Originally published at: NEWS AND VIEWS FROM THE NEFARIUM JULY 2 2026
Podcast: Play in new window | Download (Duration: 10:44 — 17.2MB)As we head into the Fourth of July holiday, there’s much to contemplate, including this story shared by M.C.: You can persuade AI models to accept falsehoods as truth, study shows
4 Likes
Does AI understand time? Even if you could program in an equation of time, would it understand the implications / ramifications. Does it need time to work out the experience first to become reliable and mature into functionality without psychosis? That, next to all the bias language programed into it.
1 Like
If people die…?
Well, all the better…?
As that ties int another Live-in-Action ongoing OP: Depopulation.
So much fodder[useless eaters].
Only so many Ops that can contibute.
[Due to time constraints.]
ALL-Too-Soon; they’ll be only 500,000 Earthlings left?
Do “THEY” have the Stones for that?
YEP! [Puposed double entendre]
Georgia, is just one place; that they’ve made sure gave warning.
1 Like
Expecting real from “artificial” is delusional. How easily buy-in was accomplished, both financially and intellectually, seems to show how willing humans are to deny reality and suffer the consequences for profit and the idea of progress.
Apropos your reference to the medical risks of AI … a ChatGPT test was undertaken for veracity / validity by published medical experts BMj and Lancet now admit that a: They are receiving high volumes of fake research >50% estinated; B) they are increasingly unable to disconfirm assertions/findings and/ or replicate the studies to ascertain reliable or valid results. https://fb.watch/I6BHFXAtBw/
Abstract :
Objective: To evaluate the quality of the answers and the references provided by ChatGPT for medical questions.
Patients and Methods: Three researchers asked ChatGPT 20 medical questions and prompted it to provide the corresponding references. The responses were evaluated for the quality of content by medical experts using a verbal numeric scale going from 0% to 100%. These experts were the corresponding authors of the 20 articles from where the medical questions were derived. We planned to evaluate 3 references per response for their pertinence, but this was amended on the basis of preliminary results showing that most references provided by ChatGPT were fabricated. This experimental observational study was conducted in February 2023
.
Results: ChatGPT provided responses varying between 53 and 244 words long and reported 2 to 7 references per answer. Seventeen of the 20 invited raters provided feedback. The raters reported limited quality of the responses, with a median score of 60% (first and third quartiles: 50% and 85%, respectively). In addition, they identified major (n¼5) and minor (n¼7) factual errors among the 17 evaluated responses. Of the 59 references evaluated, 41 (69%) were fabricated, although they appeared real. Most fabricated citations used names of authors with previous relevant publications, a title that seemed pertinent and a credible journal format.
Conclusion: When asked multiple medical questions, ChatGPT provided answers of limited quality for scientific publication. More importantly, ChatGPT provided deceptively real references. Users of ChatGPT should pay particular attention to the references provided before integration into medical manuscripts.
Source:
Gravel, Jocelyn & D’Amours-Gravel, Madeleine & Osmanlliu, Esli. (2023). Learning to Fake It: Limited Responses and Fabricated References Provided by ChatGPT for Medical Questions. Mayo Clinic Proceedings: Digital Health. 1. 226-234. 10.1016/j.mcpdig.2023.05.004.