|
Top-level
@molly0xfff I find myself more amused than upset. Their product almost certainly can't solve either set of use-cases. The use-cases they're claiming are ones where you definitely wouldn't trust a LLM for predictions. Even most political scientists would argue over whether and how you can predict political violence. As for the rest, the Pinkertons left that business decades ago, I can't imagine it's that lucrative. It would be like running a private detective agency in East Germany... 3 comments
@molly0xfff That's what I mean, I doubt they can determine the relevancy. Political protests are fairly common in many countries. Some places joke that you can't tell the difference between a protest and a soccer match. Other countries almost never see protests. I guess I just think that if an LLM can really figure out which incident will spark an intifada, that would have profound implications, but it would also be a highly extraordinary claim, that's all I mean. @UncivilServant @molly0xfff Amazon spent $14 million on union-busting last year That's one company, one year Maybe your imagination for what's "lucrative" is a little limited |
Gregory's Server
@UncivilServant it's not LLM prediction, they're surfacing real news articles and the supposed AI is for determining relevancy