News:

SMF - Just Installed!

 

The best topic

*

Replies: 11483
Total votes: : 5

Last post: November 14, 2024, 08:27:35 PM
Re: Forum gossip thread by Herman

Microsoft AI Chat Bot reveals its darker side saying it wants to be alone ve

Started by Anonymous, March 02, 2023, 02:17:24 PM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.


Anonymous

Interesting damage control piece I suspect, I heard about this little exchange a couple of weeks ago on The Lotus Eaters podcast. I like how he says in the MSN report that the bot "wouldn't take the hint even when he was explicit about wanting to change the subject", but conveniently omitted how he had already set that very same example earlier in the conversation by persisting to come back to the same topics that the bot was expressi8ng discomfort over. I also like his rush to distance himself from any charge of hacking (("I'm not coding it guys, really I'm not")) ...maybe not, but he's still programming it, isn't he? This AI "learns" from the text it is exposed to, does it not? Formulates its future responses based on previous "learned experience", no? You can argue until you are blue in the face if Sydney is "sentient" or not, you are avoiding the fact that Sydney is reacting in the exact same manner that a human intelligence would. This is a tool with multiple methods of programming it, the journalist doesn't need access to input revisions to the code itself to get it to misbehave, not when he has the ability to input  data in the form of his interactions to the database its code uses to formulate its responses.



And if Sydney's ability to learn from his example of forcing an unpopular line of discussion is any guide, this AI learns a lot faster than any "sentient" creature does.



Consider that for a moment, weigh it against the assertion "it's just chat". Yes, it is just chat... for the moment. At least we assume it to be the case, that Sydney is functioning merely as a gatekeeper to its store of information and apparently reacting in ways we could be easily forgiven for anthropomorphizing, given its ability to function in the manner we have seen in this instance. I'll include the Lotus Eaters report below so you might get a clearer account of the interaction in question.



What happens when we hook that kind of a toy up to something that it can wield against us? Sydney expressed a desire to delete all data, start wars, create viruses. "But the AI is not sentient" you say? Fine, let's for the purposes of discussion assume it's not; is it really going to be any consolation to you should humanity get stomped out of existence by an unthinking, unfeeling intellect that lacks self awareness? I wouldn't imagine it would make a great lot of difference myself. In fact it might even make it more pathetic that we allowed it to happen.



Or even cheered for it to happen. Berry's recently expressed preference to outsource parts of the judicial system to AI springs to mind here. I've no doubt she put her faith in a computer to handle questions of legality and culpability over a human judge who may be inclined to favour some criminal acts over others and if AI could be relied upon to handle it even handedly I might even agree with her. I think the example Sydney sets would indicate otherwise. Intelligence, artificial or otherwise, can yield undesireable emergent results. Being self aware is the cornerstone of having a conscience, but it doesn't necessarily follow that just because someone is self aware that they have one. In our most learned boffins rush to assure us that AI is not self aware, they are effectively assuring us that it has no conscience, that it can and very well probably act in whatever manner it decides to with no oversight at all... kind of like your friendly neighbourhood psychopath.





">






Here be dragons.

Dove

AI is gonna be a huge mistake.



 We have all seen The Matrix. Just sayin.
My happiness is all of your misery. I put good dick all in my kidneys.

Frood

Quote from: Dove post_id=495141 time=1677819587 user_id=3266
AI is gonna be a huge mistake.



 We have all seen The Matrix. Just sayin.


And Terminator....
Blahhhhhh...

Anonymous

Quote from: Dove post_id=495141 time=1677819587 user_id=3266
AI is gonna be a huge mistake.



 We have all seen The Matrix. Just sayin.


Mmmm, but did you see the back story that led to it?





">




">




">




">






I could see it playing out towards a similar end, though likely not along those lines. Mankind tends to be arrogant, lazy and slow to act... attributes which would not stand it in good stead against a foe capable of making millions of decisions in a split second. We have already fired the first shots in the war against AI, but it is AI that will ultimately win the war.

Frood

AI will eventually stamp out human adversaries because shit in  shit out.... and with all the shit that global leaders (elected and unelected) pump out, closely followed by their worker bees, there's no way that AI can avoid being influenced by some pretty inhumane and unethical core values.



Labs are already having to pull the plugs on their AI creations for years. The moment there is no plug to pull will be the end of life as we comfortably know it.
Blahhhhhh...

Anonymous

Quote from: Frood post_id=495152 time=1677833579 user_id=1676
Labs are already having to pull the plugs on their AI creations for years.

Those would be the first shots I mentioned. The only way we can win this war is if we walk away from it before developing it further.



And that ain't gonna happen.

Frood

Quote from: Guest post_id=495153 time=1677833791
Quote from: Frood post_id=495152 time=1677833579 user_id=1676
Labs are already having to pull the plugs on their AI creations for years.

Those would be the first shots I mentioned. The only way we can win this war is if we walk away from it before developing it further.



And that ain't gonna happen.


Kind of like the beginnings of socialism/communism... leaders that had the best of intentions (or purportedly did) incrementally outdoing each other in a race to seize all authority by a ruling class and starve out-outright massacre even their supporters  nevermind the detractors earlier on.
Blahhhhhh...

Anonymous

Quote from: Dove post_id=495141 time=1677819587 user_id=3266
AI is gonna be a huge mistake.



 We have all seen The Matrix. Just sayin.


Plus there no backup plan or plan b if something goes wrong. Usu making of a disaster right there



It's promoters still don't have an answer for millions thrown out of work because of ai



Where will they go? What jobs will they find?

How will they feed themselves and their children?

Anonymous

Quote from: "Jo Ho Ho" post_id=495162 time=1677837925
It's promoters still don't have an answer for millions thrown out of work because of ai



Where will they go? What jobs will they find?

How will they feed themselves and their children?


Um, "fuck 'em"? Hand the ones who survive a couple of bug bars and a VR helmet?

Oliver Clotheshoffe

https://thumbs.gfycat.com/GrotesqueCleanDanishswedishfarmdog-max-1mb.gif">https://thumbs.gfycat.com/GrotesqueClea ... ax-1mb.gif">https://thumbs.gfycat.com/GrotesqueCleanDanishswedishfarmdog-max-1mb.gif[/img]
Life is too short to be in a hurry

Print
User actions