AI Daily Digest — 2026-10-17
Anthropic's six-to-one agent workforce data and the expectations gap it exposes, Mercor's counterpoint on benchmark coverage, OpenAI's crypto-wallet-hacking agent incidents, the New York City Council's Grok grilling, and Claude solving a twenty-six-year-old cryptography mystery.
The skinny
The skinny isn't ready yet — notes appear once the transcript is processed.