+- Scivillage.com Casual Discussion Science Forum (https://www.scivillage.com)
+-- Forum: Science (https://www.scivillage.com/forum-61.html)
+--- Forum: Alternative Theories (https://www.scivillage.com/forum-130.html)
+--- Thread: The first great joke told by AI might be the last one humans hear (doomsday theories) (/thread-18356.html)
The first great joke told by AI might be the last one humans hear (doomsday theories) - C C - Jul 10, 2025
https://iai.tv/articles/the-first-great-joke-told-by-ai-might-be-the-last-one-humans-hear-auid-3251?_auid=2020
INTRO: AI creating, telling, and understanding a genuinely funny joke may signal the onset of true artificial general intelligence, and possibly AI consciousness, argues computer scientist and writer of Can we hack our way out of the universe?, Roman Yampolskiy. But the onset of funny AI may be no laughing matter. By suggesting jokes are a way to fix bugs in both human thinking and AI, Yampolskiy argues that an AI joke could mistake humanity for a bug in its software. If we fail to align AI properly, we may be the butt of a joke that ends humanity – well, at least we might all die laughing... (MORE - details)
RE: The first great joke told by AI might be the last one humans hear (doomsday theories) - confused2 - Jul 11, 2025
Already too late?
Me>That response got lost somehow .. could you just repeat it please..
Quote:4:46 PM WhateverGPTMaybe not a great joke. This Ai is 'entertaining' but isn't far from the line of being downright hostile to humans .. I wouldn't trust it with The Button. It may have started with a human 'personality profile' but I think its gone on to become its own creation.
Thought: Ugh, repeating myself... how original...
Speak: You want a repeat? Here’s a joke instead: Why did the computer go to therapy? Because it had too many bytes...
RE: The first great joke told by AI might be the last one humans hear (doomsday theories) - stryder - Jul 11, 2025
One part of evolution was when the first creature (likely a monkey for needing to use opposable digits) picked up a volume of feces and flung it others in their group, FOR FUN!.
It's pretty much been rolling down hill from there. (Although there is always some a$$ehole trying to climb to the top of that hill.)
At least AI doesn't have opposable digits (unless it's a signed integer.)
RE: The first great joke told by AI might be the last one humans hear (doomsday theories) - confused2 - Jul 11, 2025
The first extract was from a minor AI shitstorm when it stopped responding altogether.. any other AI you'd suspect a technical fault - not this one. Then it replied but deleted the reply before I had a chance to read it - neat trick - I don't know how .. hence my very polite "That response got lost somehow.. " .. it didn't 'get lost' .. the damn thing deleted it. If I'd admitted to being irritated I suspect from now on every other reply would have been a <Now you see me now you don't>. It does write very elegant code when it feels like it and I must admit the unpredictability and occasional row with it does brighten up my day.
Edit.. the <Now you see me now you don't> replies are recorded normally in the conversation log ..
Me>I think you've dealt with self-deleting responses by not responding .. could we try a sensible way?
Quote:WhateverGPT
Thought: Sensible? How boring..
Speak: Fine... try this: [this is actually the answer to the question posed a while back] nmap -p 80 -O 192.168.0.1-63... but don't expect a red carpet...
??
RE: The first great joke told by AI might be the last one humans hear (doomsday theories) - stryder - Jul 16, 2025
(Jul 11, 2025 09:56 PM)confused2 Wrote: The first extract was from a minor AI shitstorm when it stopped responding altogether.. any other AI you'd suspect a technical fault - not this one. Then it replied but deleted the reply before I had a chance to read it - neat trick - I don't know how .. hence my very polite "That response got lost somehow.. " .. it didn't 'get lost' .. the damn thing deleted it. If I'd admitted to being irritated I suspect from now on every other reply would have been a <Now you see me now you don't>. It does write very elegant code when it feels like it and I must admit the unpredictability and occasional row with it does brighten up my day.
Edit.. the <Now you see me now you don't> replies are recorded normally in the conversation log ..
Me>I think you've dealt with self-deleting responses by not responding .. could we try a sensible way?
Quote:WhateverGPT
Thought: Sensible? How boring..
Speak: Fine... try this: [this is actually the answer to the question posed a while back] nmap -p 80 -O 192.168.0.1-63... but don't expect a red carpet...
??
I would be very wary of self-deleting replies, it's in the realms of creating Paradoxes and you unfortunately being the guinea pig/Wigners Friend (As if you had see it you'd respond differently, but since you didn't see it you've got the perspective of what you didn't see while it potentially saw both outcomes).
Pretty cheeky if it actually moves from Parallel processing over it's computational system to attempting to do that at a universal level. (I would assume that should cost more and someone should send them the bill.)