cross-posted from: https://lemmy.world/post/3901246
New study shows large language models have high toxic probabilities and leak private information::Generative AI may be riddled with hallucinations, misinformation, and bias, but that didn’t stop over half of respondents in a recent global study from saying that they would use this nascent technology for sensitive areas …
Need to revise this title to proprietary LLMs. Offline and open source LLMs are a thing, a big thing, THE big thing that will displace all the proprietary junk soon. The problem is giving people your information and compromising your privacy with a company that only sees you as a revenue source, not a human, not a person, you have no rights except those insufficient odds and ends they must try to avoid, kinda, sorta. I mean the penalty is less than the cost of buying the data from one of the dozens of brokers that have been stalking you for the last decade and are selling the same thing.