Google Cloud AI Research has released RRSI, an Apache 2.0 tool whose self-rewriting agent harness lifted Terminal-Bench 2.1 scores from 74.2% to 80.2%. Any team can use it commercially, though on tasks the agent never trained against the reported gain falls to 4.7 points.
Reality
- Evidence45
- Adoption
- Insufficient
- Hype gap+15
- Incentives50
- Confidence40
A peer-reviewed audit scored the free ChatGPT model's election answers on one day in late October 2024, and it counted hedging and omission as failure categories alongside outright error.
Reality
- Evidence62
- Adoption
- Insufficient
- Hype gap+25
- Incentives55
- Confidence58
HUMID-Atlanta covers 2010 through 2023 and is publicly available. Whether its building-scale wind and humidity explain heat harm better than thermometer readings is being tested in an NCAR and Emory study of kidney injury.
Reality
- Evidence52
- Adoption28
- Hype gap+18
- Incentives66
- Confidence56
The grant to 1Day Sooner is the smallest disclosed line in the Foundation's first health-data round. It pays for a feasibility test, because the group's two attempts to get drug company files this year both failed.
Reality
- Evidence62
- Adoption20
- Hype gap+25
- Incentives68
- Confidence58
Fortune's split screen at the G20 in Chapel Hill has Sam Altman calling more infrastructure the only way to keep AI cheap while hundreds outside object; the number that decides it is a 3,000-site pipeline needing county consent.
Reality
- Evidence58
- Adoption72
- Hype gap+14
- Incentives74
- Confidence54
A Penn State and UNC team matched sea surface temperatures to child health records from 36 low- and middle-income countries. The association came out strongest for stunting, the outcome that takes years to build.
Reality
- Evidence55
- Adoption
- Insufficient
- Hype gap+12
- Incentives52
- Confidence50
In an aligned fluid, a Caltech group reports in PNAS, colonies grow as chains one cell wide that run straight and then kink in a single tight spot, which hands part of colony architecture to the medium.
Reality
- Evidence56
- Adoption
- Insufficient
- Hype gap−8
- Incentives55
- Confidence46
North Carolina, Wisconsin, Louisiana and Connecticut are routing tax dollars and betting revenue into athletic budgets. The federal bill meant to impose guardrails does not restrain them.
Reality
- Evidence62
- Adoption66
- Hype gap+14
- Incentives55
- Confidence61
Hillary Salo takes the FASB chair on July 1, 2027 through June 2034. She is the insider who helped build the current agenda, which tells reporting teams what to budget for.
Reality
- Evidence72
- Adoption
- Insufficient
- Hype gap+6
- Incentives55
- Confidence66