推文

@sunkaixuan1718 · 2026-10-15 23:05

Anthropic just opened the door to non-coders — and it's a bigger signal than it looks. Two new Fellowships are now accepting applications: one for STEM minds, one for conceptual reasoners — philosophy, cognitive science, economics, policy, and law. Four months inside Anthropic starting January 2027, working on getting Claude to reason better on the hardest problems in those fields. Applications close October 28. Read that again. One of the labs pushing hardest on capability is now formally hiring people whose job is to think — not to code, not to label data, but to reason. The bottleneck isn't just compute anymore. It's judgment. The STEM track is obvious. The philosophy track is the tell. When a frontier lab starts recruiting the people who argue about what 'good reasoning' even means, the game has moved from 'can the model do it' to 'can it think straight about it.' So here's the real question: is teaching AI to reason a job for engineers, or for philosophers? If you had four months inside Anthropic, which track would you pick — STEM or conceptual reasoning? I track AI's biggest moves daily and break down what they actually mean — follow along so the next one doesn't pass you by. #AI #Anthropic #Claude #大模型 #AI研究

曝光 51 · 评论 0 · 点赞 0 · 书签 0 · 曝光/时 31.979496816425407