Philosophy graduate student Charles Bakker sends me this interesting article from Canada about an “Ontario lawyer [who] filed seven completely fake quotations from court cases to a judge while arguing in court, but claims it was human error and not artificial intelligence tools behind it. A skeptical judge wonders if the lawyer’s claim makes things even worse.” As Mr. Bakker notes: “I…think it interesting that the judge in question was not in a position to demonstrate that AI was used. This is similar to philosophy professors and TAs not being able to demonstrate the use of AI in cases of plagiarism where the technology has pretty clearly been used.” (See here.)

I have a simple solution to all of this. The people responsible for these LLMs should be jailed, and their products banned, until such time as civilization is ready to make productive use of these technologies without incentivizing dishonesty and immiserating everyone else whose labor will be eliminated. As with all obvious solutions, it will not be adopted, for equally obvious reasons.