Hallucinations are deeply embedded in large language models because they are designed to produce the most plausible, most likely word given training data, not to distinguish truth from falsity, so the model produces high-probability words regardless of accuracy.

causalpending

Speaker

Michael Wooldridge

Evidence Quote

“Hallucinations are deeply embedded...because the AI doesn't know truth from falsity”

Source

This is not the AI we were promised | The Royal Society— The Royal Society
Created: 8/12/2026, 6:23:19 PM

Your notes

Loading notes...