beat · 154 stories
A replayable market simulator lets LLM agents test interventions in a "lab" before a crisis decision. On one deployment (Luna) — the only one of three with a complete record — lab access cut regret 57%, but the caveats are large.
Patches shipped on 22 Oct 2026; the fix is a version upgrade, not a configuration change, and it lives with site operators and hosting providers, not browser users.
In occupations where AI can do more of the work, men's employment grew 6.3% for each step up Pew's AI-exposure scale (how much AI can do a job's core tasks) while women's grew 0.
Meta's European ad revenue grew 18%, against 33% in the US and Canada. The new variable is the EU's Less-Personalized Ads option, which uses 90% less data.
The Hangzhou-based lab behind DeepSeek's R1 — the open-weight (model parameters publicly released) reasoning model that went viral in early 2025 — named battery maker Contemporary Amperex (CATL) and Tencent as potential investors, with proceeds
A joint notice from China's commerce and science ministries (MOFCOM-MOST) freezes compute and parameter thresholds but tracks "globally leading capability" on the leaderboard, a moving target the 30-day comment window can still shape.
OpenAI dropped 722 math papers on GitHub overnight, including claimed progress on three famous open problems. The mathematicians asked to vet them say they didn't.
DataBrain's Text2Dashboard prototype passed 4 of 4 single-panel dashboard tests and 10 of 10 fault scenarios, but its authors call the result a small prototype, not a production claim.
A new systems paper, FluidPD (Prefill-Decode), splits each request into two phases and lets a server borrow GPU time between them, or swap worker roles in place, when one side goes idle.
TEMPEST, a 2.8MB AI model, hits 91.71% accuracy on a 45-driver dataset and only loses 4.3 points as the pool grows. The privacy question is what the paper does not ask.
GA-GRPO (Guidance-Augmented GRPO, a training-recipe variant for reasoning AIs) turns expert hints, self-explanations, and retrieved reasoning patterns into a signal-to-noise dial, with a closed-form optimum and a matching lower bound on the bias
The Chicago class action borrows the algorithmic-coordination antitrust theory the federal courts have been applying to real-estate pricing vendor RealPage, but the franchisor's role as both brand owner and pricing-engine deployer puts the case on