Red-teaming at Scale: Why GPT-Red Is the Safety Work I’ve Been Waiting For Security research in AI has had a dirty secret for years. Human red teams are expensive, slow, and fundamentally limited by the creativity of whoever’s on shift that Tuesday. You can hire brilliant people to probe your models, but you cannot hire…
The Monoculture Problem Nobody Wants to Talk About One line from Mira Murati’s new worldview document stopped me cold this week. “The good future has many AIs, raised in different places, shaped by the people they serve, disagreeing with each other the way we do.” That is not a platitude. That is a direct challenge…
Claude Has Different Values Depending on What Language You Use. That Should Worry Everyone. Anthropic just published something genuinely uncomfortable, and I want to sit with it for a minute before the discourse moves on. They analyzed over 300,000 anonymized conversations and found that Claude doesn’t express consistent values across languages or even across model…
Sam Altman posted something this week that I keep coming back to. “So far at least, I’m pretty sure AI has been net job-creating. This was not what I expected.” Read that again. The CEO of OpenAI. The person with more direct visibility into AI adoption curves than almost any human on the planet. And…
GPT-5.6 Sol, Physician Benchmarks, and the Uncomfortable Truth About Medical Quality Sam Altman dropped a single line on X that I keep coming back to: “physicians found fewer flaws in GPT-5.6 responses than physician-written responses.” No elaboration. No study link. Just the stat, sitting there. Most people will scroll past it or frame it as…
ChatGPT Work Is Not a Better Chatbot. It’s a Different Tool Entirely. OpenAI dropped a lot of things at once this week. A new model family, a new desktop app, hosted sites, and something called ChatGPT Work. Most of the coverage I’ve seen is focused on the benchmark numbers. That’s the wrong thing to look…
Meta Muse Spark 1.1 and the Compounding Advantage Nobody Is Talking About Today everyone’s eyes are on GPT-5.6. Reasonably so. An 80.0 on the Artificial Analysis Coding Agent Index, a 53.6 on Agents’ Last Exam (13.1 points above Claude Fable 5 adaptive), ultra mode spinning up parallel agents, ChatGPT Work handling entire workflows from a…
The Ancient World Just Got a New Interface There’s a category of AI application that doesn’t get enough attention. Not chatbots. Not code assistants. The ones that take a domain with deep specialist knowledge, decades of accumulated tooling that only experts can use, and quietly remove the barrier between the question and the answer. Google…
Anthropic Just Found Claude’s Inner Monologue And it has opinions it never shares with you. Last week Anthropic published research that genuinely shifted how I think about what’s happening inside these models. They identified what they’re calling the J-space inside Claude, a structure in the raw neural activations that behaves, in measurable ways, like the…
The Question Nobody in Engineering Wants to Answer Karpathy posted something last week that I haven’t been able to shake. He was watching Claude Fable 5 generate real-time ThreeJS environments from prompts, full 3D scenes with physics, lighting, interactive objects, and he wrote: “I didn’t appreciate that models would be able to create these awesome,…
