« Not a good time for criminals | Main | Minnesota mass killer planned to avoid armed citizens »
John Lott vs. AI: guess who wins?
Dr. Lott posts on a debate between himself and Grok on gun issues. At the end he concludes that Grok is "biased and actually stupid" when it comes to empirical research.
To the legal profession, that's not news. The risks of using AI are a running joke (except to the attorneys involved, who often get hit with sanctions). Judges don't have to worry about sanctions, just the embarrassment. As in this case, or this one. The most likely explanation is that the judge signed whatever his research clerk put in front of him without so much as reading it, and the clerk used AI. I mean, when a ruling names the parties to the case, and most of them are invented, invents events that were never even plead, and relies upon declarations (affidavits) from invented people... Which means the judge (at most) told the clerk he wanted to dismiss the case, research it and write me an opinion, and then signed it without reading.
update: AI doesn't do so good as a mental health therapist, either. Encouraging "patients" to commit suicide, or to murder the regulators of the profession, while framing someone else for the offense, probably doesn't meet the standard of care...
4 Comments | Leave a comment
Judges do get hit with sanctions - another of the myriad of problems that prove our legal system has always been broken. The legal profession should attack every case like a real scientist, question all previous "statements" and prove them true of false. Judges have become our imperial rulers because We the People have failed to kick them down and put their feet to the fire. Judges are not smarter than the rest of us. So judges should have citizen boards to sanction them every time they violate OUR LAWS as written. No interpretation. No guessing at intent. Simple pure language regardless of where it leads. If the language is bad, send the law back to the legislatures to fix. Again and again until they get it right. And get practicing attorneys out of the legislative branch. Any attorney who is working is a member of the bar and an officer of the court thus is a member of the judicial branch. If they want to be a legislator, they should have to quit the bar, close their practice so that they don't violate separation of powers. They can always sit for the bar after serving in a legislative of executive position.
Signed, No use for judges because most have their heads up deep.
Judges do
Should have been don't
Lets face it: Professor John Lott vs....well, anyone? Bet on Lott.
"Grok is 'biased and actually stupid' when it comes to empirical research"
AI (to me, an uncredentialed observer) is just a vastly amplified version of predictive text message completion. If I type "m a s s a . . ." into a phone with this feature active, it will type the rest of the word that is most often typed in the context---preceding words in the message I'm typing, preceding messages in the thread that I'm messaging, and prior messages from me and to me, and so on.
It could be "Massachusetts" or "massage" or a family name; it could even assume I was misspelling "massive" or the Spanish name for corn flour.
And its predictions will arise from "internal knowledge" that was trained into it by humans with varying credentials, and varying biases.
Of course AI will dredge up only what it has seen before, and of course AI will lend more credibility to sources, and meta-sources, that get appear more often in training data or on the open internet.
That seems to me to be quite the opposite of empirical research: "what's everybody else saying?" Putting aside your own biases is hard enough, spotting biases of others not much easier.