 01

  Yeah, I totally agree. It's just that that space hasn't been like captured as much.

• 26:56Speaker 00

  Yeah.

• 26:56Speaker 01

  Like, I guess maybe now, actually I'm interested in this like shift to a way into, in like, into knowledge work tasks with coding agents.

• 27:04Speaker 00

  Okay.

• 27:05Speaker 01

  I wonder.

• 27:05Speaker 00

  Using coding agents for non-coding tasks?

• 27:07Speaker 01

  Correct.

• 27:08Speaker 00

  Do you, do you do that personally?

• 27:09Speaker 01

  Yeah, yeah, yeah.

• 27:10Speaker 00

  Yeah, what, what do you do?

• 27:11Speaker 01

  Well, like I was, this summer I was doing, like I was helping, I was trying to automate some of my dad's workflows and stuff like that. And just like some of his, he's got some Excel spreadsheets and uh, for like accounting, like, like uh, financial accounting, or managerial accounting, I guess. Um, and uh, yeah, just like point cloud code of that stuff and see what happens. And like, it's, it ends up doing Python and, and, and generating some scripts and it kind of got off, down on the hairs. But it was like, even he saw that it was better at it than the chat client. That's excess.

• 27:41Speaker 00

  Super obvious.

• 27:42Speaker 01

  It became kind of obvious. Yeah, it felt better.

• 27:44Speaker 00

  I wonder if he can try Cloud for Excel and see.

• 27:47Speaker 01

  Yeah, yeah, you got me right. The um, and then of course you got the browser, the, I don't know what, not browser, browser space agents, but not, not computer use, but browsers with agents, agent browsers. Agent browsers.

• 27:59Speaker 00

  For Flexity.

• 28:01Speaker 01

  And it's like, is that the better? If that's true, then maybe the, the general purpose injection point is there.

• 28:05Speaker 00

  Yeah.

• 28:06Speaker 01

  Uh, what do you, or have, which, have you tried any of the agent browsers?

• 28:09Speaker 00

  All of them.

• 28:09Speaker 01

  All of, which, where?

• 28:10Speaker 00

  I, I am very, uh, I'm currently maining Atlas, mostly because I just want to give ChatGPT a fair go.

• 28:16Speaker 01

  Okay.

• 28:16Speaker 00

  Um, I, but I'm very stuck to the arc and like the vertical tab.

• 28:20Speaker 01

  Oh, okay.

• 28:21Speaker 00

  I think any pro user, like I have multiple businesses.

• 28:24Speaker 01

  How many tabs you?

• 28:25Speaker 00

  I'm context switching. I have hundreds of tabs open. I made a, I made an open source tool called Chrome Dump, you can find it on my GitHub, uh, where it literally dumps all the tabs open, it summarizes them, and then I can close it by deleting them on Markdown.

• 28:37Speaker 01

  That's pretty cool.

• 28:40Speaker 01

  So you just go on like a, you just go on like a, like a, like a, a bender?

• 28:44Speaker 00

  Yeah.

• 28:44Speaker 01

  And then you just dump it?

• 28:46Speaker 00

  It should be as easy to close as Markdown and Chrome isn't that good at the performance side of things yet.

• 28:51Speaker 01

  And you were working on like some browser comparisons?

• 28:54Speaker 00

  I, I was, but so I tried to build it in Tori.

• 28:56Speaker 01

  Okay.

• 28:57Speaker 00

  Tori explicitly doesn't want you to build a browser and I, I tried to fight it too much.

• 29:02Speaker 01

  I see. Yeah, I think. Very cool.

• 29:04Speaker 00

  So just to wrap things up and, and we're, we're around about time. There are other side projects, uh, tasks and things that you've, you've, uh, announced here. Uh, first of all, redesign GitHub homepage, which a lot of people don't even know GitHub has a homepage. I, I'm legitimately one of them.

• 29:19Speaker 01

  Tweet, which is, uh, Riz's tweet, like printed out, like, you see the tweet, like, there's a tweet, uh, from Riz, and it's like, no one uses. None, all of this stuff is totally useless. I'll, I'll pull it up, like.

• 29:31Speaker 00

  Yeah.

• 29:32Speaker 01

  I, I have, I cordial today, because we were like, when we launched, let me get it right, because I gotta do it right. Hold on.

• 29:37Speaker 00

  Okay.

• 29:38Speaker 01

  Was, um. Incredible how pretty much the entire GitHub homepage is useless. And is 1.3 million views and 19,000 likes. And this was May 2, 25. So, hurt. The team, the team made improvements and today they launched a new GitHub homepage. Yeah, which we're very proud of. And they should be really proud of. It's got tasks at the top. It's got recent PRs. Some stuff is still there, like your recent repositories. I still, I think there's more work to do, but it's like really overhauled and they did an amazing job with it. So they nailed it. But you know, more work to do, never done. And like, hopefully we can keep iterating with the community and, and everyone. And, uh, and keep going.

• 30:18Speaker 00

  The last thing I want to hit you on is stack tips.

• 30:20Speaker 01

  Oh, yeah.

• 30:21Speaker 00

  You asked everyone when you joined, what should I work on or something?

• 30:24Speaker 01

  Yeah.

• 30:24Speaker 00

  I don't know if this is your job specifically.

• 30:26Speaker 01

  It wasn't.

• 30:27Speaker 00

  But why do people want stack tips so much? Let's, I, I think you have some history there.

