To the editor: Visitor contributor Iddo Gefen not solely laments the analogy between the human mind and synthetic intelligence, however he additionally means that human minds don’t be taught or recall like an AI (“The human brain doesn’t learn, think or recall like an AI. Embrace the difference,” July 9). Actually, Gefen will get it backwards: Latest massive language reasoning fashions of AI be taught, recall and remedy issues a lot in the way in which people do.
People (and different animals) be taught most of our vital behaviors via reinforcement, the foundational regulation of studying first instructed by psychologist Edward Thorndike’s Legislation of Impact and later experimentally investigated and confirmed by the behavioral psychologist B. F. Skinner and his colleagues.
As well as, verbal people ceaselessly remedy issues by speaking (and imagining) to ourselves till we discover a resolution. Reinforcement performs a crucial position in studying and fascinating in problem-solving conduct. Reasoning LL fashions of AI which can be additionally programmed to be taught via reinforcement mimic such problem-solving behaviors in people and at the moment are among the many strongest AI machines.
Sadly, alongside the way in which, Gefen ventures past his experience as a neuroscientist and mischaracterizes “behaviorist psychology.” If he appeared deeper, nonetheless, he would discover that behavioral psychologists have found legal guidelines of conduct which have revolutionized the therapy of a variety of conduct issues, in addition to revolutionizing schooling. They now present the muse for an entire new era of AI machines that assume and be taught like people.
Henry D. Schlinger Jr., Glendale
This author is a professor of psychology at Cal State Los Angeles.
