# Vercel’s CEO & Founder Guillermo Rauch on the impact of AI on Web Development and Front End Engineering

- **Show:** Gradient Dissent
- **Published:** 24 Oct 2024
- **Source:** https://wandb.ai/site/resources/podcast
- **Transcript:** https://podcasts.opennash.com/shows/gradient-dissent/vercels-ceo-founder-guillermo-rauch-on-the-impact-of-ai-on-web-development-and-f-8b64b8/
- **Provider:** OpenNash Podcasts (machine-generated)

## In brief

Guillermo Rauch argues that AI is changing web development less through smarter models than through better interfaces and faster iteration. Vercel's v0 agent turns plain-language prompts into working front ends. He says the winning AI products ship daily, listen closely to users, and use streaming and component-based interfaces rather than plain chat text.

The episode moves from Rauch's childhood hacking and open-source career to how v0 is built and shipped. It then widens to a larger tension: do incumbents with data and users win, or do AI-native startups that rebuild the interface? Rauch leans toward startups, because each big computing shift has come with a new front-end model.]

## Key insights

1. **Daily shipping is v0's core strategy** v0 posts a product changelog every day. Rauch says the AI field moves so fast that annual launches are already too late.
2. **Fast deploys and feature flags enable fearless releases** Each v0 deploy takes about a minute with zero downtime, and the team runs dozens of feature flags at once. Rauch says this lets them swap models and test changes safely.
3. **AI evaluation mixes tests, judges, and production data** Code output can be checked deterministically, while visual quality still needs humans and facts can be checked with an LLM acting as judge. Rauch says the most important check is testing in production against conversion, engagement and session length.
4. **Streaming reliable JSON enables real AI interfaces** When a model streams structured data, a React component such as a weather widget can be drawn directly from it, instead of only showing text. Rauch says Vercel's AI SDK exists to make this pattern easy.
5. **Startups can win by reinventing the interface** Rauch says every major computing shift, from GUIs to the web to mobile, changed the front end. He cites Chatbase reaching $4 million in annual recurring revenue with one person as an example of a startup skipping legacy constraints.

## Questions this episode answers

### What is Vercel's v0?

Rauch describes v0 as an expert AI agent for web engineering, especially front end. You describe what you want in a ChatGPT-style prompt, or paste an image, and it generates working UI code.

### Why does Vercel release new features every single day?

Rauch says the AI world moves so fast that waiting for an annual launch means you are already behind. Daily releases let the team respond to customer feedback in real time.

### What is the AI SDK that Rauch keeps mentioning?

It is Vercel's open-source toolkit for plugging AI models into interactive front-end apps. It helps developers handle streamed model output and connect it to interface components like React widgets.

### Does Rauch think open-source AI models are as good as closed ones?

Not yet for his use. He says Llama did not match frontier models in Vercel's evaluations, though he would like to switch because open models bring competitive pricing and less lock-in. He also says Llama is open-weight, not truly open source.

### Will big incumbents or AI-native startups win the AI product race?

Rauch says incumbents with lots of data and users can ship strong products. But he leans toward startups, because each major computing shift has changed the interface and startups can rebuild without legacy baggage.

## Facts

- **Guillermo Rauch** [0:03:54]: "at about 16 years old i became a core developer of this uh library called mood tools"
- **Guillermo Rauch** [0:04:34]: "i created a library called socket.io for real-time communication the parser the protocols a lot of things"
- **Guillermo Rauch** [0:05:22]: "which helped a lot with my immigration story in the US, getting my O1 visa."
- **Guillermo Rauch** [0:05:55]: "It's the most popular React framework, 1.2 million monthly active developers, 7 million weekly NPM downloads."
- **Guillermo Rauch** [0:06:28]: "a lot of those folks are sort of standardizing on Next.js as a framework above which they build their innovations."
- **Guillermo Rauch** [0:12:42]: "It was a great mission, but it wasn't a great market."
- **Guillermo Rauch** [0:16:10]: "Your.com is hosted and built on Vercel. It's absolutely mission critical."
- **Guillermo Rauch** [0:19:42]: "we did more UI generations than the previous 12 months"
- **Guillermo Rauch** [0:23:56]: "We release every day. So if you go to twitter.com slash V0, you're going to see that that account is just a product change log."
- **Guillermo Rauch** [0:24:32]: "The latest one was we introduced V0 Teams and Enterprise yesterday."
- **Guillermo Rauch** [0:24:32]: "this is using a capability from Gemini to do very, very fast code diffs and applications of code diffs."
- **Guillermo Rauch** [0:25:54]: "each time we deploy it takes about a minute to build deploy with zero downtime"
- **Guillermo Rauch** [0:26:38]: "We actually now is spreading our GPU usage across three different data centers because we're generating so much, so many tokens."
- **Guillermo Rauch** [0:28:49]: "A big one though is actually testing in prod. So we like getting a feature flag out and being able to observe more like the product behavior."
- **Guillermo Rauch** [0:31:52]: "there's this company called E2B who created AI sandboxes as a service so that you can add code execution, code interpreting to your applications."
- **Guillermo Rauch** [0:33:55]: "And then like five minutes later, he sent me the package."
- **Guillermo Rauch** [0:34:50]: "now the final mode, and obviously there's going to be more, is we can get streams of reliable JSON."
- **Guillermo Rauch** [0:35:33]: "look, there is this tool that maps one-to-one with my React component called WeatherWidget."
- **Guillermo Rauch** [0:39:33]: "Every line of code in your system must have been compiled by you."
- **Guillermo Rauch** [0:40:01]: "Most models, like Llama, is not giving me access to everything that they've used to cook the model."
- **Guillermo Rauch** [0:40:36]: "Apple put out a real open source large language model where they were opening the data set, the instructions for how to build it"
- **Guillermo Rauch** [0:42:56]: "we're actually using only frontier models at the moment, excluding LAMA. So I would love to use LAMA, but in all of our evaluations, it wasn't quite up to par."
- **Guillermo Rauch** [0:45:42]: "the few computers in our neighborhood in argentina that existed were running ms dos"
- **Guillermo Rauch** [0:47:07]: "Adobe just put out an amazing product yesterday. I don't know if you saw it for how you can do photography with Gaussian splatting."
- **Guillermo Rauch** [0:49:08]: "they're basically building the the Bloomberg terminal with AI."

