The AI Frontier: from open weights to open research — Eiso Kant, Poolside AI
- Source
- youtube.com
- Author
- Latent Space
- Date
Context
Eiso Kant, co-founder of Poolside, explains to Latent Space why the company released the weights of a new . Poolside started on two premises: capabilities would keep compounding, and reinforcement learning, by rewarding good outcomes, would drive them. Open source did not come up again until early this year. Kant and the co-founder then concluded they would rather live in a world with a hundred foundation-model companies than five, even if Poolside were one of the five, and that opening research and weights was the smallest meaningful contribution they could make toward that.
Kant says big questions remain open: how to build a business on open models, when misuse risk should stop a release, and how governments will respond. This is a founder's rationale, not evidence about the model's quality or the effects of open release.
- Model — A computational system that uses learned patterns to produce an output from information supplied to it.
- Training — The process that adjusts a model using examples and a measure of how well it performs.
- Model Context Protocol (MCP) — A published agreement for an AI application to discover what a separate program offers and ask it to run one of those offerings.
- Tool — An operation that the software around a model makes available for the model to request, with a name, expected inputs and a kind of result.
videoExtreme Harness Engineering: 1M LOC, 1B toks/day, 0% human code or review — Ryan Lopopolo, OpenAILatent Space
videoNotion’s Sarah Sachs & Simon Last on Custom Agents, Evals, and the Future of WorkLatent Space
videoThe Inference Frontier: from 100 to 10,000 tokens per second — Sean Lie, Cerebras CTOLatent Space
Checking sign-in…
Loading comments…



