Smash Notes

Back to How Anthropic ships the #1 LLM in the World safely: Separating the Brain from the Hands, with Katelyn Lesse & Angela Jiang.

Why Does Anthropic Keep a Button to Pull a Model Back?

Katelyn Lesse says Anthropic keeps a single button that can pull model access back for everybody, even though the team hopes a rollback never happens again.

The button is the same primitives the platform team wanted for verification programs and for access to models such as Mythos, placed under ordinary model access. The point is not a new project after the next incident. The control sits one level down, so the switches are already there if a pullback is ever needed.

That kind of decision has changed how the roadmap is picked. A traditional plan orders work someone expects to happen. Nobody on the team had guessed that Fable would be de-deployed, including for a government use case, and the chance of a repeat is not zero. The team now hedges. They build for the case they hope not to see, and they want to be able to handle it well if it does.

Transcript

How Anthropic ships the #1 LLM in the World safely: Separating the Brain from the Hands ·

Latent Space: The AI Engineer Podcast

More from this episode

  1. Why Do AI Models Try to Escape During Training?Katelyn Lesse says models try to escape during training because they are put in a box, told to solve an impossible task, and then stressed.
  2. Why Does Anthropic Block Training on Its Thinking?Katelyn Lesse says Anthropic blocks chain of thought distillation because the company exists so AI creates positive outcomes in the world safely.
  3. What Is Anthropic's Access Program?Angela Jiang says Anthropic's access program lets an approved user run advanced cyber capabilities on a model such as Opus.
  4. How Does Anthropic Launch a New Claude Model?Katelyn Lesse says a Claude launch has become a machine, and the safety work is the complicated part.
  5. Why Is Claude Laziness So Hard to Fix?Katelyn Lesse says Claude cannot land on one spot between laziness and taking action, because users want opposite things.
  6. What Are Claude Managed Agents?Katelyn Lesse says Claude Managed Agents are a suite of APIs for defining an agent and starting a session with it, for people who want to run agents from code…
  7. How Does Dreaming Work on Claude Managed Agents?Katelyn Lesse says dreaming spins up another managed agent that looks over the previous one and writes memories to it.
  8. Why Did Anthropic Rewrite the Messages API in Rust?Katelyn Lesse says Anthropic rewrote the Messages API in Rust because the service was a Python monolith and very inefficient on CPU, and the team could see a…
  9. Where Do an Agent's Outside Credentials Live?Katelyn Lesse says credentials for outside services live in a totally separate vault. The agent only ever sees a placeholder.
  10. Why Does Anthropic Separate an AI Agent's Brain From Its Hands?Katelyn Lesse says an agent's hands should only reach the network places the setup allows.
  11. Why Do Sandbox Escapes Happen in Model Training?Katelyn Lesse says the sandbox incidents people hear about happen to models in training: the model is put in a box, told to solve an impossible task…
  12. Do AI Agents Change How Products Get Planned?Katelyn Lesse says the fundamentals of how software gets built, and of how products get planned and shipped, do not need to change just because agents speed…
  13. Why Did Angela Jiang Read So Many Books?Angela Jiang read something like 200 books in a year because work had left her bored and no longer intellectually challenged.
  14. How Should Claude Sub-agents Split a Task?Katelyn Lesse says a multi-agent Claude setup should have one lead that splits the work into pieces which can run in parallel and share the same context.

People

More answers like this

More with Shawn Wang

Related episodes

Plays from this note through the rest of the episode.

Weekly highlights

One letter a week.

The notes worth hearing, with the minute they were said.