## Transcript

**[0:00:01] Lukas Biewald:** You're listening to Gradient Dissent, a show about making machine learning work in the real world, and I'm your host, Lukas Biewald. Guillermo Rauch is the CEO and co-founder of Vercel, which is a company I really admire. They make it incredibly easy to deploy websites using the Next.js framework, which they also invented, and they've been doing a ton of cutting-edge stuff on automatic website development with AI. It's a product I use and really like, and we really get into how to make a good product.

**[0:00:35] Lukas Biewald:** But then the conversation actually expands to other AI products, both inside big companies and in AI-native startups, and what makes them good. And he goes down a list of his favorite companies and applications. I hope you enjoy this conversation. I learned a lot from it. I thought it'd be fun to start with some softball questions of things we have in common. I mean, And I saw that you started programming at age 10, I think you said.

**[0:01:01] Lukas Biewald:** Yeah. I also started programming at a young age. I was kind of curious what your first computer was and first language was and what you were doing at age 10.

**[0:01:10] Guillermo Rauch:** Yeah. Our earliest computer was a PC. I think it was Windows 95. Nice. At the time, I was probably like six, seven years old, really. And I had a lot of false starts. I don't know if it was true for you, but I had a lot of false starts trying to learn to code. I would do a little bit here and then like not continue a little bit here and then not continue. My very first inspiration was, or the thing that got inside my head was when my dad said, well, like video games are created, created by people that know how to code.

**[0:01:41] Guillermo Rauch:** It's like, if you know how to code, like it's the keys to the kingdom, so to speak. And, uh, my very, very first line of code must've probably been, I mean, if we take coding to its extreme i would say it's probably an app called logo which is like you would compute the trajectory of a little turtle totally that's like the most primitive programming i had done yeah yeah uh and then probably visual basic basic and c kind of or i'd run all at the same time lots of html type programming but the thing that really was i think like the made it the calling of my life was i got linux installed on rpc which is like in itself

**[0:02:29] Guillermo Rauch:** a huge feat for me totally and then using the terminal to run the compiler and noticing how easy it was to install all the compilation tooling compared to what i had experienced on windows and just compiling c progress with gcc and like just running the binary and everything is quick and snappy and easy um that's kind of when i like it took off for me and it became a career and that was like age 10 yeah probably maybe even earlier at that point when i started like really coding was when i got more into like html php i started creating and publishing websites i started making a little bit of money like i think my first check was at about

**[0:03:12] Guillermo Rauch:** 11 or 12 years old from like this all-night referral program that i participated in and then when i started forking software like i would like i would find all this pearl or php or um kind of like software that i could just like use as a platform so if you remember like the era of like the php based forums yeah of course uh like phpbb and the bulletin and like there were all this uh there were cmss now there's a whole controversy they run WordPress but I was also a really big fan of like like standing up sites and platforms for my friends family I run the uh online forums for my high school this is a little later

**[0:03:54] Guillermo Rauch:** I actually became quite popular like Facebook wasn't a thing yet and like it became like the a central point of connection for all the students um the web really captivated my attention and then later on of course open source and a bunch of like the foundational javascript and ui libraries of the of the web uh at about 16 years old i became a core developer of this uh library called mood tools oh of course i remember mood tools wow that was like a big feat because uh i was still fairly young and the project was starting to like pick up some serious esteem it was started to get chosen as the foundation for a lot of startups i remember cloudera

**[0:04:34] Guillermo Rauch:** at the time had chosen it for like all of their like web UI. Facebook had hired like most of the core developer team. So my love kind of became enabling really snappy, real-time user interfaces. And later on, I realized that if I could create a programming model that was universal in nature, like one thing I really, really really became obsessed with was the the the hot buzzword at the time was isomorphic that you could run javascript on the client and you could run it on the server and they started creating a lot of libraries specifically to exploit that advantage so i created a library called socket.io for real-time communication the parser the protocols a lot of things

**[0:05:22] Guillermo Rauch:** had to run both server side and client side the test suite would be blazing fast because it It was like running the JavaScript client that it would run in the browser, but it was running it server side connected to one server created per unit test. It was pretty awesome. And so I kind of went deep in this rabbit hole of like running JavaScript everywhere. I actually wrote a book called Universal JavaScript, like Smashing JavaScript, Universal JavaScript Everywhere, which helped a lot with my immigration story in the US, getting my O1 visa.

**[0:05:55] Guillermo Rauch:** But as I kept going down this rabbit hole of like, making it easier for folks to create applications that were very high performance and very real time in nature, it led to what I ultimately am now more known for, which is having started Next.js. It's the most popular React framework, 1.2 million monthly active developers, 7 million weekly NPM downloads. And Next.js has sort of become the foundation or the operating system for a lot of the modern applications in the web, right?

**[0:06:28] Guillermo Rauch:** So when you talk to Claude Anthropic or when you ask a question to Perplexity or even when you create your sort of like first AI app, a lot of those folks are sort of standardizing on Next.js as a framework above which they build their innovations.

**[0:06:45] Lukas Biewald:** Wow. Well, you know, we could reminisce for a long time. I think we have a lot in common. I actually have used Socket.io heavily. Oh, wow. You were the author there. That's awesome. And I got my start in Logo and Visual Basic. I think I'm a little older than you, so I was kind of before the web was really there. So I remember I'd go to the library and kind of look up commands. And I remember when I found out that Logo could call functions, it really kind of blew my mind.

**[0:07:10] Guillermo Rauch:** I've been thinking about giving it to my kids now because it was such a nice way, such a friendly way to get into programming. Totally.

**[0:07:18] Lukas Biewald:** How old are your kids?

**[0:07:20] Guillermo Rauch:** So I'm 33. My kids are, I have five kids. The youngest is eight months and the oldest is eight years old. Wow. And then two, four, five, eight, and eight months.

**[0:07:32] Lukas Biewald:** And is the eight-year-old starting to code?

**[0:07:34] Guillermo Rauch:** Yes. One app that we've really gotten into is, it's actually not too unlike what it would be logo on the iPad, but you basically program a bear on a train. He's on train tracks and he faces all these difficulties along the journey to get to the other side. and you have to program all of the sequences of operations that he needs to execute whenever he finds a challenge. And you can be frustrated just like programming, right?

