Guillermo Rauch says he became a core developer of the MooTools JavaScript library at about age 16.
“at about 16 years old i became a core developer of this uh library called mood tools”
Person
This page collects what Guillermo Rauch said on the Gradient Dissent podcast about AI and web development. He is the CEO and founder of Vercel, and the episode covered front end engineering, the v0 product, and open source AI models.
Rauch says he became a core developer of the MooTools JavaScript library at about age 16 and created Socket.io. He says Next.js has 1.2 million monthly active developers, and that v0 produced more UI generations in one month than in the previous twelve combined. Each fact below gives his exact words and a link to the moment.
1 episodes · 25 quoted claims
Guillermo Rauch says he became a core developer of the MooTools JavaScript library at about age 16.
“at about 16 years old i became a core developer of this uh library called mood tools”
Guillermo Rauch says he created Socket.io, a library for real-time communication that runs on both client and server.
“i created a library called socket.io for real-time communication the parser the protocols a lot of things”
Guillermo Rauch says a book he wrote on universal JavaScript helped him obtain a US O1 visa.
“which helped a lot with my immigration story in the US, getting my O1 visa.”
Guillermo Rauch says Next.js is the most popular React framework, with 1.2 million monthly active developers and 7 million weekly NPM downloads.
“It's the most popular React framework, 1.2 million monthly active developers, 7 million weekly NPM downloads.”
Guillermo Rauch says AI companies such as Anthropic and Perplexity are standardizing on Next.js for their web applications.
“a lot of those folks are sort of standardizing on Next.js as a framework above which they build their innovations.”
Guillermo Rauch says his first startup, in education technology, had a great mission but was not in a great market.
“It was a great mission, but it wasn't a great market.”
Guillermo Rauch cites the search company You.com as hosted and built on Vercel, as an example of a mission-critical customer.
“Your.com is hosted and built on Vercel. It's absolutely mission critical.”
Guillermo Rauch says v0 produced more UI generations in a single month than in the previous twelve months combined.
“we did more UI generations than the previous 12 months”
Guillermo Rauch says the v0 team ships every day and posts a daily changelog on its Twitter account.
“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 says v0 introduced Teams and Enterprise offerings shortly before the interview.
“The latest one was we introduced V0 Teams and Enterprise yesterday.”
Guillermo Rauch says one v0 improvement used a Gemini capability to produce and apply code diffs very quickly.
“this is using a capability from Gemini to do very, very fast code diffs and applications of code diffs.”
Guillermo Rauch says a v0 deployment takes about a minute to build and deploy with zero downtime.
“each time we deploy it takes about a minute to build deploy with zero downtime”
Guillermo Rauch says v0 spreads its GPU usage across three data centers because of the volume of tokens it generates.
“We actually now is spreading our GPU usage across three different data centers because we're generating so much, so many tokens.”
Guillermo Rauch says a major part of v0's evaluation is testing in production behind feature flags and watching product behavior.
“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 describes E2B as a company offering AI sandboxes as a service for code execution, used by Perplexity to run Python.
“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 says he messaged the owner of the unrelated npm package named 'ai' and received it about five minutes later, giving the AI SDK its package name.
“And then like five minutes later, he sent me the package.”
Guillermo Rauch says a key recent AI capability is streaming reliable structured JSON, which lets clients render interface components as data arrives.
“now the final mode, and obviously there's going to be more, is we can get streams of reliable JSON.”
Guillermo Rauch gives the example of a tool mapped to a React component called WeatherWidget, so a model returns props and the app renders UI instead of text.
“look, there is this tool that maps one-to-one with my React component called WeatherWidget.”
Guillermo Rauch says Linux From Scratch requires that every line of code in the system be compiled by the user.
“Every line of code in your system must have been compiled by you.”
Guillermo Rauch says Meta's Llama releases only the weights and not the full training recipe, so it does not meet his definition of open source.
“Most models, like Llama, is not giving me access to everything that they've used to cook the model.”
Guillermo Rauch says he recalls Apple releasing a fully open-source language model with its dataset and build instructions. Host Lukas Biewald responds that Snowflake did something similar.
“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 says v0 uses only frontier closed models because Llama fell short in Vercel's evaluations.
“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 says the few computers in his Argentine neighborhood when he was young ran MS-DOS.
“the few computers in our neighborhood in argentina that existed were running ms dos”
Guillermo Rauch praises a new Adobe product that combines generative AI with Photoshop-style controls using Gaussian splatting.
“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 describes the startup Thin Tool as building a Bloomberg-terminal-style product with AI.
“they're basically building the the Bloomberg terminal with AI.”