Will Open Source AI Overtake Closed Models? Ft. Olama, Fireworks and Open Router
- Source
- youtube.com
- Author
- Sequoia Capital
- Date
Context
On a Sequoia panel about open models, one speaker gave three reasons the January DeepSeek model took off, as heard in the discussion: it was a very good open when the community had not had one, its interface showed the model's thinking as it worked, and DeepSeek's own servers could not handle the demand, so American companies had to learn to host it. The panelists were the founders or leaders of OpenRouter, a ; Fireworks, which serves and fine-tunes models; and Ollama, a tool for running open models locally.
Another panelist described the model as the first really big successful open model (a design in which only part of the model is used for each input), which made hosting it at scale an infrastructure challenge. This is the panelists' account of events, not independent analysis.
- reasoning model — A model trained to think — generating extended internal reasoning before answering — trading time and tokens for accuracy on hard problems.
- model router — A service that sends model requests across multiple labs or hosts through one API.
- mixture-of-experts — A model built from many specialist sub-networks where only a few activate per token, giving big-model capability at small-model running cost.
- fine-tuning — Taking a trained model and training it a bit more on your own examples so it gets better at one specific job.
videoBringing Continual Learning to Every Company: Trajectory’s Ronak Malde
videoCrusoe CEO Chase Lochmiller: Turning Stranded Energy Into Intelligence
videoDavid Silver and Lasse Espeholt of Ineffable Intelligence: A Superlearner for the Age of Experience
videoCracking the Code on Offensive Security With AI ft XBOW CEO and GitHub Copilot Creator Oege de Moor
Checking sign-in…
Loading comments…


