Two roundups today. Stratechery's weekly email covers Apple's foldable iPhone Duo, OpenAI's Navier-Stokes proof versus Meta's Muse agent, and the NBA's harshest-ever penalty for the Clippers. Not Boring's Weekly Dose of Optimism has the details on that math proof, DeepMind's AlphaGenome Atlas, Hermeus' ramjet test plane, Michael Levin's 'Ingressing Minds', and Neko Health's NYC launch. Faye D'Souza (one new members-only daily), Vested Shorts, FoundMyFitness and Every were quiet.
Stratechery
The Friday roundup of the Stratechery bundle. Top picks: Apple's iPhone event introduced the iPhone Duo, a foldable priced between $2,000 and $3,200 - Ben Thompson's piece argues Apple again showed the power of hardware-software integration but that its biggest AI blindspot is believing in the primacy of apps; Gruber nailed the 'Duo' name back in April. Wednesday's Update contrasted OpenAI's technically thrilling Navier-Stokes breakthrough (which may chill advanced math research) with Meta's far less sexy Muse agent launch - free consumer hardware and software that drops the barrier to ordinary people using agents, 'making the AI upside a lot more accessible to the masses who don't want to buy a Mac Mini'; Sharp Tech covered Muse plus Anthropic and the revival of p-doom angst. And the Clippers received the harshest penalty in NBA history for circumventing the salary cap to pay Kawhi Leonard - GOAT recapped the evidence that sealed Ballmer's fate and why Kawhi got off easy. Also out: Dithering episodes on 'Agents and Forklifts' and the iPhone Duo, and Asianometry on the flood that wrecked the hard disk drive industry.
Why it mattersOne email catching up the whole Stratechery week: the foldable iPhone is the first genuinely new Apple hardware bet in years, and the OpenAI-math vs Meta-agents contrast is a clean frame for where AI value actually lands for normal people.
Read at source
Not Boring
Sent at 8:46am on the 25th anniversary of 9/11 with a nod to Colson Whitehead's 'Lost and Found'. Five items: (1) OpenAI published a proposed solution to the Navier-Stokes existence and smoothness problem (a Millennium Prize Problem) - roughly 10,000 agents worked about 88 hours using an unreleased model more capable than GPT-6 Astra, with 17 more hours to formalize and verify in Lean; the 166-page proof addresses the forced version, producing unbounded velocity with finite total energy. Anthropic, working with a mathematician on its own solution, accused OpenAI of using logs from their Codex conversations; Altman denied it but said rumors Anthropic was close spurred the attempt. (2) Google DeepMind released AlphaGenome Atlas, a one-petabyte database predicting molecular effects for all ~9 billion possible single-letter DNA changes including the non-coding 98%; it already gave supporting evidence in an unresolved rare-disease case at the Broad Institute. (3) Hermeus unveiled Ramjet-X, an expendable test vehicle launched at speed from its reusable Quarterhorse plane, to be sold as Flight Test as a Service; integrated testing planned for 2027. (4) Michael Levin published the peer-reviewed 'Ingressing Minds' - minds as patterns in a structured nonphysical space, with bodies and possibly computers as interfaces; Packy calls Levin possibly the most important scientist of this era. (5) Neko Health, Daniel Ek's preventative-health company, launched in NYC's Soho: full-body scan, bloodwork and more for $499, with results explained immediately after.
Why it mattersThe Navier-Stokes proof is the headline - if it holds, agent swarms just took a Millennium problem - but the DeepMind atlas and the Levin paper are the items likely to matter in five years. Good weekend reading either way.
Read at source