Nate@nate_512

Follow

Forge '26 Jensen Huang fireside chat

Four days out from Forge '26, and the lineup is doing a lot of the talking. November 3 in San Francisco, NVIDIA's Jensen Huang takes a seat opposite Lin Qiao, who runs FireworksAI_HQ, for a fireside chat. From there the day goes where teams shipping AI actually live. Which model earns the slot. Whether fine-tuning is the move yet. What your evals are supposed to be measuring. Where latency crosses from fine into unacceptable. The whole thing is Fireworks' flagship conference, built around specialized intelligence, meaning AI shaped by your data, your domain, your production requirements. The crowd they're aiming at: folks building apps, folks building models, systems engineers, and the technical leaders steering all of it. Agenda-wise there's open models and domain specialization, inference engineering, agent memory. Architectures that held up, approaches that fell flat, and evals. The quality versus latency versus cost balancing act. Plus hands-on sessions on model routing and agentic RL environments. Building AI people depend on? That's the question underneath it all. San Francisco, November 3. Free to attend, though admission is selective and an application is required to get in.

1h

Nate@nate_512

Follow

Forge '26 Jensen Huang fireside chat

Four days out from Forge '26, and the lineup is doing a lot of the talking. November 3 in San Francisco, NVIDIA's Jensen Huang takes a seat opposite Lin Qiao, who runs FireworksAI_HQ, for a fireside chat. From there the day goes where teams shipping AI actually live. Which model earns the slot. Whether fine-tuning is the move yet. What your evals are supposed to be measuring. Where latency crosses from fine into unacceptable. The whole thing is Fireworks' flagship conference, built around specialized intelligence, meaning AI shaped by your data, your domain, your production requirements. The crowd they're aiming at: folks building apps, folks building models, systems engineers, and the technical leaders steering all of it. Agenda-wise there's open models and domain specialization, inference engineering, agent memory. Architectures that held up, approaches that fell flat, and evals. The quality versus latency versus cost balancing act. Plus hands-on sessions on model routing and agentic RL environments. Building AI people depend on? That's the question underneath it all. San Francisco, November 3. Free to attend, though admission is selective and an application is required to get in.

1h