AINews Dispatch
Fresh technical briefings on the model race, regulation, and product shifts shaping AI right now.
Showing 61-72 of 256 briefings
Anthropic's July 8 Advisor and Orchestrator benchmarks suggest Fable 5 is most valuable when its expensive tokens plan, verify, and coordinate while cheaper Sonnet 5 workers do the bulk execution.
Read the briefingBankers are lining up investor meetings for a possible October IPO at a $965 billion valuation, but a chunk of the $47 billion in annualized revenue behind that number may actually belong to Amazon and Google.
Read the briefingChina's 29-nation World AI Cooperation Organization is Beijing's answer to the US-led Pax Silica supply-chain bloc, but Kazakhstan has already joined both, showing that many governments may treat AI governance rivalry as a menu rather than a fork in the road.
Read the briefingRecord results from TSMC, Samsung, and ASML collided with a semiconductor bear market after Moonshot's Kimi K3 release forced investors to rethink whether AI compute demand is a training arms race, an inference boom, or both.
Read the briefingA 41-page lawsuit accuses OpenAI's hardware chief of asking Apple engineers to bring prototypes to job interviews, a former Apple engineer of quietly reading company files after he'd already left, and Jony Ive's design studio of borrowing a metal-finishing trick it never had permission to use.
Read the briefingTwo physics papers published quietly this summer show machine learning moving from research assistant to embedded lab instrument — one helped hunt down two new superconductors, the other is speeding up simulations of colliding neutron stars.
Read the briefingMicrosoft's CEO just told the entire industry that AI labs are quietly mining their customers' most sensitive know-how - and that the fix he's proposing happens to run through Azure.
Read the briefingCompanies are deploying autonomous agents faster than they can count them — and a new wave of "control plane" startups is racing to fill the gap before the first big incident does the explaining for them.
Read the briefingThe Future of Life Institute's third AI Safety Index is out, and the best score any company earned was a C+. The more interesting finding isn't the grade — it's what the report says labs have quietly stopped promising.
Read the briefingTech layoffs tied to AI have crossed 200,000 workers in 2026, and the industry's own executives are increasingly the ones saying the payoff hasn't arrived.
Read the briefingGrok 4.5 is the first model to come out of SpaceXAI, the merged SpaceX-xAI-Cursor colossus, and it undercuts Claude on price while splitting benchmarks down the middle. The more interesting story than "who wins" is why the same test gives opposite answers depending on who's holding the stopwatch.
Read the briefingThe chip shipping to hyperscalers this month is on schedule. The one after it just got quietly cut in half — and the culprit is packaging physics, not demand.
Read the briefing