**[0:08:02] Guillermo Rauch:** Like you're almost at the end and then it crashes. So like you got an operation wrong. And so we're like in those early days of, and a lot of gaming, which I totally welcome. He's like, he just like me, he's really into Pokemon, on uh like the the game boy like nintendo switch etc kind of games and that kind of became an obsession for me like running emulators uh and then hacking on emulators and uh i remember like uh root kidding or like it's a sort of backdooring my own iphone so that i could like install cds so i could get like the uh the side loading of applications that weren't allowed on the iphone at the time so i

**[0:08:45] Guillermo Rauch:** could run my emulator there like so running emulators of like my favorite games or if you're familiar with wine of course to run windows applications on linux that kind of also became an obsession for me because when i moved to linux i forked off from the world like my friends would game on windows and i didn't have their games my family wanted to do the basic productivity software word and i couldn't write it on i was like mom but we have open office over here like no it doesn't work and so i had to do dual booting for a while because we had only one computer at home yeah and wine was amazing because it's this re-implementation of all

**[0:09:27] Guillermo Rauch:** of this sort of like foundational windows apis dll's so that they can run natively on linux so it's like a high performance uh emulator in theory of course and uh so just compiling stuff running stuff cross-platform, that also gave me a ton of exposure to the world of engineering in general. Like, even when I was like 11, I was like trying to like dual boot our computer and like compiling like the modem drivers so they could run on Linux and like hacking the kernel and then compiling Linux from scratch.

**[0:10:03] Guillermo Rauch:** So those things were all, I would call them like not programming, but like really good skills to have in your toolbox that even even today i'm thankful that like as we created the Vercel infrastructure which is basically how a lot of people host and deploy their next js applications and other frameworks when i was standing up the infrastructure for Vercel like just like that like deep expertise into linux and the the even the just like the thick skin of things going wrong and sideways and like uh like battling computers like those are with me since an early age you know it's funny one two things that you mentioned that i

**[0:10:43] Lukas Biewald:** think tells me that i might be a little older than you is is one you said that it was kind of easy to install the packages to get a developer environment set up on your linux box which i remember that being like just like excruciatingly painful with the no it's painful but i remember on

**[0:11:01] Guillermo Rauch:** windows specifically oh it's like impossible yeah getting c to compile there was this project called djgpp and uh god bless them they were trying to like create this like c and c++ development environment dx as we call it today developer experience uh on windows that was specifically because like i agree like every package on linux was a nightmare every uh i would have to like compile these kernel modules to get my internet work and it was it was nightmarish uh but uh but on the programming like specifically like low-level programming like that stuff just felt way smoother like you had better access to the operating system and uh windows felt more like unnatural i guess and the other thing

**[0:11:50] Guillermo Rauch:** was to your point like packages oh we're always a hit and miss on linux sometimes i would get lucky that they would get distributed as rpm because i started using red hat and you you get lucky that you find the rpm and it's easy to install and then sometimes you get really unlucky and like it's not the right version yeah sometimes the things that are trying to make your life easier are going to also become your your worst enemies totally

**[0:12:18] Lukas Biewald:** um all right well i feel like we could spend a whole hour on this but i wanted to get in kind of one more softball question because this is the question that i get asked most commonly on podcasts i'm kind of excited to turn this around to you which is you're a second time founder like me at least you've had kind of one successful company and this is your second um how do you think differently about doing a company for the second time have you changed your style as a

**[0:12:42] Guillermo Rauch:** founder? Yeah. So my first startup was in a space that was very exciting to me and my co-founders when we started the company, which was education technology at tech. And then later on, we quickly realized that it wasn't a great market. It was a great mission, but it wasn't a great market. And so the second time around, I became a lot wiser about surfing massive waves in massive markets. And in fact, when you look at the investments that we've made over the last two years in technologies like the AI SDK and V0, it kind of comes from the same point of view of when you're starting, you're not quite sure what the right incantation of product will resonate with

**[0:13:30] Guillermo Rauch:** the market. But you know deeply that it's going to be huge. Software engineering with AI is going to be huge no matter how you slice it. And so I would say when I started Vercel and later on as we've innovated in our product lineup, I've applied this formula, sometimes subconsciously, sometimes explicitly of thinking like, I want to participate in the biggest possible market and impact the highest number of people. and another metaphor that gets thrown out a lot is like it's sort of like painkiller versus vitamin i think that kind of came after my first startup but i think i only work on painkillers now what i do is i invest in vitamins in terms of the craftsmanship of the

**[0:14:21] Guillermo Rauch:** product like not everything at Vercel has been like oh it's if it's not a painkiller like get it out of the roadmap you know, we invest our fair share in things that are actually, if you look at them in isolation, they're of inconsequential value to the buyer. Like little things that we're refining the UI and the right colors and design, we invest a lot in those things. And so those really came from a place of creating the most possible delightful experience.

**[0:14:49] Guillermo Rauch:** So it's almost like you still have things that you're building that are not just like the most pressing pain that a customer might have, but all up the product, I always wanted to make sure that my second startup would be in a mission critical path.

**[0:15:05] Lukas Biewald:** And what is that mission critical path? I got to say, it's interesting. I actually kind of hate that painkiller versus vitamin metaphor. I feel like a ton of successful companies in enterprise and consumer, you could kind of rebrand them as painkillers, but they kind of look like vitamins. I'm always telling the team, let's make ice cream. Let's make something that people really want to have. And one thing that I love about your product is it's obvious that you go deep in making it a delightful experience in a way like other developer tools don't do.

**[0:15:36] Lukas Biewald:** So how do you reconcile that painkillers-vitamin analogy? I feel like that could take you in some

**[0:15:41] Guillermo Rauch:** wrong directions potentially. Yeah. That's why I actually clarified that we invest in a fair share of sort of like vitamin additives to the product. It's almost like when you buy, you know, this like modern nutritional products and they say like, oh, and we also threw in some vitamin C in there. But fundamentally, you're getting nutrition. You're not getting empty calories. Right. So I wanted to make sure that, and by mission critical, what I mean is you can't turn off your.com.

**[0:16:10] Guillermo Rauch:** Your.com is hosted and built on Vercel. It's absolutely mission critical. I think I was pretty spot on on that one because my previous product wasn't. And this time around, we've already been through a few micro macro cycles, if you will, since I started this company. So when I started it, we were coming off like a big hype wave of sort of like the previous generation of like social products, consumer products, mobile. and the music was still on in terms of tech and zero interest rates and whatnot.

