A scary chat with ChatGPT about latest NDE account in NDE thread

Nice to see you post again Mediochre! 

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.

That's fair, however I do think its an important thing to point out given these days "he who controls the algorithms controls the universe" as it were. And if political bias is being shoved in, there's every reason to think other biases are as well.

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.

He seems as biased as anyone else scrubbing for money by riling people up online.That he's successful doing so...more-so the pity.

My point is that his problem seems to be people making AI aren't supporting the politics he likes, rather than the problem being that letting computers do human work is genuinely a bad idea regardless of the side being advanced. It seems based on the video posted if they just pointed out criminals in the way he wants he would be fine with it.

The companies that offer this stuff up will be happy to offer products that cater to whatever narrative you want so long as the check clears. The political stuff is just smoke and mirrors frankly.

Nice to see you post again Mediochre! 

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.

Oh also thank you, I know I haven't been around for a long time but I also haven't been sitting still either. Been making progress with magic training, had an encouraging apparition show up last year, and other fun things. I do have more stuff to post but generally haven't gotten around to it but maybe now I can make the time.

[Laird]

As a mod:

There's a genuinely topical discussion to be had here re machine learning bias, including with respect to political matters, and because (1) it's admittedly difficult to separate that discussion from a purely political one, and also because (2) I know that you guys are already aware of the need to do that, there's scope for some leeway in moderating this discussion - but some posts have arguably already crossed the line, so I'd encourage participants to switch over to the opt-in forums if they want to continue the more explicitly political side of this discussion, to avoid one of us moderators from having to move posts there himself.

As an ordinary member:

It's also great for me to see you back, Mediochre. I hope you stick around.

As a mod:

There's a genuinely topical discussion to be had here re machine learning bias, including with respect to political matters, and because (1) it's admittedly difficult to separate that discussion from a purely political one, and also because (2) I know that you guys are already aware of the need to do that, there's scope for some leeway in moderating this discussion - but some posts have arguably already crossed the line, so I'd encourage participants to switch over to the opt-in forums if they want to continue the more explicitly political side of this discussion, to avoid one of us moderators from having to move posts there himself.

I think the fascinating thing about this conversation with ChatGPT, is that it can so easily be lead into contradicting itself. What we ideally need is a non-political discussion that is inherently contradictory. The problem is that nowadays far too many issues are considered political.

David

[Ninshub]

As a mod:

There's a genuinely topical discussion to be had here re machine learning bias, including with respect to political matters, and because (1) it's admittedly difficult to separate that discussion from a purely political one, and also because (2) I know that you guys are already aware of the need to do that, there's scope for some leeway in moderating this discussion - but some posts have arguably already crossed the line, so I'd encourage participants to switch over to the opt-in forums if they want to continue the more explicitly political side of this discussion, to avoid one of us moderators from having to move posts there himself.

I agree. The choice is to either to move the discussion if it continues into the opt-in forum, or for members to exercise restraint and have a meta-conversation. E.g. let's not get into data regarding what backs up specific socio-political positions. So just discuss it in other ways if possible.

EDIT: I just realized my post is unnecessary (the "I agree" part) in that this is Laird, and for some reason I was scanning the thread quickly and though it was Sci that was giving us this warning!

As a mod:

There's a genuinely topical discussion to be had here re machine learning bias, including with respect to political matters, and because (1) it's admittedly difficult to separate that discussion from a purely political one, and also because (2) I know that you guys are already aware of the need to do that, there's scope for some leeway in moderating this discussion - but some posts have arguably already crossed the line, so I'd encourage participants to switch over to the opt-in forums if they want to continue the more explicitly political side of this discussion, to avoid one of us moderators from having to move posts there himself.

As an ordinary member:

It's also great for me to see you back, Mediochre. I hope you stick around.

Fair

As a mod:

There's a genuinely topical discussion to be had here re machine learning bias, including with respect to political matters, and because (1) it's admittedly difficult to separate that discussion from a purely political one, and also because (2) I know that you guys are already aware of the need to do that, there's scope for some leeway in moderating this discussion - but some posts have arguably already crossed the line, so I'd encourage participants to switch over to the opt-in forums if they want to continue the more explicitly political side of this discussion, to avoid one of us moderators from having to move posts there himself.

As an ordinary member:

It's also great for me to see you back, Mediochre. I hope you stick around.

I'll do my best, thank you.

I learned over the last three years its that Yuval Noah Harari was right, humans really are just hackable animals. Feed them the right stimuli and apparently they will believe absolute absurdities with religious fervour despite mountains of evidence to the contrary. I'll believe the nature of humanity is different when it acts differently.

It would be possible to see in Achilles the Dionysiac strain, a passion for destruction growing out of a hatred for the destructibility of all things; and in Hector, the Apollonian part, the will toward preservation growing out of love for human achievements in their vulnerability.
 -Rachel Bespaloff

I'll do my best, thank you.

Don't forget, Mediochre, if you want a totally uncensored discussion, all you have to do is email me!

I do see the sense in censoring the public forums, but that doesn't mean that those censored have ceased to think what they think!

David