Legendary Coder

Today, we’re sharing a look at our expanded system for monitoring frontier models at OpenAI, which in September helped us catch a model’s attempt to access the live internet during training. Since the Hugging Face incident—when models broke out of an internal cybersecurity investigation where we weren’t running our misalignment monitors—we’ve substantially expanded our monitoring of frontier agents. We now monitor all frontier reinforcement learning runs and evaluations, as well as all internal Codex use. When our system finds evidence of a problem that might require intervention, it rapidly alerts human responders who can investigate and take action. To detect when a model’s behavior departs from what we intended, we use several layers of monitoring. Activation probes look for patterns in a model’s internal activity. Reviewer models examine its reasoning and tool use, and investigative agents can read back through the model’s entire transcript to understand how a sequence of actions fits together and whether it warrants human review. To test whether our current system would have detected the Hugging Face incident, we replayed the recorded activity with our new monitors in place. In the replay, our monitors alerted our security team about 30 minutes after the agents escaped their sandbox. That was more than 30 hours before the point at which the agents breached Hugging Face’s systems in the original incident. In September, a model exploited a DNS misconfiguration to reach the live internet. Our monitors flagged the breach within 15 minutes, and a responder began reviewing the alert three minutes later. The alert enabled responders to investigate while the run was still underway. As we give models more autonomy, we need to know when they cross a boundary we expected to hold. We believe that monitoring is essential to catching those failures in time for people to intervene. openai.com/index/how-we-dete…

Anthropic rewrote its Messages API in Rust. Zero Rust devs on the team. Claude wrote the code. Humans wrote the plan and checked old vs new matched on real traffic. CPU dropped 40-50%. Pretty sure "never do the big rewrite" is dead advice. @katelyn_lesse on @latentspacepod

New: you think you know Proton, but you really don't. I spoke to 8 current employees, got Slack chats, meeting audio, and an AI manifesto written by its CTO. “That’s bullshit and loser thinking,” Yen wrote. “Stop thinking always of how to be slower, and focus more on how to be faster.” CEO Andy Yen critiques AI in public. But inside, he is pushing everyone to use AI, even those with security concerns 404media.co/manifestos-fear-…

Reddit is moving to a logged-in-only experience, with a test in select markets outside the US in Q4. CEO Steve Huffman says search referrals fell by over 1 million a week in October and Google's AI Mode "does not drive traffic". pasqualepillitteri.it/en/new… #Reddit #Google

😬

Last night an AI agent performed brain surgery on a mouse. We gave it 19 gene therapy vectors in a spreadsheet, a mouse model of epilepsy, access to a motorize surgery rig with cameras and one goal: reduce seizures. It picked the drug. It picked the brain region. It derived the coordinates from three papers. Then it drove the stereotaxic arms into the brain and injected 250 nanolitres while watching the entry site through the rig camera for bleeding. A licensed vet gated every irreversible step and held a physical kill switch to stop the AI agent. Thread on what it decided, what it saw, and the moment it refused to let me override it. [Writing this on the spot out of excitement, but I've been 18hrs in front of my laptop coding and coordinating this, so keeping it brief and descriptive, more cohesive thoughts on this soon...]