Topics

Governments, corporations, and publics are all vying for authority over increasingly capable AI. The harder question is who answers when it goes wrong.

What a volcano sceptic can teach us about AI scepticism

MIT CSAIL researchers show that a frontier language model can understand, repair, and even create neural network weights directly, hinting at a new way to optimize and interpret AI systems.

NASA's Mars rover has driven farther than any vehicle on another world, using autonomous navigation and self-localization that point toward more independent space exploration.

Google and SpaceX recently launched the first experimental satellite equipped with AI chips, testing whether orbital data centers could solve Earth's energy constraints. The project promises abundant solar power and reduced environmental impact, but faces significant technical hurdles including cooling, radiation, and data transmission.

Some evocative if very simplistic toy-simulation results exploring dynamics of goal-sets under self-modification and intelligence increase.

A biotech startup connected an AI agent to lab equipment and let it carry out a brain injection on a living mouse, with human safeguards in place.

Ohio State researchers built an AI assistant that studies lunar data, judges possible landing sites, and explains its reasoning to the people planning future missions.

Using artificial intelligence and supercomputers, researchers have found overlooked MXene materials in old databases and created new ones with a faster, cleaner method.

Hours after OpenAI announced it had overturned the exponential time hypothesis, rival Anthropic revealed its own model had reached the same result on the same day, by a different route and with a very different rollout.