Google's new AI mode is good, actually

The example itself that Simon puts up is questionable. I might be wrong about this, but I thought I read elsewhere that the “buy, scan, destroy” method was explicitly not the problem, and instead the issue was that anthropic downloaded libgen, and the settlement was for libgen.

Edit: https://news.ycombinator.com/item?id=45143392

Yeah, you're right! The answer is definitely misleading at best. It would be better if the sentence "This method was a major component of a copyright lawsuit settlement that Anthropic paid in September 2025." was removed.

I'm sure this method _did_ come under discussion in the lawsuit & settlement, but as you pointed out the settlement itself was only about pirated works.

Yeah, reading the response raised a lot of questions for me too. I find it striking that he doesn't comment on whether the presented facts were true or relevant. What does it mean that "In 2025, a search tool revealed that Meta [...]". Is it the search tool that the "AI mode" is using under the hood?

Welcome to subtle misinformation as truth everywhere. "Ministry of Truth" might be a good name for this phenomenon.

i don’t read the AI answer as saying otherwise, but one might read that the physical purchasing was used as part of the settlement. it was used as part of the defense to mitigate though

Yeah, I agree: "This method was a major component of a copyright lawsuit settlement that Anthropic paid in September 2025" is a bit misleading, I actually wrote a whole separate thing about that here: https://simonwillison.net/2025/Sep/6/anthropic-settlement/

Yeah, it's amazing how fast people can overlook incorrect results when the form is in the spotlight. Or the AI rhetoric is convincing or plausible.

I recall a tweet by Altman, leaking the launch of GPT-5, praising their new model's answer to a prompt about thought provoking TV shows about AI. The X thread that followed was about the form ("em-dashes are still there!") and nearly nobody cared to evaluate that neither shows recommended were about AI. They weren't, or at least, were very debatable as belonging to the genre.

I especially liked how the image inside of a link I shared of an example of GPT-5's (in)ability to rank states was decented sans warning...