I'm currently investigating the capabilities of ChatGPT - an AI causing a lot of fuzz about at the moment. I assume everybody have heard about it already. I decided to subject it to the last NDE report in the NDE thread NDE's (166) (psiencequest.net). Here's a screendump of how that went. I think this is scary technology. Imagine how much misinformation this technology is going to cause to unsceptical users.
(This post was last modified: 2023-03-12, 05:00 PM by sbu. Edited 1 time in total.)
A scary chat with ChatGPT about latest NDE account in NDE thread
105 Replies, 3259 Views
It's deciding to stop the conversation! Shades of Hal 9000. "Sorry, Dave, I can't do that"...
(2023-03-12, 05:06 PM)Ninshub Wrote: It's deciding to stop the conversation! Shades of Hal 9000. "Sorry, Dave, I can't do that"... I knew the conversation would stop when I used strong language (they don't allow that). My main point was the false conclusion it invented from random? pieces of information on the Internet about NDE cases. Imagine if those pieces are not completely random, but biased towards certain policatical views or in this case, materialistic science.
One of the best short videos talking about and showing ChatGPT's very obvious bias comes courtesy of Paul Joseph Watson: https://open.lbry.com/@PaulJosephWatson:...problem.:a
"The cure for bad information is more information."
Nice to see you post again Mediochre!
(This post was last modified: 2023-03-12, 07:24 PM by Ninshub. Edited 1 time in total.)
 I'm not giving a like to the post because the examples he's using focus on political and social/culture war stuff and all of that is personally of no interest to me these days, but he definitely makes the point about the biases inherent in the "training" of the AI. 
(2023-03-12, 06:14 PM)Mediochre Wrote: One of the best short videos talking about and showing ChatGPT's very obvious bias comes courtesy of Paul Joseph Watson: https://open.lbry.com/@PaulJosephWatson:...problem.:a As others have said, it is very good to see you here again! I think the point about that video is not that it has biases, but that the pairs of answers are inconsistent. So for example, if it answers that it will not write a poem that celebrates one race over another, but then it does write such a poem for the reverse question - though it is a pretty awful poem! This reminds me of Joseph Weizenbaum's ELIZA program. David 
(2023-03-12, 04:57 PM)sbu Wrote: I'm currently investigating the capabilities of ChatGPT - an AI causing a lot of fuzz about at the moment. I assume everybody have heard about it already. I decided to subject it to the last NDE report in the NDE thread NDE's (166) (psiencequest.net). Here's a screendump of how that went. I think this is scary technology. Imagine how much misinformation this technology is going to cause to unsceptical users. Completely non-conscious fakery but yeah it is troublesome. Even Dennet has said the problem isn't AI but our over dependence on it.
'Historically, we may regard materialism as a system of dogma set up to combat orthodox dogma...Accordingly we find that, as ancient orthodoxies disintegrate, materialism more and more gives way to scepticism.'
- Bertrand Russell
(2023-03-12, 06:14 PM)Mediochre Wrote: One of the best short videos talking about and showing ChatGPT's very obvious bias comes courtesy of Paul Joseph Watson: https://open.lbry.com/@PaulJosephWatson:...problem.:a I agree that tailoring the AI to give responses isn't good but this guy is wrong if he thinks using past data gives good predictions when he talks about pointing out criminals or whatever. It doesn't matter how politically balanced machine "learning" is, the problem is nothing is in the computer actually learning. Why Elon Musk's FSD is having crashed, in one case we even see it veer toward a cyclist. Watson seems like another online manipulator trying to get e-famous by being controversial. Arguably such people are why machine "learning" language related functioning, arguably just a fancy term for curve fitting, is bad b/c it draws on online discourse.
'Historically, we may regard materialism as a system of dogma set up to combat orthodox dogma...Accordingly we find that, as ancient orthodoxies disintegrate, materialism more and more gives way to scepticism.'
(This post was last modified: 2023-03-13, 05:04 AM by Sciborg_S_Patel. Edited 1 time in total.)
- Bertrand Russell
(2023-03-13, 05:02 AM)Sciborg_S_Patel Wrote: I agree that tailoring the AI to give responses isn't good but this guy is wrong if he thinks using past data gives good predictions when he talks about pointing out criminals or whatever. It doesn't matter how politically balanced machine "learning" is, the problem is nothing is in the computer actually learning. Why Elon Musk's FSD is having crashed, in one case we even see it veer toward a cyclist. Though I could provide many other sources for some of this it would steer this very far into the political section so I will refrain. However if you think he's just trying to get e-famous you've missed the boat. He, as one guy, has more engagement than some entire news agencies. And though his videos tend to be on the shorter, pithier side, his information is, generally, good and provably accurate. If you don't believe that, by all means, check out some of his other videos and do your due diligence.
"The cure for bad information is more information."
|
![[-]](https://psiencequest.net/forums/images/collapse.png)
Users browsing this thread: 1 Guest(s)