News

Can AI chatbots ever become as smart as human intelligence? New study reveals a fatal flaw in the plan

A new commentary in Nature Human Behaviour argues that the race toward human level AI is built on a flawed assumption. Neuroscientists Evelina Fedorenko, Steven Piantadosi and Edward Gibson show that language is a communication tool, not the basis of thought, meaning better language models will not produce human style intelligence. Evidence from brain imaging and clinical cases reveals that humans think independently of language, suggesting AI fluency cannot translate into true reasoning or cognition.