Scott Wu: prediction

Transcript context

…The founding team, I mean most of us have known each other for years and years and years actually. And for almost everyone, this is our first time working together, but we've known each other a long time. And we all actually had our own kind of journeys in AI for the last decade or so. And so for myself, I ran a company before this called Lunchclub, which was an AI for a professional networking product and I ran that for about five years. And one of my co-founders, Steven was one of the first engineers at a company called Scale AI, which has obviously grown a lot and done very well. My other co-founder Walden, was an early engineer at a company called Cursor, which has also obviously grown a lot and done really well. And our whole team was kind of like that. Many of us knew each other from competitive programming and math competitions, but we had stayed very closely in touch in the decade since then and we've all kind of all had our own journeys. And so we had one person who was running teams at Neuro, we had one person who was at Waymo, someone who had their own YC tools startup for machine learning, and we were really excited to build something together. And this was around late 2023, so about a year and a half ago at this point. And yeah, when we got started, I mean I think there were a couple of things that we felt really strongly about and one was that reinforcement learning was really working and was going to be the next big paradigm shift in capabilities. Back then it was the initial ChatGPT launch in 2022, and those models were, to first order were what we would call imitation learning in AI, right, which is basically you have the model read all the texts that you can find on the internet and then train it to talk like somebody on the internet would talk. Right. And there are kind of obviously a lot more details on top of that, but that's kind of the first order pass of what was really done. And it was amazing. Right. I mean, it passed the churn test, it was able to respond and to have encyclopedic knowledge about a lot of things. And I think this new paradigm which we've gotten into over this last year or year and a half is really high compute RL, which is a very different paradigm, right, which is basically the ability to go and do work on task and put something together and then be evaluated on whether that was correct or incorrect and use that knowledge to decide what to do and to learn from that. Right. And so we felt very strongly that that was going to happen. I think for us, code was the natural thing to work on for a couple reasons. One, because we're all programmer nerds ourselves, and so teaching AI to code is about as cool as it gets for us, but also because code has this whole automated feedback loop, right, where you can run the code and that is the kind of automated feedback that really feeds into the RL, which makes these models so great at coding. se code has this whole automated feedback loop, right, where you can run the code and that is the kind of automated feedback that really feeds into the RL, which makes these models so great at coding. And then the other thing that we felt very strongly about was that the product experience was going to shift from what I'll call text completion to agents basically. Right. And to first order, I would kind of say there's been a lot of great experiences in text completion. It's been used for marketing, it's been used for customer support, it's been used for education and in Coda obviously as well. The GitHub copilot was kind of really the dominant product of that initial wave. Right. But I think that the big shift that we really felt we would see is moving from kind of this text to text model to an actual autonomous system that can make decisions, that can interact with the real world, that can take in feedback, that can iterate and take multiple steps to solve problems. And now we call that agents, but that was what we were really excited about at the time. So it was always coding, it was always agents. And in some ways that kind of feels like it should have been cleared from the start. But even with that, I feel like we've pivoted eight times or something within coding agents over the last year and a half, so. I just noticed recently all the AI, top AI companies sort of, not all, but many of them, the product that is winning is different, has a different name from the company, which is not typical, Cursors, Anysphere, Bolt, StackBlitz, you guys are Cognition Labs, like V0 is Vercel. And it just tells me these all emerge later in the company's journey and they tried a bunch of stuff and like, oh wow, this thing worked and it's so interesting that it's so common amongst these top AI companies.…

Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.