this post was submitted on 10 Dec 2024
181 points (98.9% liked)
chapotraphouse
13594 readers
613 users here now
Banned? DM Wmill to appeal.
No anti-nautilism posts. See: Eco-fascism Primer
Slop posts go in c/slop. Don't post low-hanging fruit here.
founded 4 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
Amazing, it can pass tests which it churned through 1000 times but cannot produce simple answer a child might stumble through. It's not cognition, it's regurgitation. You do get diagnosed at llm-shop mate, have fun
Yeah you're right! What use is having the entirety of medical knowledge in every language REGURGITATED at you in a context aware fashion to someone who can't afford a doctor? After all it's not cognition in the same way that I do it.
How many shitty doctors getting nudged towards a better outcome for real people does this tech need to demonstrate to offset the OCEAN BOILING costs of this tech do you think?
Can't wait to pick up my prescription for hyperactivated antibiotics.
https://www.cio.com/article/3593403/patients-may-suffer-from-hallucinations-of-ai-medical-transcription-tools.html
How often do you think use of AI improves medical outcomes vs makes them worse? It's always super-effective in the advertising but when used in real life it seems to be below 50%. So we're boiling the oceans to make medical outcomes worse.
To answer your question, AI would need to demonstrate improved medical outcomes at least 50% of the time (in actual use) for me to even consider looking at it being useful.
50% is the number yeah? I wish yall took "no investigation no right to speak" more seriously.
They've provided a source, indicating that they have done investigation into the issue.
The quote isn't "If you don't do the specific investigation that I want you to do and come to the same conclusion that I have, then no right to speak."
If you believe their investigation led them to an erroneous position, it is now incumbent on you to make that case and provide your supporting evidence.
Y'all are suffering because of the lack of downvotes, so you need to actually dunk on someone instead of downvoting and moving on
We need to make a chat gpt powered dunking bot
ChatGPT is censored, this calls for some more advanced LLMing, perhaps even a finetune based on the Hexbear comment section argument corpus. It's only ethical if we do it for the purpose of dunking on chuds/libs