Comment on Lawyers increasingly have to convince clients that AI chatbots give bad advice
ToTheGraveMyLove@sh.itjust.works 3 hours agoExcept with a plane, if you know how to fly it you’re far less likely to crash it. Even if you “can use LLMs” there’s still a pretty strong chance you’re going to get shit back due to its very nature. One the machine works with you, the other the machine is always working against you.
a4ng3l@lemmy.world 3 hours ago
Nha that’s just plain wrong…also you can also fantastically screw flying a plane but so long you use LLMs safely you’re golden.
It also has no will on its own; it is not « working against you ». Don’t give those apps a semblance of intent.
ToTheGraveMyLove@sh.itjust.works 3 hours ago
If I’m canoeing upriver, the river is working against me. That doesn’t mean it has a will. LLMs don’t need to have a will to work against you if your goal is to get accurate information, because by its very design it is just as likely to provide innnacurate information based on the way the tokens it applies to your query are weighted. You cannot control that. Its not plain wrong. Jfc, you slop apologists are fucking delusional.
a4ng3l@lemmy.world 3 hours ago
Slop apologist because I argue that correctly using a tool to restructure pre-existing information I’m inputting under my oversight is risk free?
You crazy ass end-of-world lunatic…
As far as I know slop always presupposes generation of derivatives, not restructuring or manipulation. You argue out of your ass and that’s just a bad opinion.