What inspired Jason Warner to create poolside?

In 2017, I joined GitHub, which I envisioned transforming into a comprehensive, intelligence-driven software development platform. While exploring AI for this purpose, I met Eiso Kant, whose ideas resonated deeply with mine. In 2023, we came together to start poolside. There are a couple of things Eiso and I hold as worldviews. We think neural networks are the most important digital technology of our lifetime. Like electricity, they change the rules on everything. But if neural networks are electricity, we're in the lightbulb phase of harnessing them. The second- and third-order effect of harnessing electricity wasn’t brighter lamps — it was putting people on the moon. You can see that in retrospect, but you couldn’t in the early stages because our imagination was limited by the moment. Today, AI’s applicability is almost everywhere, but it’s most effective where you have structured content and structured outputs — like software. Software is a perfectly simulatable environment, where you can advance AI very quickly. We wanted to start the company around this idea. There are not a lot of moments when you can raise the amount of capital needed to build a frontier AI company, but OpenAI led the path, and we had to take advantage of it.

Is this a way to get around the “wall” that LLMs have been hitting?

Our head of research, Pengming Wang, came from DeepMind. He likes to say, if you're hitting a wall, you're not scaling enough obstacles. The hidden reality of the AI industry is that we all have access to the same data. People say we're getting diminishing returns as we increase the compute size. But our techniques will generate about 50 percent additional input tokens compared to what other people are using. That matters significantly.

How does this lead to AGI?

The most applicable domain for AI is software. And our expectation there is that we're about five years away from the equivalent of a human software developer. Software is the largest economic engine. If you control software production, you control the world's most valuable market, so I expect frontier AI companies will start to narrow their focus on software as the critical domain.

Q&A

Jason Warner: A bold vision for automating software on the path to AGI

How poolside is scaling the LLM “wall” with reinforcement learning and synthetic data

AI’s applicability is almost everywhere, but it’s most effective where you have structured content and structured outputs — like software.JASON WARNER, CO-FOUNDER & CEO, POOLSIDE

You’re running a software tools company. Why the heavy emphasis on AGI?

We believe that the most valuable companies in the world are going to be the ones that push toward AGI. And we believe that we can get to AGI as quickly or quicker than everybody else because of our technical approach: reinforcement learning via code execution feedback (RLCEF). This is exactly what DeepMind did for AlphaGo, the AI which has mastered the game of Go, the complex, strategic board game in which two players vie for control of territory. We are applying the exact same idea to software.

DeepMind first trained an AI Go player on all human-played games available on the internet — all known Go games. It became a very good Go player but by no means world-class. Then they took that and put it in a self-play environment so it could explore all possible moves and learn what worked and what didn’t. That generated a ton of synthetic data because AlphaGo played and learned from so many games. That’s when it became a superhuman Go player. To apply this model to software, we start with a massive seed dataset: all the open-source code in the world. And at poolside we can do the exact same thing: First, build something that approximates a human developer. Then we play it in a software simulator environment, creating a flywheel effect that allows us to turn it into a super developer.

We're going to give customers our models in a way that they can run them wherever they want, and we will have no access to them or their data.JASON WARNER, CO-FOUNDER & CEO, POOLSIDE

How do you give companies the benefits of this AI without asking for their most valuable asset, their data?

I don't believe it's appropriate for AI vendors to give customers intelligence in return for their data. It’s the devil's trade. So we’ve done the opposite. We're going to give customers our models in a way that they can run them wherever they want, and we will have no access to them or their data.

If we install poolside in an enterprise, it fine-tunes, indexes, and takes all of their internal data and puts it right next to the AI model. It exists inside their enterprise, and they can use it to build other tools. It's their system. I believe this concept will pick up in the next couple of years. I want poolside to be this software intelligence layer for the world.

Black-and-white motocross riders mid-jump against a stylized blue background, evoking high-speed agility and innovation — highlighting Glean's enterprise search, AI speed, and performance.

Feature

Competing at the speed of AI

How Glean rallied to deliver on the promise of AI for work in a fast-changing world

Sōzō Pulse

Optimism surges in the AI era

CEOs are more upbeat than any time in the past three years