Gowers tests ChatGPT 5.5 Pro on open math + Anthropic gives Petri to Meridian

Mathematician Tim Gowers, a Fields medalist, asked ChatGPT 5.5 Pro to attack open problems on sumset diameter from a Mel Nathanson paper in additive number theory. In under two hours the model improved a known exponential bound to a polynomial one — work the original researcher Isaac Rajagopal called "original and clever" and Gowers judged at "the level of a perfectly reasonable chapter in a combinatorics PhD." Gowers concludes that PhD-style "gentle problems" have been crossed off the LLM frontier and researchers must now aim above what these models can prove.

Anthropic released v3.0 of Petri, its open-source toolbox for evaluating large language models on deception, sycophancy, and cooperation with harmful requests, and donated stewardship of the project to Meridian Labs, an independent AI-evaluation nonprofit. The move is framed as keeping alignment tooling lab-neutral as Petri scales — sister labs can adopt and contribute without depending on Anthropic's release cadence. v3.0 also ships an add-on called Dish that increases scenario realism so models can't easily detect they are being evaluated.

The Walton Family Foundation's sixth Gen Z AI poll, fielded with Gallup across 1,572 respondents aged 14 to 29, finds weekly AI use stuck at 51% — growing only four points year-over-year — while anger toward AI jumped to 31% from 22%. Eighty percent now say relying on AI to finish tasks faster will make learning harder, and fewer than 20% would choose AI for tutoring, financial advice, or customer service. Schools tightened in parallel: 74% of Gen Z students report a school AI policy, up 23 points from the prior year.

In a new paper on arXiv, researcher Valerio Capraro coins LLMorphism — the biased belief that human cognition operates the way a large language model does. He argues the framing flows from two channels: analogical transfer (projecting LLM features onto humans) and metaphorical availability (LLM vocabulary becoming the dominant way to talk about thought). Capraro: "the issue is not only whether we are attributing too much mind to machines, but also whether we are beginning to attribute too little mind to humans." The paper traces implications across work, education, healthcare, and human dignity.

Wispr Flow, the AI voice-input app, told TechCrunch that India is now its second-largest market by users — with growth that doubled month-over-month following its recent Hinglish voice-support launch — but India still produces only 2% of revenue against 14% of global downloads. The company introduced localized pricing at ₹320 (about $3.40 on annual plans) and plans to scale to 30 India-based employees over the next year, eventually pushing the entry tier toward 10 to 20 cents per month to reach household users beyond knowledge workers.