Large language models aren't people. Let's stop testing them like they were.

When Taylor Webb played around with GPT-3 in early 2022, he was blown away by what OpenAI’s large language model appeared to be able to do. Here was a neural network trained only to predict the next word in a block of text—a jumped-up autocomplete. And yet it gave correct answers to many of the…

Related Keywords

Stanford University ,California ,United States ,United Kingdom ,Israel ,New Mexico ,Cambridge ,Cambridgeshire ,Taylor Webb ,Geoffrey Hinton ,Melanie Mitchell ,Lucy Cheke ,Michal Kosinski ,Google Deepmind ,Natalie Shapira ,Los Angeles ,Laura Weidinger ,Harvard University ,University Of Cambridge ,Ilan University In Ramat Gan ,Microsoft ,Santa Fe Institute ,University Of California ,Twitter ,Google ,United States Medical Licensing ,Bar Ilan University ,Ramat Gan ,Progressive Matrices ,Reasoning Challenge ,Tomer Ullman , ,