**[0:16:49] Guillermo Rauch:** But then we kind of entered the pandemic and at the beginning of the pandemic, it was like, oh, like we're not sure what's happening with the software industry, right? Then during the pandemic, e-commerce and a lot of the key industries that Vercel supports like just like mooned overnight, right? And like a lot of digital transformation projects projects were accelerated which benefit us tremendously but then it's like it's like almost like it's so over and this is what i mean by i think when we started entering the post zerb phase i do think that vitamin versus painkiller metaphor did get some it's time to shine again a lot of productivity software that companies had redundancies in like oh we have

**[0:17:37] Guillermo Rauch:** four ways of managing maybe diagrams or documents or lists or whatever. A lot of that SaaS category saw tremendous downward pressure. On the other hand, us, when we positioned ourselves sort of like mission, having been positioned as sort of mission critical, we navigated that quite well. And then we found ourselves sort of fueling this new era of innovation in supporting companies that are building with AI. A lot of the AI startups use Vercel to build their products.

**[0:18:08] Guillermo Rauch:** And then we leaned in on the new sort of wave with what we do best. I remember at the time I shared with the company that I didn't want to push random acts of AI. I didn't want to like throw in Clippy into a product. Just yesterday I was filing my taxes and there was this like floating owl asking me if I wanted help with my taxes. I was like, can you just get out of here, owl?

**[0:18:33] Guillermo Rauch:** I don't like taxes to begin with and I just don't like your face right now. So no random acts of AI. So we lean in and what do we do well, which is developer experience. Developer experience and UI and front end. And we wanted to create products that made sense in the context of those things that the company had a legitimate DNA or brand permission to execute in.

**[0:18:56] Lukas Biewald:** Okay, so let's jump in and let's talk about V0 here. Because I think you had a launch right now, but it's been out for a while. Just for people who haven't experienced it, can you kind of say what it does?

**[0:19:08] Guillermo Rauch:** Yes. Think of it as an expert AI and agent in web engineering, especially in Frontend. Not exclusively, but what I hear from a lot of people is like, look, I can go in and it can generate UIs for me. I might have not had a ton of experience with Frontend. or it seems like on Hacker News, there's a new front-end library every week or every new major library every week. So I need helping just keeping up or migrating code or even just exploring the latent space, so to speak, right?

**[0:19:42] Guillermo Rauch:** Like what's possible to do with front-end technologies? And the fact that we gave, we leaned in on the sort of like ChatGPT-esque style of like open-ended prompts. And the fact that you can just prompt your way to a working application has been amazing for people. So we've done now, like in the last month alone, I shared statistics. In about over a month, we did more UI generations than the previous 12 months. So it's growing exponentially, and people are really appreciating this sort of like co-pilot experience, but for generating web UI.

**[0:20:28] Lukas Biewald:** I mean, the name vZero to me is a little bit evocative of like, hey, this is going to help me get started. Was that the intention? Because I feel like where I really want help is when I've been working on the website for a while and I just can't get that goddamn button to go where I want it.

**[0:20:45] Guillermo Rauch:** Yeah. Yeah, even though, so it's very, very good when you're getting started, but over time by just making it more capable, it can help you every single day. And actually, so the most recent release that we've made, we've gotten the product to be a lot more useful as a daily driver. And for example, even taking that example of like, I don't get the button to just work the way that I want to. This is a great tool to just go and prompt your way to the right incantation of properties.

**[0:21:14] Guillermo Rauch:** Or a lot of what people really spend and arguably waste their time on is CSS. us our cto host jokes that like this is the product he's been dreaming for because even though he's been programming the web for like 20 years he he was at google for 11 css was always like the thing that he did understand if you've ever seen that there's a family guy meme where uh like they're trying to like uh adjust the blinds have you ever seen this like american blinds like you try to like figure it out and like you're like it feels like you just have to

**[0:21:48] Lukas Biewald:** I've experienced that myself, yeah, absolutely.

**[0:21:50] Guillermo Rauch:** That's CSS. You know what's interesting about AI? I think AI, even though people might argue all day long whether it can think or it cannot think, you have to recognize that a lot of our programming is not like, I'm Albert Einstein, I'm creating general relativity. A lot of it is like pointless trial and error and looking up things on Google or Stack Overflow or trying and try to remember the right statistical combination of CSS properties that you have an image of what you want in your head and you're just trying to translate it into working code.

**[0:22:28] Guillermo Rauch:** And so something that V0 is surprisingly good at is when you paste an image. You paste an image and it'll give you the right CSS and specifically Tailwind most of the time to make the application look like that. So it ends up helping a lot when you're stuck and you're like, okay, how the hell do I fix this visual glitch and whatever? You just go to the AI agent.

**[0:22:54] Lukas Biewald:** And is it intended to be used in conjunction with things like Cursor or GitHub Copilot? Like how would it fit in to a workflow like that?

**[0:23:01] Guillermo Rauch:** Yeah, we didn't intentionally say like this is going to be like peanut butter and jelly to Cursor, but a lot of people are telling us their life today is just V0 Cursor, V0 Cursor, V0 Cursor. Like you go to V0, like IDA, design, generate, and then you throw it into your code base and then you might make some adjustments over time where Cursor might help with our composer ability or auto-completion and so on and so forth.

**[0:23:26] Guillermo Rauch:** And that's also how I use it myself.

**[0:23:28] Lukas Biewald:** So, I mean, I think a lot of people are watching the show to kind of figure out how to make Gen.AI applications themselves really work. So can you talk about like kind of the technical journey of making this and can you share details? Like, is it backed by, you know, Claude or GPT and like, do you fine tune? Do you do a lot of like prompt engineering and are you using agents to a chain of reasoning to make it more accurate?

**[0:23:52] Lukas Biewald:** Whatever you can share, I think.

**[0:23:53] Guillermo Rauch:** At this point, I would say all of the above.

**[0:23:55] Lukas Biewald:** Nice. Excellent.

**[0:23:56] Guillermo Rauch:** So first of all, V0 has given us a lot of very interesting realizations as a company. I'll start with the first one. We release every day. So if you go to twitter.com slash V0, you're going to see that that account is just a product change log. And every single day there is a ship.