• 30:31Speaker 01

  Yes.

• 30:32Speaker 00

  Anyone who's interacted with anyone at Facebook knows about Fabric.

• 30:35Speaker 01

  Yeah, we're just about it. Yeah.

• 30:37Speaker 00

  So can you explain why, why it's been so, like what it is, why is it so hard?

• 30:41Speaker 01

  Okay, so this, um, this concept of pull requests, which we're all familiar with. You write some commits, you open a PR and then you merge the PR and

• 30:55Speaker 01

  you go about your day. So as you scales like larger organizations,

• 31:00Speaker 01

  and like you look at your history and, and, and there are people who are like very,

• 31:05Speaker 01

  I'll say like have like near religious beliefs about how to Git. How did you get right? Rebase versus, rebase versus merge. There's, there's a crowd that wants to fast forward the repository so to preserve all the history. And then there's a crowd that wants to squash and, and merge into the.

• 31:19Speaker 00

  I'm being squashed.

• 31:20Speaker 01

  Okay. Anyway, Facebook, and I've, I've never worked Facebook, but in my previous startup, before Versal, Turbo Repo, I did a lot of research on build systems. And at Facebook, they have, not only they have their custom build tools called Buck. They also have a custom file system and they don't use Git. They use Mercurial, which, and then now it's sort of custom and it's all wired together. And at, at, at Facebook, they don't use pull requests. They have a different sort of philosophy. You can, um, it's sort of like the best, like, think about this is like, imagine every PR just had one commit in it.

• 31:59Speaker 00

  Yeah.

• 32:00Speaker 01

  You could branch them. And re, and the critical thing is you can restack them. And then if you restack or make a change later, like earlier in the stack than later in the stack, and these stacks are just diffs, right? The commits are just diffs, and it's the term stack diffs. You can then collapse them and merge the last one, merge them towards it, and it just gives you a little bit of a nicer workflow. And it's what people, if you work on a monorepo or you work on a very, very large codebase, it's a really, really nice way to work, especially if you've got a system that will automatically restack. And then if you think even more deeply about it, like, and get really deeper into the weeds,

• 32:41Speaker 01

  you can decide which diffs in the stack CI should run against, if you get fancy. Okay, like some kind of commit message or something. There's always, yeah, you could decide, like, maybe this one doesn't need it, or skip that one, or whatever. And you end up getting these, like, sort of, these groups, these stacks. And it's really nice from a code review perspective, because when you go to update, or you can update a different part of the stack, it just makes it a little bit more fluid. And so it's what people want. There are a couple of tools out there in the market that do this kind of behavior. One's called Graphite. There are a couple others. So many tools called Graphite. There's another Graphite right here. Yeah, yeah. And it's a very workflow. And so it's been the top pull request, or sorry, the top. Feature request.

• 33:23Speaker 01

  Feature request, thank you, at GitHub for years, from the community's perspective. I don't know at GitHub, but. And then, so as soon as I joined, the first thing I did, I was gonna look this up, and, well, that's the first thing. I asked how should we make it better, and it was the top feature request, right? And then I went to go like, okay, investigate, like any good product person would. And there's been multiple attempts at this internally, going back to like 2020. Okay. And there was one very, very, very polished attempt too in 2022. And it just, I don't have all the context, so, but it was, it was, there was a pretty good implementation. All of the work was done on the client, and it reintroduced this new concept called stacks

• 34:05Speaker 01

  outside of the pull request into GitHub, and it was a little too risky. It was sort of deemed too risky, too big of a change. That's just what I was told. So, anyway, we're, we have a couple of meetings internally already, and we're trying to weave it into the planning and the roadmap, and so hopefully we'll be able to share more updates soon. But like, it's a top a little less known feature. Again, heard. Yeah, and so, and like, we're working on it, obviously like something the size of GitHub to move to like support this kind of new. This your feature is like not just like a walk in the park because of the size of GitHub's and GitHub's Git implementation, but it's something that we're actively exploring.

• 34:43Speaker 00

  Yeah, well, I think, you know, just to wrap all that up, you know, I think it's really nice for someone who's so deeply engaged and like coming from like one of us, literally.

• 34:54Speaker 01

  Yeah, yeah.

• 34:54Speaker 00

  That you now run things at GitHub and we can just at you.

• 34:57Speaker 01

  You can do that, I mean.

• 34:59Speaker 00

  And I think like Ajay Kapati was the other day was saying like every company needs one of these where you can just like, hey, like this, this really shouldn't exist at GitHub. We love GitHub, we use GitHub, but like, come on and.

• 35:09Speaker 01

  Well, yeah, feature requests, welcome. Like my DMs are always open.

• 35:13Speaker 00

  Oh, airful. I don't know. 180 million developers.

• 35:15Speaker 01

  I like 100, whatever. I am of the philosophy that like all feedback is a gift. Like it's all a signal.

• 35:22Speaker 00

  Yeah.

• 35:22Speaker 01

  And the more signal we can collect, the better decisions we can make and truly build this really, really useful website and company like together and that's going to be the future. And if we focus just on that, we're going to be okay.

• 35:35Speaker 00

  Yeah, we're going to be okay. All right. Well, thanks so much, Jared.

• 35:37Speaker 01

  Yeah, it's been a pleasure.

• 35:39Speaker 00

  Awesome. Catching up.

• 35:39Speaker 01

  Yep, likewise.

• 35:40Speaker 00

  Congrats.