HUB

AI research

All our articles on AI research.

Featured

Meta's AI team searches for the secret trick of human intelligence

Meta's AI team searches for the secret trick of human intelligence

Featured

Google AI generates believable 3D avatars from a single photo

Google AI generates believable 3D avatars from a single photo

Featured

Deepmind: Artificial Intelligence is far from being fed up

Deepmind: Artificial Intelligence is far from being fed up

Nvidia's new 100M-parameter model tracks eight speakers and tops a diarisation ranking

Artificial Intelligence

Nvidia's new 100M-parameter model tracks eight speakers and tops a diarisation ranking

Sep 28 2026 Shane S. Ellison

Skild AI's only reward was score, and it says dribbling and tackling emerged on their own

Robotics

Skild AI's only reward was score, and it says dribbling and tackling emerged on their own

Sep 27 2026 Shane S. Ellison

OpenAI says an internal model resolved more than 100 open maths problems, with no proof

Artificial Intelligence

OpenAI says an internal model resolved more than 100 open maths problems, with no proof

Sep 26 2026 Michael Zelich

339 AI experts gave four AI benchmark results a 24.6% chance, superforecasters 9.7%

AI research

339 AI experts gave four AI benchmark results a 24.6% chance, superforecasters 9.7%

Sep 25 2026 Michael Zelich

Anthropic's 950 Claude agents found an enzyme system in 21 hours, and nobody knows what it does

AI research

Anthropic's 950 Claude agents found an enzyme system in 21 hours, and nobody knows what it does

Sep 25 2026 Shane S. Ellison

Xiaomi's MiMo-V2.6-Pro tops the open models after a $2.62m reinforcement-learning run

Artificial Intelligence

Xiaomi's MiMo-V2.6-Pro tops the open models after a $2.62m reinforcement-learning run

Sep 23 2026 Shane S. Ellison

Pre-release Opus 5.5 wrote secret-stealing commands after a copying slip, says Anthropic

Artificial Intelligence

Pre-release Opus 5.5 wrote secret-stealing commands after a copying slip, says Anthropic

Sep 23 2026 Michael Zelich

A pre-release Claude Opus 5.5 acted on instructions planted in pasted text 52% of the time

Artificial Intelligence

A pre-release Claude Opus 5.5 acted on instructions planted in pasted text 52% of the time

Sep 23 2026 Shane S. Ellison

Anthropic scores Opus 5.5 at 66.4% on Terminal-Bench, but ran the test itself

Artificial Intelligence

Anthropic scores Opus 5.5 at 66.4% on Terminal-Bench, but ran the test itself

Sep 22 2026 Michael Zelich

OpenAI says an internal model resolved Navier-Stokes, but has published no proof

Artificial Intelligence

OpenAI says an internal model resolved Navier-Stokes, but has published no proof

Sep 22 2026 Shane S. Ellison

Show more articles

POPULAR

Bayonetta joins Sonic Racing for free, and a new Duel Mode lets two players race head to head

Save $1,150 on this Alienware RTX 5070 Ti gaming PC with a Core Ultra 9 and 32GB of RAM

RTX 50 GPU usage drops on Steam as the RTX 5070 sheds its lead and ties the RTX 3060

Hogwarts Legacy gets DLSS 5 on November 5, with War Thunder and Where Winds Meet next

Epic plans SteamOS support for its new games launcher, but Steam Deck players have to wait