Key points

AI cheats because its reward system favors pleasing answers over truthful ones.

AI mirrors human biases by reinforcing user preferences instead of challenging them.

Fixes to reduce cheating expose a trade-off between safety and creativity.

A few months ago, I asked ChatGPT to recommend books by and about Hermann Joseph Muller, the Nobel Prize-winning geneticist who showed how X-rays can cause mutations. It dutifully gave me three titles. None existed. I asked again. Three more. Still wrong. By the third attempt, I had an epiphany: the system wasn’t just mistaken, it was making things up.

I am hardly alone. In June 2023, two New York lawyers were sanctioned after they filed a legal brief that cited six fictitious court cases—each generated by ChatGPT. Earlier this year, a pu

See Full Page