The World Times
Sanderson's million-view video proves the discipline is thriving. Stockton's layoffs prove the department isn't — because the thing that pays mathematicians isn't enthusiasm for math.
OpenAI pulled three of 722 math manuscripts a day after posting them. The retraction is the honest part. The 419 papers with no Lean certificate are the part nobody wants to discuss.
The fight over whether Claude 'discovered' ART misses what Anthropic actually built this month: a wet lab, a gated verification program, and the only instrument that can check its own claims.
A charming Raspberry Pi gadget turns backyard birds into 1800s illustrations. It's also a beautifully engineered substitute for the thing conservation actually needs.
Twenty-five Fields Medalists say AI labs have misaligned mathematics. They're right — but the misalignment is a labor bill, not a philosophy problem.
The September declaration on AI and mathematics isn't a capability panic. It's a bill for verification labor that neither academia nor industry has ever agreed to pay.
The Navier-Stokes credit fight isn't a scandal — it's the first crack in a mathematical credit system built for a world that no longer exists.
A post hoc analysis of the SELECT trial shows semaglutide shifts a dementia-risk biomarker. The gap between 'predicted' and 'actual' is the whole story.
OpenAI dropped ten math breakthroughs in a single paper with no named human authors. The real story isn't the AI — it's what happens to scientific credit when discovery becomes a product launch.
A preclinical HIV vaccine study hit the top of Hacker News this month. That says more about the collapse of trust in science communication than it does about the study itself.
Terence Tao's public ChatGPT session wasn't about verifying a counterexample; it was about establishing a new ritual of trust in an age of machine-generated knowledge.
Thursday's detection of an atmosphere on a distant rocky world is a triumph, but not for the reasons the headlines suggest. The instrument that found it is a rebuke to the privatize-everything gospel.
A new study on how modern design strains the brain is really a warning about the digital environments we inhabit for hours each day—and the business models that keep us there.
The ESO's call for a hard cap of 100,000 low-orbit satellites is scientifically rigorous and politically naive — and it dodges the question of who funds the alternative.
Scientists built a cell that grows and divides from scratch—an astonishing, historic, genuine achievement. But the way we talk about it says more about our hunger for revelation than about the science itself.
The outcry over political appointees reviewing grants misses a deeper rot — decades of academic science producing findings nobody asked for, on questions no voter would fund with their own money.
A University of Kansas pilot study found creatine slowed early Alzheimer's cognitive decline by roughly 30%. The real story isn't the science — it's what the silence from pharma tells us about which breakthroughs get funded.
