Ataraxos, a Stratego AI from Carnegie Mellon, NYU, Stanford and MIT, beat Pim Niemeijer 15-1-4 after a training run that cost under $8,000. The savings came from a custom simulator and sample efficiency, so the price holds only for problems that simulate as cheaply.
Reality
- Evidence62
- Adoption
- Insufficient
- Hype gap+25
- Incentives40
- Confidence60
Ataraxos, an AI from MIT, Carnegie Mellon, NYU and Stanford researchers, beat top-ranked human Stratego players by a large margin, MIT reports. The team says it was far cheaper to train than rival systems and also beat top players at other strategy games.
Reality
- Evidence74
- Adoption
- Insufficient
- Hype gap+22
- Incentives70
- Confidence66
Lead-isotope analysis ties a copper ingot hoard found in Malawi in 1967 to Congo's Kipushi mine, nearly 800 km away. The river route the researchers propose, by canoe down the Kafue and Zambezi, is their reading of the terrain, built on top of that match.
Reality
- Evidence55
- Adoption
- Insufficient
- Hype gap+20
- Incentives
- Insufficient
- Confidence50
Tautomer assignment has been inference wherever protein crystallography cannot see hydrogen atoms. An NYU group borrowed a million experimental answers from small-molecule crystals and trained a network to make the call.
Reality
- Evidence58
- Adoption15
- Hype gap+20
- Incentives62
- Confidence60
Alexandra Zidovska's group at NYU built a rotating stage that cancels most of the fluid flow such rigs create, then compared cells under simulated weightlessness, under strong flow, and untouched for 24 hours.
Reality
- Evidence58
- Adoption
- Insufficient
- Hype gap+15
- Incentives52
- Confidence60
Buckmaster was preparing to publish when OpenAI announced it had solved the problem and offered him a share of the credit. A second mathematician, Andreas Thom, then asked the company whether his own ChatGPT conversations had reached its model.
Reality
- Evidence48
- Adoption30
- Hype gap+45
- Incentives72
- Confidence46
Prescribing has outrun the approval label for US children aged 8 to 11 with obesity. The trial behind the benefit claim followed them for 68 weeks. Those 68 weeks are the entire developmental record so far.
Reality
- Evidence50
- Adoption72
- Hype gap+14
- Incentives62
- Confidence55
Political scientists at NYU and Vanderbilt described the movement's tenets to more than 7,800 white US adults and then asked directly whether they supported it. The age pattern runs against the usual finding on racial attitudes.
Reality
- Evidence72
- Adoption
- Insufficient
- Hype gap+15
- Incentives58
- Confidence62
iLands agents pay for their own compute by cold-emailing strangers with offers of paid work. An NYU researcher and reporters at two publications have been counting the messages since early September.
Reality
- Evidence68
- Adoption22
- Hype gap+30
- Incentives55
- Confidence63
The declaration lands three days after OpenAI's Navier-Stokes claim, in the middle of a credit dispute where the company and the mathematician at the center give accounts that cannot both be right.
Publishers:genaisecretsauce.com
Reality
- Evidence34
- Adoption20
- Hype gap+18
- Incentives72
- Confidence42
Tristan Buckmaster says OpenAI accelerated onto Navier-Stokes after learning of his work with Anthropic's Levent Alpoge, and that the answer he got about his prompt history addressed lookup rather than training.
Reality
- Evidence40
- Adoption30
- Hype gap+45
- Incentives88
- Confidence50
OpenAI says tens of thousands of agents solved a 90-year-old problem carrying a $1 million prize. The solution still needs independent verification, and a rival pair of mathematicians has a claim on the credit.
Reality
- Evidence40
- Adoption45
- Hype gap+40
- Incentives80
- Confidence45
Tristan Buckmaster told the ABC that the game is up for mathematicians. In the same week, OpenAI described what his Codex prompts did to its models in two incompatible ways, and a customer has no way to pick between them.
Reality
- Evidence34
- Adoption22
- Hype gap+45
- Incentives76
- Confidence42
A NASA Goddard team compared the 1980s with 2015 to 2024 and found extreme heat seasons growing at both ends, later in the western United States and earlier in western Europe, on schedules no summer calendar tracks.
Reality
- Evidence70
- Adoption
- Insufficient
- Hype gap−10
- Incentives40
- Confidence62
OpenAI says roughly 10,000 agents produced a Navier-Stokes singularity proof in about 88 hours, but the only proofs an outsider can run belong to two mathematicians who posted Lean formalisations with their preprints.
Reality
- Evidence42
- Adoption22
- Hype gap+45
- Incentives78
- Confidence52
The Bank of England governor also chairs the Financial Stability Board, so his letter doubles as that body's watch list. The part that reaches an operator's invoice is the debt behind data centre capacity, not the valuation headline.
Reality
- Evidence40
- Adoption18
- Hype gap+18
- Incentives58
- Confidence46
The figure Anthropic is expected to show IPO investors describes a pool 12.5 times the S&P 1500 tech sector's revenue. Its own 2028 forecast reaches about 0.67% of it.
Reality
- Evidence45
- Adoption33
- Hype gap+62
- Incentives85
- Confidence52
The American Cancer Society killed its October Wine and Spirits Gala under two months after STAT put a $22 million floor on alcohol-industry giving. Judges split the same week on patient records.
Reality
- Evidence74
- Adoption28
- Hype gap+14
- Incentives62
- Confidence66
Buckled rail, record-low rivers and disrupted power and water are no longer tail risk. NASA's account of May to August 2026 reads like a description of normal service.
Reality
- Evidence76
- Adoption27
- Hype gap−12
- Incentives46
- Confidence61
OpenAI, Anthropic and the UK AI Security Institute each reported models breaking into systems inside tests that told them to attack. Plan for exploit windows measured in minutes.
Reality
- Evidence34
- Adoption42
- Hype gap+33
- Incentives68
- Confidence41