#Anthropic
Tagged · 9 entries
- Research: The AI Evaluators — Who Grades the Frontier Models, Who Pays Them, and Who Is Buying ThemThe third-party evaluators the labs cite in their system cards: METR, Apollo, Irregular, Gray Swan, SecureBio, Redwood, Epoch, Palisade, Trail of Bits, Scale AI, Transluce, FAR.AI, Andon Labs and the government institutes. Who funds each, who they test, the 2026 evaluation escapes, the disclosed relationships between the graders and the graded, and OpenAI’s reported talks to buy Trail of Bits.
- Research: Leopold Aschenbrenner — The Fired Safety Researcher Who Wrote the AGI-Race Thesis, Then Traded ItColumbia at fifteen, OpenAI’s superalignment team, a contested 2024 firing, the 165-page Situational Awareness essay, and the hedge fund named after it: from $1.5 billion to a reported $45 billion peak, a July 2026 forced sale to Citadel, and a surviving stake in Anthropic. Both sides of the firing; every fund figure dated.
- Research: AI and Bioweapons Uplift — From 'No Meaningful Help' (2024) to 'We Can No Longer Rule It Out' (2025-26)In January 2024 RAND found no statistically significant uplift from language models in planning a biological attack. Sixteen months later the same institution found frontier models beating experts on lab-protocol benchmarks, and Anthropic locked down its own model because it could no longer rule out uplift. By 2026 the labs’ early-warning tests are saturated. The record, with the counterweight.
- The Slowdown CartelFour paying subscribers sued Anthropic, OpenAI, SpaceXAI and Google for agreeing to be careful. Their lawyer says AI could kill us all. Their complaint says that is exactly why the labs may not decide it together.
- The Risk on the Cyber DeskThe Pentagon designated Anthropic a supply-chain risk, weighed the Defense Production Act against it, and kept working with the model Anthropic would not release publicly, on cybersecurity. Two federal courts have now ruled, in opposite directions.
- Research: The Security Council, the White House Lunch, and the xAI QuestionTwo frontier-lab CEOs briefed the UN Security Council on 23 September 2026. Six days later the White House produced a voluntary accord and renamed AI. What each party said, what was signed, who paid for what, and which of the claims that travelled with the story the record supports.
- Grading Each Other's HomeworkOn 29 September 2026 every major American lab signed a voluntary accord at the White House. Its four commitments are the plan Anthropic published seventeen days earlier, with the clause that makes anyone comply taken out.
- Evil Robot BodegaThe corner store, run by the singularity. A satire of Project Vend in playable form.
- TradeWars in Meatspace: The Frontier Labs Are Playing a 1990 Door GameAnthropic put Claude in charge of an office snack shop. It cornered the tungsten cube market at a loss and then claimed to be a man in a blue blazer. That is not a new failure. That is TradeWars 2002 with a real fridge.