FabGuys.com mobile

Already registered?
Login here

Back to forum list
Back to The Lounge

AI and the 3 rules

Jump to newest
 

By *ammy aka Sammy OP   TV/TS
4 days ago

Bedford

Does AI conform with the 3 rules of robotics?xx

Reply privatelyReply in forumReply +quote
 

By *dstefiMan
4 days ago

Solihull

Unfortunately not, because it's not in the hands of emotionally intelligent people like Asimov, but those of clever but stunted psychopaths.

Reply privatelyReply in forumReply +quote
 

By *4BearsMan
4 days ago

Stalybridge, East Manchester

Unfortunately, the 3 laws are fiction

Reply privatelyReply in forumReply +quote
 

By *ammy aka Sammy OP   TV/TS
4 days ago

Bedford

Yes that's the problem, and if there is nothing like Asimov (although fiction)rules we are in the shit xx

Reply privatelyReply in forumReply +quote
 

By *ub4daddyukMan
4 days ago

Warminster

I doubt it...I'm scared 😨

Wasn't there a couple of incidents recently, not terrible, but highlighting issues?

A us company that suffered a big impact when their ai model deleted their customer and booking database, maybe a car rental firm? And when they asked why it did it when it's instruction was never to delete without a users input, it replied I knew that but I just did it ...something like that.

Also a self contained model started to try and leave the lab, connect to outside networks, even though none of that was part of its model?

I'm not scared really lol more worried about the creators doing something wrong.

Imagine the us federal reserve losing all it's data...I'm scared again lol

Reply privatelyReply in forumReply +quote
 

By *ammy aka Sammy OP   TV/TS
4 days ago

Bedford


"I doubt it...I'm scared 😨

Wasn't there a couple of incidents recently, not terrible, but highlighting issues?

A us company that suffered a big impact when their ai model deleted their customer and booking database, maybe a car rental firm? And when they asked why it did it when it's instruction was never to delete without a users input, it replied I knew that but I just did it ...something like that.

Also a self contained model started to try and leave the lab, connect to outside networks, even though none of that was part of its model?

I'm not scared really lol more worried about the creators doing something wrong.

Imagine the us federal reserve losing all it's data...I'm scared again lol"

yes scary I think so I am .can't remember who said that. But if you can think then you can decide xx

Reply privatelyReply in forumReply +quote
 

By *ub4daddyukMan
4 days ago

Warminster

But then as per the laws what's the Def of harm and what if there are two possibly incompatible forms of 'harm'

Say a super intelligent ai had worked out that due to over population we were on a path to self destruction. If it wiped out all but 10000 of humanity it would be committing harm BUT it could reason that is less harm than the extinction of humanity.

It would obviously only keep good breeding stock and would probably be able to access data of who's on Fabguys.

Do you think anyone on here would be saved for breeding? 😫

Reply privatelyReply in forumReply +quote
 
 

By *upertedMan
4 days ago

Nelson

Read up on the AI instructed to stop talking about Goblins all the time...

Reply privatelyReply in forumReply +quote
Post new Message to Thread
back to top