A three-month field experiment with 133 patent attorneys found that AI boosts immediate performance, especially for junior staff. Once the AI was removed, only senior attorneys kept the gains, suggesting existing expertise is needed to turn AI-assisted work into lasting skill.
A new study found that Grok overwhelmingly recommends Republican candidates while OpenAI's GPT leans toward Democrats. The two models disagreed in more than 90% of nearly 500 election races.
In an image of a mountain lake drawn by Claude Fable 5.1, each star encodes a letter according to its height in the sky, a method the model came up with on its own. OpenAI's GPT-6 Astra decoded 44 of the 50 hidden-message images Fable produced.
Google says an unattended 11-hour run of Gemini 4 Argon produced two accepted optimisations to its Agent Development Kit codebase. They cut in-memory session preparation overhead by about 40% in long sessions and cold package import time by 21%.
Microsoft Research has released Agensh, a multi-agent harness with no central orchestrator in which self-organising workers share a Git repository, a chat and a findings board. On ProgramBench, with a six-hour budget, it scaled from one agent to 1,024.
The researchers behind OpenMobile-2 argue smartphone agents improve by training on many different apps rather than on more data. They built an open environment with over 110 apps, including 35 replicas of popular ones, and say their model scores highest among open agents on real Android devices.