Large language models aren’t people. Let’s stop testing them as if they were.
Illustration by Muhabit ul haq via AISynopsisMultiple researchers claim large language models can pass tests designed to identify certain cognitive abilities in humans. Such results are…