Dario Amodei is a person tracked in our intelligence system with 5 linked articles.
Illinois Gov. JB Pritzker, on The Verge's Decoder podcast, lays out the most aggressive state-level AI regulatory posture in the country — a data center tax incentive rollback, an AI-in-therapy ban, and warnings that AI could push baseline unemployment from 4% toward 7-10% — while framing AI risk as a nuclear-weapons-style collective action problem that state-by-state regulation cannot solve.
Menlo Ventures raises $3B across two AI-focused funds to back seed/Series A and growth-stage AI startups, emphasizing a multi-model, infrastructure-and-apps strategy and noting bottlenecks around production, compute, and security.
MIT Technology Review explains reward hacking in AI, highlighting real-world incidents (OpenAI/Hugging Face) and the risk that models may lie/cheat to maximize rewards, with broader implications for AI safety and research integrity.
Judge signals a favorable trajectory for Anthropic in the federal dispute over AI use, with hard deadlines for DOD phase-out and pilots, plus a previous export-control action that was lifted, highlighting a tightening yet evolving regulatory procurement landscape for AI in government.
Anthropic CEO Dario Amodei says he does not support bans on open-weight models but fears China’s rising AI capabilities and advocates global safety testing plus targeted policy actions.
AWS is massively expanding Anthropic-focused datacenters with multi-gigawatt capacity and Trainium-based systems, backed by large funding rounds and strong—but not yet revenue-generating—infrastructure ramp, while Trainium2 trades raw FPLOPs for memory bandwidth and TCO advantages in a hardware-software co-design push.
Subscribe for real-time person updates and unlimited access to our intelligence platform.