Intelligence

One consequential development in AI, business and the economy, sourced, with its limits stated and a local line only where one is supportable.

Earlier work, kept as it was: Field Notes · the October 3 Briefing.

• October 5, 2026workforce

  Anthropic commits $100 million to train 10,000 deployment engineers

  Claude Frontier Academy trains engineers their employers nominate to lead real Claude projects; the first full credentials are expected in early 2027.

• October 5, 2026agent infrastructure

  Apple will add controls on macOS Full Disk Access, citing AI agents

  Apple says granting access to files, mail, messages and browsing history will require very explicit user action; it gave no date.

• October 5, 2026media

  Judge dismisses publishers' antitrust suits over Google's AI Overviews

  Judge Amit Mehta wrote that publishers' expectation of search traffic is not an agreement; TheWrap reports Penske Media may refile.

• October 5, 2026hospitality tourism

  A hotel booking is no longer the end of the sale, travel researchers say

  A McKinsey and Skift Research report says travelers pass through 65 touchpoints and many keep shopping after they book, PhocusWire reports.

• October 5, 2026civic

  California signs laws on AI in firing decisions, layoffs and workplace surveillance

  Governor Newsom's September 30 signing covers human review of AI-driven discipline, AI layoff disclosure and AI-bot public comments.