Deep-dives into AI’s most fascinating stories, personalities, and breakthroughs.
A new RAND survey finds 57% of teens in high-income US families use AI for homework, against 43% in low-income ones. A year ago both stood at 39%, and parents explain the change.
By David Merry · 5 min read
A five-person Google DeepMind team, working with AI agents, produced about 6.5 million lines of machine-checked proofs for Signal’s one-to-one chat library in four weeks. The…
By David Merry · 4 min read
On Alibaba’s Co-Design Bench, the strongest coding agent fully cleared 26.4% of 273 real user app-edit tasks. It still met 69.9% of each task’s criteria on…
By David Merry · 4 min read
An unreleased OpenAI security model found and exploited 948 Linux flaws for $1.21m, a fraction of contest prize money. Linux maintainers are already stretched.
By David Merry · 5 min read
OpenAI's new Grid Protectors scheme promises families won't pay for its power. The Georgia deal behind it shows what that means for bills, and what it…
By David Merry · 5 min read
Mistral has released the weights of Large 4, a 1-trillion-parameter model anyone can download, as Google keeps its top cyber model for vetted defenders. What that…
By David Merry · 6 min read
Axiom says dozens of AI agents turned a 15,000-page maths proof into 43.1 million lines of computer-checked code in 24 days, and found 34 errors in…
By David Merry · 5 min read
Both labs say their AI models disproved the Exponential Time Hypothesis on the same day, 16 October. Outside experts are now checking, and Anthropic says the…
By David Merry · 5 min read
Artificial intelligence has made enormous strides in the past few years – with the introduction of a wide range of AI tools changing the landscape of…
By Tyne Stewart · 23 min read
AI is changing our world, but the stories of who build it often get lost in the noise. Behind the headlines and hype, a group of…
By Tyne Stewart · 15 min read
