アップルのMacチップ戦略転換、自動運転車の「失敗」がAI基盤築く－Power On
THE SO WHAT
Apple’s pivot from a shelved autonomous car program into Mac-first custom silicon is a case study in repurposing long-horizon R&D into a defensible AI compute base. Operators with stranded or ‘failed’ bets in adjacent domains should be asking which capabilities — tooling, teams, IP — can be re-centered as AI infrastructure rather than written off.
READ THE SOURCE
MORE FROM THE WIRE
Tech & InnovationThe people set to influence tech policy in Burnham’s Britain
UK tech operators are about to get a new set of power centers — who sits around Burnham’s table will determine how AI regulation, data access, and digital infrastructure budgets actually move. Map your regulatory exposure now and build relationships early; once agendas harden, you’re just a lobbying line item.
Tech & InnovationForget the Kindle Scribe: I've tested the ViWoods AiPaper and it's the better epaper writing tablet at 33% off for Prime Day
A sub-AU$600, AI-leaning e-paper tablet undercutting Kindle Scribe shows how quickly ‘smart’ hardware categories commoditize once components and software are off-the-shelf. If your hardware roadmap leans on AI as the differentiator, assume price compression and focus your moat on ecosystem, workflows, and recurring services, not the device itself.
Tech & InnovationA profile of Asha Sharma, an outsider elevated by Satya Nadella to lead Xbox, where she made sweeping job cuts and shed game studios within her first few months
Putting an outsider in charge of Xbox and backing sweeping cuts this fast is a signal that even flagship consumer franchises are not exempt from hard P&L discipline. If you run an adjacent content or platform business, assume your counterpart may now be optimizing for margin and focus over empire size.
Elon Musk and Sam Altman are fighting in public again
Public conflict between major lab leaders keeps AI in the political and regulatory spotlight — that raises the odds of more personality-driven scrutiny rather than purely technical or economic debate. For operators, assume AI policy risk will stay noisy and path-dependent on a small set of visible figures, not just on formal processes.