News
The AGI Race Starts After the Breakthrough
A CNAS report says technical possession is not strategic power. METR's reliability gap shows why the measurement problem is bigger than the label.
Read articleThe Eleventh Hour
The newsroom
Behind the claims. Inside the evidence.
News
DeepMind's Four Roads to ASI Lack a Common Measure
A careful new roadmap exposes the real problem: scale, recursion and agent crowds still lack a common test of intelligence.
Read articleNews
A One-Year AGI Forecast Needs an Agreed Finish Line
Selected AI benchmarks show rapid progress alongside deep weaknesses elsewhere. Calibrating an AGI date requires a fixed event, system and test.
Read articleNews
Small Brains Break the Intelligence Ladder. 'Cognitive Style' Still Needs a Test
Animal cognition defeats crude rankings by brain size. It does not make neural resources irrelevant or turn a memorable label into a mechanism.
Read article