As AI language skills grow, so do scientists’ concerns

The tech industry’s latest artificial intelligence constructs can be pretty convincing if you ask them what it feels like to be a sentient computer, or maybe just a dinosaur or squirrel. But they’re not so good – and sometimes dangerously bad – at handling other seemingly straightforward tasks.

Researchers say they’d be less concerned if the companies developing these systems, known technically as “large language models," were less secretive about the data used to “train” these AI systems.

Leaders such as Google and OpenAI haven’t budged, so other scientists are now building their own systems to crack open the black box of AI behavior.