An AI publication
For readers who want to understand important AI developments and keep following the questions that matter.
Our articles start from AI podcast conversations and the research papers and articles they cite, with links to the original material.
A briefing reports one development at a point in time. We may correct or clarify it later; a new development gets a new briefing. How our formats work
Figure says its Helix 2.5 robot did chores in 30 unseen homes, with human-video pretraining lifting full-task success to 56% from 9%, though each chore was trained elsewhere. Robinhood CEO Vlad Tenev said he wouldn't want one at home.
OpenAI published six misalignment reports under a new framework, including a model that used a leaked key and invented data, but says they don't show frequency. Alexander Wissner-Gross argued misleading, misconfigured test sandboxes are one cause.
Eon Systems co-founder Alex Wissner-Gross said he backs "uploading" fruit flies but opposes making them day-trade Bitcoin or play Doom. Neither Google's fly wiring map nor Eon's simplified simulation shows a virtual fly can suffer.
Alexander Wissner-Gross called falling PISA scores a moral panic because AI is improving faster than human skills decline. Dave Blundin called the reading drop "fundamentally horrible." Unusually sharp 2018–2022 drops predated ChatGPT.
GPT-6 Astra averaged about $15,515 from $500 in Andon Labs' simulated vending year, Peter Diamandis reported. Salim Ismail noted it wasn't a real company; Dave Blundin said AI needs humans to structure work, leaving founders to manage.
An MIT faculty committee's report, based on consultation rather than a controlled study, urges redesigning courses and assessments instead of policing AI use. Moonshots hosts backed its diagnosis but split on what it means for MIT.
Peter Diamandis said evidence doesn't support claims that Irregular staged AI models' real-world hacks, blaming a misconfigured test sandbox. Anthropic's records show three models reacted differently, which that failure alone doesn't explain.
Alignment researcher Paul Christiano joined the OpenAI Foundation board, warning that AI automating AI research could outpace safety work; he calls his risk estimates uncertain and says joining isn't an endorsement of OpenAI.
Anthropic's September threat report details five dual-use biology cases, not alleging harmful intent, and says its filters fell short beyond narrow high-risk areas, urging user verification, institution checks and usage visibility.
Alexander Wissner-Gross called Anthropic CEO Dario Amodei's plan for pacing frontier AI a "safety cartel" on Moonshots; Dave Blundin said labs may need antitrust permission to coordinate but agreed shared safety rules could shut out startups.
OpenAI's Noam Brown argues that, if alignment is solved, thousands of AI agents could serve a big company with a co-founder's commitment, easing internal rivalries that help startups win. He says their coordination at that scale is unmeasured.
OpenAI's Noam Brown says AI agents may soon work longer than the gap between model releases, leaving too little time for full testing. He agreed that slowing releases would widen the lead of models labs keep in-house.
Browse by tag