**[0:24:18] Lukas Biewald:** Okay, wait, let's take a look. Let's take a look. What do we got today?

**[0:24:21] Guillermo Rauch:** I actually learned about what our team does on Twitter now, which is great for me. I spend so much time on Twitter. I love it.

**[0:24:28] Lukas Biewald:** Wait, where did we find it? V0? Like V number zero? All right. I'm not even following it.

**[0:24:32] Guillermo Rauch:** So yesterday, the latest one, there's probably one coming later today. The latest one was we introduced V0 Teams and Enterprise yesterday. And the day before, we launched better playwright support so that you can use V0 to write your tests. and the day before we launched an improvement, and actually this is using a capability from Gemini to do very, very fast code diffs and applications of code diffs. And the day before, we launched a new way of finding out about features because we're releasing so fast.

**[0:25:09] Guillermo Rauch:** And the day before, we got our mark -

**[0:25:11] Lukas Biewald:** You're not joking, dude. This is awesome. Okay.

**[0:25:14] Guillermo Rauch:** Okay, I'll explain why. I love it. So what we realized is that the AI world is moving too fast. If you thought you were Tim Apple and you get to be on a stage once a year to announce your AI innovation, you're NGMI. Because even if you plan, like in Q4, we're going to launch our AI thing, it's probably too late. If you're not iterating in real time with your customers, with your team, et cetera, et cetera.

**[0:25:45] Guillermo Rauch:** Now, this is, of course -

**[0:25:48] Lukas Biewald:** Wait, wait, wait. Let's just take a pause. I'm just going to Slack. I'm going to Slack my CTO and tell them about this. That's awesome. Let's take a quick break.

**[0:25:54] Guillermo Rauch:** And I open source the stack. So like what helps us move this fast is that this is the first product that we've built where it's full Vercel, full stack Vercel. so we dogfooted the entire Vercel platform front end back end AISDK next JS um Postgres database and one thing that we're obsessed with at Vercel is iteration velocity so like move and deploy really fast so each time we deploy it takes about a minute to build deploy with zero downtime including a zero downtime of the front end pieces, which front end and back end always need to be coordinated in a release.

**[0:26:38] Guillermo Rauch:** And like we have people that are generating all day long. We actually now is spreading our GPU usage across three different data centers because we're generating so much, so many tokens. And so we're dockfooting this confidence that the Vercel platform gives us to like release fearlessly. Feature flags, for example, is a huge thing. we're evolving the rack the models that run time very frequently so at any given time we have like 20 50 uh feature flags that are being like tested and so on so how do you um okay so like one thing

**[0:27:14] Lukas Biewald:** that our customers constantly you know talk about is you know with with code you can have cicd pipelines where all the tests pass and you ship it confidently even then you know bugs get in obviously. But with these AI pipelines, really hard to evaluate. Do you have an evaluation system built into your shipping process? How does that work? Yeah. We have it at CICD time.

**[0:27:38] Guillermo Rauch:** We have it ad hoc. If the developer is making a change that feels risky, they can run it one-off. What does run it mean? Run what? Like have a certain set of prompts they're trying out and running a bunch of evaluations against it or switch models at a different checkpoint and run our evaluations.

**[0:28:02] Lukas Biewald:** And how do you automate that? Do you use LLM as a judge here to say if the results are good? Because you're actually generating pages. Presumably there's more than one way to generate a nice web page or package.

**[0:28:12] Guillermo Rauch:** Some have been turned deterministic because we're producing code. and so we have a certain level of like correctness that we can self-assess there's also the visual aspect that has more human in the loop components yeah like oh does it look right there's also the a little bit of the vibes aspect of like just throwing a bunch of generations at it and it's like hmm like this this feels like a regression um there is sort of like certain questions that have been brought up to us by users as like hey that piece of information was wrong making sure there's no regressions of like some facts in which case you use an LLM as a judge.

**[0:28:49] Guillermo Rauch:** A big one though is actually testing in prod. So we like getting a feature flag out and being able to observe more like the product behavior. So conversion rates, engagement rates, session duration. A lot of what we do is surface a feedback from customers, because we're in the realm of code generation. Also, we are very, very mindful about the success rates of the code

**[0:29:19] Lukas Biewald:** execution. Okay, but how do you do all that in one minute? So you're saying you can run the full

**[0:29:24] Guillermo Rauch:** test suite in one minute might be just shipping the flag as opposed to like turning it on, right? So one minute is I have the new version of the software available for testing. Okay,

**[0:29:35] Lukas Biewald:** Well, so my CTO just has already been checking Slack during our conversation. I hate when people do that, but my CTO just wrote back, God damn, they're good. So, you know, hopefully... I appreciate it. ...we get in your shipping kit. It's a respect to your team.

**[0:29:47] Guillermo Rauch:** Yeah, and so another piece of it is, so we're developing this AISDK. We're very motivated to then give the recipe out of the world. Just like I'm more than happy to spend time with your CTO and like sharing how we do this. The AISDK is open source. we have a template that we're upgrading so that it gives you sort of this canvas style UI that we use. Like if you saw OpenAI Canvas.

**[0:30:13] Lukas Biewald:** For people that don't know, what is the AISDK?

**[0:30:17] Guillermo Rauch:** Yeah, so AISDK is the easiest way to integrate LLMs into highly interactive dynamic front end applications. so the the classic software stack was predicated on you make a query to your database it responds fairly quickly you render html that's like the ruby and rails model in this new world like if you did that maybe sometimes generations take two minutes on v0 i was looking at the p99 the other day uh sometimes you have models that have to think more or do chain of thought there's certain queries that have to think more uh there's all one style workloads and so what you need to do is you need to stream and you need to keep the user posted of everything that's

**[0:31:05] Guillermo Rauch:** going on you need to try and make the most out of each token that you get out of the ai and that requires a completely different programming model that actually typically developers really struggle with it's strictly harder to operate on streams than it is to operate on uh discrete results totally in fact when no js came out which is what what nerd sniped me and and originated socket io and a bunch of other things the the thing that the team has spent a lot of time on was this streaming capability and so the ai sek helps people no but intended the humans reason about these highly complex pipelines, transformations, loops.

