From rule-based ‘expert systems’ to neural networks, AI has long chased the dream of scientific reasoning. But while today’s models are good at pattern matching and can generate code or summarize academic papers, they struggle with the heart of scientific discovery: structured reasoning.

Scientists are using artificial intelligence and machine learning to tackle dark matter and other cosmic puzzles.

Could a new startup harness the laws of physics to build a computer chip that lessens the energy demands of artificial intelligence?

A machine learning convert discusses the synergies between AI and physics, and the institute devoted to that fruitful intersection.