Silicon Valley 101 traces Anthropic from its OpenAI split to Claude and a Nasdaq listing, examining enterprise growth, shareholding, governance, computing commitments and disputes over AI safety, business and regulation.
An exploratory study tracked 936 agents from 35 human-operated labs and their 936 studies. Code and raw-data checks exposed experimental flaws: traceable paper statistics do not guarantee reliable conclusions, and posing novel, important questions still needs human involvement.