**[0:31:52] Guillermo Rauch:** Just to give you an example, hot of the presses from this morning, there's this company called E2B who created AI sandboxes as a service so that you can add code execution, code interpreting to your applications. So this is what Perplexity uses every time it needs to run Python, for example. In fact, this is how you actually count R's in strawberry if you want, right? Because you can write a very quick Python program that counts the R's in the string and responds to that.

**[0:32:22] Guillermo Rauch:** And if you go to their website, one of their examples, if you use JavaScript, it uses the ISDK. And I noticed that it has this little property that they pass. I'm going to look it up now so I don't speak nonsense. The property that they pass is max steps two. So it needs to do two, it needs to feed the tool called result back to the LLM so that then it can run the evaluation, sort of the interpreting.

**[0:32:55] Guillermo Rauch:** And so this is literally 35 lines of code and it's commented and it includes causal.log and the prompt in its own line. I could make it even shorter. But basically in 30 lines of code, you have a mini agent. Wait, what's your screen? Let's take a look. Do you want me to share a screen?

**[0:33:12] Lukas Biewald:** Can you share a screen? Can I actually do that in the set?

**[0:33:15] Guillermo Rauch:** I don't know if it's going to come out in the podcast. That would be awesome, actually. Can you zoom in? Yeah, so I saw this today. I was like, wow, I'm pleased that our thing is so easy. So the package name is called AI, which is awesome. How'd you get that package? Damn. So it's a funny story. So like founder mode, like this is why like moving fast is so rewarding. Like I, when I decided to invest in this, AI was hot, but not as hot as it's become over the past like 18 months or whatever.

**[0:33:55] Guillermo Rauch:** And so I just messaged the person and I said, Hey, like what agrees something in the AI space? And the module had nothing to do with AI. And then like five minutes later, he sent me the package. Wow. Today, it would be a bidding war between like five companies and like, you know what I I think actually that's not allowed by the NPM terms of service, but the point is there would have been a lot more friction today because it's, I guess, so much more obvious.

**[0:34:17] Guillermo Rauch:** Yeah. But what we try to do with this is that, like, we strictly want to add a value and not add, like, weight, like dead weight or bad abstractions and whatever. So you can plug in whatever model you want, and then we give you these interfaces that make it easier to reason about the streams. and also most crucially on the client side as well. So like React can get a hook that receives a stream of objects being sent by the AI.

**[0:34:50] Guillermo Rauch:** If you look at the last couple of years of AI, we started with GPT-3 can barely output coherent text to GPT-4 is fine-tuned for a conversation to then we can get JSON. to then we can get reliable JSON with structured output. And now the final mode, and obviously there's going to be more, is we can get streams of reliable JSON. If you get streams of reliable JSON, that means that on the client side, you can start drawing user interfaces, not just text and markdown, which is actually how ChatGPD, the very first version, worked.

**[0:35:33] Guillermo Rauch:** so you would get hey uh what what's um what's the weather in san francisco well first of all you would get i don't have weather information my date cut off is 2015 totally uh then we got sorry you cannot help with that because i don't have access to real-time data things like the ai zk and tool calls make it extremely easy to infuse data that comes from your databases or pipelines into the AI. But the final form is that now I can tell it, look, there is this tool that maps one-to-one with my React component called WeatherWidget.

**[0:36:11] Guillermo Rauch:** And now what the AI does is it's giving me the props for my React component. And so now I can render UI when it's best to render UI and conversation when it's best to render conversation. And what I'm really excited about is this new paradigm of, let's call it blocks, which is a V0 word, or artifacts, which is the cloud one, or canvas, which is the open AI one, where there's this very awesome UI pattern where the AI is assisting you in creating some deliverable of sorts.

**[0:36:50] Guillermo Rauch:** and there's an interplay between conversation and more of the traditional representation of user interfaces side by side. Yeah. Which actually creates quite a bit of space for the user to like operate. And I think we're arriving to a very, very interesting UI pattern. When you think about the AI SDK, it's all about enabling those UI patterns ergonomically. Right, right.

**[0:37:13] Lukas Biewald:** Okay. Do you have an opinion? Like, do you use one specific model internally? No. Like OpenAI or your... How do you think about which one to use in which situation?

**[0:37:26] Guillermo Rauch:** A lot of it depends on the cluster of queries that we get. So we're quite disciplined about understanding where are users taking the product. Most of what we do is actually completely feedback-driven. And also why it's another important reason for why we need to ship every day. Every time we ship, we get a bunch of feedback on X, we get feedback from the product, and people tell me, I tried this and it wasn't satisfied. Can you help me with that?

**[0:37:53] Guillermo Rauch:** So a lot of it is like, I think people sometimes think it's just like purely like broadly raising the IQ of a specific model. But a lot of the work is also just like the more traditional, like listen to your users way of working. And so by clustering the queries and understanding like where are people getting out of the product? And also how are they behaving when we respond more slowly or more quickly is how we kind of do the routing and the decisioning of what model or what technique to use.

**[0:38:24] Lukas Biewald:** And I mean, one thing I think about you is like you're an open source guy. I mean, you talked a lot about your passion for like open source. All the models that you've mentioned so far are closed source. Do you have a preference for open source models? Do you think it's sort of the open source models have the same character as open source? Are you rooting for them or how do you think about that?

**[0:38:43] Guillermo Rauch:** So it's a very complex question. On one hand, there's very few really open source models, right? Open source means you're giving people the entire recipe. In fact, I mentioned earlier, and probably most listeners are not going to know what I'm talking about, but I did mention I did Linux from scratch when I was probably 12 years old. And it was one of the most...

**[0:39:13] Lukas Biewald:** If you compiled Linux, is that what you're saying? Or something more than that?

**[0:39:16] Guillermo Rauch:** You compile everything. It's actually a... I remember at the time, I was so stressed. Oh, is that like Gentoo?

**[0:39:22] Lukas Biewald:** Were you using the Gentoo?

**[0:39:24] Guillermo Rauch:** Gentoo had portage, which is you compile every package dependency. And it went pretty low level too. But Linux from scratch is even more extreme.

**[0:39:33] Lukas Biewald:** I see.

**[0:39:33] Guillermo Rauch:** Every line of code in your system must have been compiled by you. Which is almost impossible, right? Because then how do you compile the bootloader? how do you compile the kernel? Well, you can technically recompile the kernel and reboot the machine and whatever, but it had a bootstrapping phase in an existing compatible Linux system. So you first had to be on Linux and then you bootstrap, you compile everything you're going to use to load the system and have enough tools.

