3 ms·
I would be exceedingly worried about someone trying to take medical advice from chatgpt, given that chatgpt's training data set will include anything from mayoc
by PuppyTailWags 4y ago
I would be exceedingly worried about someone trying to take medical advice from chatgpt, given that chatgpt's training data set will include anything from mayoclinic to mommy blogs to new age stuff. Similarly, I worry about law, because chatgpt's training data will include old/expired laws mix with supreme court opinions mixed with sovreign citizens nonsense.
- anonporridge 4y agoI consider ChatGPT to simply be a word like Kleenex or Google at this point. A brand name that has become synonymous with the generic. Also, don't discount the vast amount of cooky doctors and lawyers with fully legitimate licenses. The humans who occupy these roles make countless mistakes, and so will and AI version, at least initially.
- PuppyTailWags 4y agoAt least if a doctor or a lawyer gives you bald-faced wrong information that meaningfully harms you, you have some, limited recourse, and that doctor/lawyer takes on some, limited risk if they're found to be straight up negligent. With ChatGPT there is no recourse and no risk to spew harmful falsehoods. [To be clear: I state how limited this is because if doctors were genuinely held to give beyond cover-your-ass standard of care, there would be way less medically traumatized disabled people out there. Unfortunately...]