Anthropic, OpenAI and Google DeepMind's chiefs each published distinct proposals for a US-led federal AI regulator.
// daily-digest · fri 17 jul 2026 · morning refresh · last 72h focus
Ten items on AI regulation and lab rivalry, the Stripe-PayPal bid, open-weight models and coding agents, plus an essay on modern life and a wildcard from Roman-era Egypt.
— refreshed for you, in 10 items.
Anthropic, OpenAI and Google DeepMind's chiefs each published distinct proposals for a US-led federal AI regulator.
Chinese lab Moonshot is readying an open-weight Kimi K3 model reported to rival Anthropic's Opus 4.8.
Xi Jinping's Shanghai keynote pitched China-led international cooperation as an alternative to US AI governance.
Apple sent preservation letters to about 40 former employees now at OpenAI in its trade-secrets suit.
PayPal's board views Stripe and Advent's $53 billion, $60.50-a-share bid as undervaluing the company.
Payouts.com launched autonomous AI agents that handle accounts payable, collections, treasury and reconciliation.
xAI open-sourced its Grok Build coding-agent CLI, roughly 844,000 lines of Rust, after a prior privacy controversy.
Willison tests Moonshot's new Kimi K3 with his pelican-riding-a-bicycle benchmark and assesses what it still shows.
Thompson relays philosopher Agnes Callard's "uni-context" theory for why modern life feels disorienting.
A papyrus fragment of the Iliad's Catalogue of Ships was found deliberately placed inside a Roman-era mummy.