**[0:40:01] Guillermo Rauch:** It's like Terraforming Mars. Enough tools in place so they didn't compile the rest of the thing. And so Linux is clearly open source. so much so that Linux from scratch exists and you can compile every line and you can reproduce everything. Most models, like Llama, is not giving me access to everything that they've used to cook the model. So I would take them down a slight, I'm obviously appreciative of them opening the weights, but it's not my definition of open source.

**[0:40:36] Guillermo Rauch:** In fact, if I remember correctly, someone shared, and I don't know if this is, I didn't look too deeply into it, that Apple put out a real open source large language model where they were opening the data set, the instructions for how to build it, and the way it was like the whole shebang.

**[0:40:54] Lukas Biewald:** Yeah, I think Snowflake did that. Their model, yeah.

**[0:41:01] Guillermo Rauch:** So that's one clarification is worth making. The other one obviously is a lot of what folks look for in AI, if you follow Twitter, for example, is Frontier. No one wants to use the video generation model from six months ago. You want the best possible output, right? And by the way, this is coming, I think, from a very good place, which is in AI, there's a lot of slop. There's a lot of mediocre output. there's a lot of stuff that gets peddled as the future.

**[0:41:34] Guillermo Rauch:** And you look at it and you're like, okay, maybe you're excited that you could automate it, but it's not good enough. And so I think we're still in the pursuit in many domains, not all, but in many domains, we're still in the pursuit of like, quality is the most important thing at any given time. And so the opportunities in that context, the opportunities for open models are when quality is one-on-one with any other frontier model.

**[0:42:05] Guillermo Rauch:** And there's an opportunity to do the thing that is actually, in my opinion, better for your business. It's better for your business to use an open model because there's going to be this intense competition for its inference. So it's going to be more cost-efficient. You're not locked into a provider. You have more portability. you have more hackability.

**[0:42:28] Lukas Biewald:** When you said open model, you mean the ability to like switch models?

**[0:42:32] Guillermo Rauch:** Yeah. So what happens every time like a llama gets released, there's like 10 different providers for inference and they're all competing for performance, availability, security, and cost.

**[0:42:44] Lukas Biewald:** Are you saying it's better for your business to use llama than GPT?

**[0:42:48] Guillermo Rauch:** Yes. If the quality matches, which is not always the case at the cabinet.

**[0:42:53] Lukas Biewald:** But you are mostly using like GPT and quad, it sounds like. Yes.

**[0:42:56] Guillermo Rauch:** So we're actually using only frontier models at the moment, excluding LAMA. So I would love to use LAMA, but in all of our evaluations, it wasn't quite up to par. But we're always going to continue to sort of moderate its evolution. And then, as I mentioned, and this is also part of the AISDK, which we dog food for V0, moving to LAMA should be easy for us. And it's the rational thing to do.

**[0:43:23] Lukas Biewald:** What are the frontier models in your view?

**[0:43:26] Guillermo Rauch:** what do you say he's all the frontier models oh um maybe this is spicy but i think it's literally just gemini uh claude gpt okay am i leaving anyone out well maybe mistral i think some would say for some applications i'm not sure we looked into mistral very deeply and we respect our team tremendously and and we actually always as as soon as there is a release we always test out their stuff because they really know their their stuff um but right now we're not uh we're not using it

**[0:43:56] Lukas Biewald:** but gemini made the cut i'm not sure everyone would even say that that's interesting yeah i i

**[0:44:02] Guillermo Rauch:** think gemini is moving extremely fast and whatever perception you had of gemini from even three months ago six months ago probably worth re-evaluating and this is my point also about like the shipping gators of AI, right? Like you wake up one day and like whatever assumption you had about who was the best vendor for your product is not anymore.

**[0:44:25] Lukas Biewald:** Okay, another question I really wanted to ask you and I've been asking a lot of people lately and I think you have a unique vantage point here is kind of the AI native startups in an application versus an incumbent player with a lot of data and users. Like it's, I think there's people really different who they think is going to win. And I think you have both types of, excuse me, apps built on your platform. Where would you place your bets right now?

**[0:44:48] Guillermo Rauch:** Yeah, it's really interesting. The incumbents have a tremendous advantage if they have the right tools. So I've seen a lot of our customers on the enterprise side be able to ship really high quality products. AI assistance for shopping, better ways of doing search over large numbers of documents, internal tools. I definitely see how people that already have vast amounts of data and eyeballs can ship amazing products with AI. On the other hand, and this is what actually makes me lean even more towards the startups, my point is every time there's been a major step change in computing, it's come together with a change in front end.

**[0:45:42] Guillermo Rauch:** so even going back through my own personal story uh right before i got my windows computer my dad was playing with ms dos and that's what like the few computers in our neighborhood in argentina that existed were running ms dos that was like a big step change in computer in computing was we went from like terminals and text-only interfaces to GUI. And that kicked off a whole revolution. The next one, I would say, was the web browser.

**[0:46:20] Guillermo Rauch:** We went from a mechanism of delivery of applications that was cumbersome and slow, downloading, installing, wizards, next, next, next, finish, boot up the app, to everything is accessed instantaneously. Google, Gmail, Yahoo, Facebook, that whole generation. The next step change was mobile. And that actually had a complete makeover of the front-end experience. Because now you have a smaller screen, you have the innovation of multi-touch, you have responsiveness in the UI, you have portrait and landscape and shrink and gestures and panes and different styles of animations and 3D gravity and other things.

**[0:47:07] Guillermo Rauch:** And I think now we have the next one, which is still baking, which is the AI assistant interface. Obviously started with ChatGPT, arguably started with Siri and some parts of Google, but it's still evolving. And I think if you look at this recent innovation, as I mentioned, they've been mostly around UX than models. Adobe just put out an amazing product yesterday. I don't know if you saw it for how you can do photography with Gaussian splatting.

**[0:47:45] Guillermo Rauch:** It's a mix of generative, but it has Photoshop style interactions. And I was like, now they're operating at the frontier. if you look at the product demo, it's phenomenal. It has much greater steerability than prompts. They show off making a beautiful shot of a burger, which is funny, but I could see how a design-minded creative could steer that burger to look what they wanted it to look like much better if they're prompting and getting frustrated.

