Ben Thompson's read on the Apple event: the iPhone Duo and the new Watch audio features show Apple's hardware-software integration still works, but its AI strategy is hobbled by the assumption that apps remain the center of the experience.
Anthropic safety researcher Jacob Coxon publicly quit, saying self-improving AI could kill everyone, and a colleague, Evan Hubinger, said the company earnestly believes there is a greater than 10% chance of that within a decade. The posts landed in the middle of Anthropic's IPO marketing and by this morning had lawmakers calling for new AI rules.
Automattic's board voted to put founder Matt Mullenweg on paid leave against his will and named CFO Mark Davies interim CEO. Mullenweg told all employees on Slack that the board 'conspired' behind his back and gave him 50 minutes to review the resolution before the vote.
Tyler Cowen flags a new paper from Anthropic's economics team (Korinek, Jones, Sacher, Cotter, McCrory) laying out a framework for the economic consequences of AI between 2026 and 2030, with scenarios he calls sober, scientific, and mostly dynamically consistent.