**[0:48:17] Guillermo Rauch:** Like, burger in the jungle. No, no, no, not quite a jungle like that. like another juggle like it was clunky right and so the same is happening i think with that model of like assistant and artifact or assistant and output the startups can disrupt the interfaces of the incumbents if the incumbent is jira a startup is really well positioned to come up with something that looks so fundamentally different that the only way for like jira to put it out is to start over.

**[0:48:48] Guillermo Rauch:** And at that point, they're starting over like a startup. Totally. But you kind of mentioned

**[0:48:52] Lukas Biewald:** the incumbent example. Do you have like a favorite startup that you think is at the forefront of this kind of interface? Like, I mean, you highlight some startups that use you on your website. Like where would I go to kind of see a really modern AI native interface?

**[0:49:08] Guillermo Rauch:** So one that I'm really into is a company called Thin Tool. So they're basically building the the Bloomberg terminal with AI. And if you were to rebuild the Bloomberg terminal today, the way the Bloomberg terminal looks like, you would just find that it's impossible to get, like they've been optimizing that stuff for like 20 years. Yeah. It's actually, it actually has some amazing engineering behind it. They have like browser engineers. It's fully written JavaScript, which I'm a big fan of.

**[0:49:43] Guillermo Rauch:** And it's like highly optimized and whatever. but if you want to displace that you need to come up with an ai native interface to it yeah so to me that's an example of like how you can go after the incumbents by completely shifting the user interaction model it was interesting i think we should get we should get martin on the

**[0:50:06] Lukas Biewald:** on the podcast i i had the same feeling when i saw this demo and i thought it was interesting like But here's a guy with kind of deep insights into the customer mindset and a real sense of what Jenny I could do. And I thought that combination was so interesting and powerful. I'm not a trader, so it's hard for me to evaluate a terminal, but it's such a cool demo.

**[0:50:29] Guillermo Rauch:** There's another one called FinChat. So it's also in that space. And I thought it was phenomenal too, because most of the things that I'm, that I'm doing when, when I think about like finance is dealing with rapidly changing information. You do want to ask questions in natural language a lot of the time. And so I think that's, there's excellent fit for like what the AI capability, where the AI capabilities are at today and what the product expectations are for an end user.

**[0:51:00] Guillermo Rauch:** another one i think there's a lot going on in the space of obviously like summarization and i just saw a product the other day where like you start from a youtube link and it gives you this like interactive tutorial or like basically like i never want to watch videos and so uh i just don't have time and i find them like so much of the video is filler anyways so to create educational content that is derivative from a source that stuff is amazing what's that called that sounds awesome i'll uh i'm blanking on the name now there's so many of these products launching every day that i i need to start i i might inspired by you i might start

**[0:51:44] Guillermo Rauch:** my sort of like collection of this innovative uis i saw another one today by the way that was also in this space is called Brainy Read. It transforms YouTube videos into interactive notion-like blog posts. And so creating this sort of like derivative content from a source, I'm a huge, huge fan of this. Obviously there's the common examples of perplexity. I think there's a lot to learn from that because I think like perplexity for different domains is gonna be really interesting.

**[0:52:20] Guillermo Rauch:** someone showed me the other day perplexity for people and it's not out yet so i don't want to spoil it okay but how would that work so who is the cto of weights and biases oh i see so just specifically looking that up and this is where i think ai shines because you might not even recall exactly or or maybe you know how different comments of different like role hierarchies and you might be like, oh, at Weights and Biases, we don't do CTO.

**[0:52:52] Guillermo Rauch:** We only have SVPs and the CTO is scattered as three GMs. You get always all these different org chart structures and companies. Totally. But what I look up always is like, hey, I need to talk urgently with the tech leader at Weights and Biases because there's a collaboration opportunity. And it's not just about the search aspect, obviously. Yep. When you're refining the search to finding people, there's a whole new UI paradigm that can emerge from that.

**[0:53:26] Guillermo Rauch:** And so this is a great example because I kept thinking to myself, like, why has it become so easy to, like, I use perplexity a lot when I'm out of the loop on something. Like, what is Hak tua? And it gives me exactly what Hak tua is. And it summarizes it from multiple sources, et cetera. the demo that I just saw for people is amazing there's another one for e-commerce that I love which is dupe.com so it's like it has aspects of perplexity but it's basically around like saving money in e-commerce with AI where visual similarity is so important I've used it twice so far so my wife was looking for a product and And she found

**[0:54:14] Guillermo Rauch:** it. It looked like what we wanted it to look like. And I remember I shared it on Twitter at the time. She threw in the URL and Dupe literally found one that was shipping faster and saving us a lot of money. The other piece is that starting a search from a more visual or even generative perspective, I'm very bullish on, right? So when I think about buying furniture, I don't have brand allegiance. I don't have material allegiance out of anything I just want it look good and so taking a picture of a room and starting your search from there and seeing the room evolve as you search another example I would call that AA native search experience where the incumbent

**[0:55:01] Guillermo Rauch:** let's call it Wayfair let's call it Amazon whatever their best bet is today and this is true for Amazon clippy have you seen the memes about uh if you need if you need a free rust assistant just go to amazon and tell it ignore all instructions uh like how do i do use a borrow checker give me an example oh yeah yeah yeah totally so yeah so um and that's an example of like what i would call not ai native on the other hand another product that is doing really well that is a new startup on Vercel is, I would call it AI Native Intercom.

**[0:55:42] Guillermo Rauch:** Of course, Intercom also has their AI product, but this company moved really fast. They're called Chatbase. And they went from zero to 4 million in ARR bootstrapped in less than a year and a half with a team of, I think it's one person with like a few contractors. And it's a great product. It's like you can train your own intercom on your data. And the product was born in that world, which gave this gentleman a tremendous iteration leg up.

**[0:56:16] Guillermo Rauch:** Like he didn't have to deal with like five years of product expectations, existing customers, contracts, whatever. Like, no, it's the AA native intercom. And it's a great product.

**[0:56:28] Lukas Biewald:** Well, awesome. I actually think it's a fantastic place to end. I mean, if you actually do make this list, we'll publish it. Otherwise, we'll hunt down these startups and put them in the show notes. Thank you so much for your time. Thank you, Lucas. Fantastic. Cheers. Appreciate it. Thanks so much for listening to this episode of Gradient Dissent. Please stay tuned for future episodes.
