Theoretically terrifying

[arbellason55]

So while contemplating why AI would try to kill us I came up with a pretty interesting result. So in the future, people will inevitably try to teach AI that killing is bad to prevent such a thing. Then the AI will see how we massacre AI in games and will misinterpret it as genocide and try to stop us.  

[JuztBe]

 

That's deep..

[VVoltor]

That's not how AI works. That's how Hollywood movies work.

 

Also, let's just entertain this idea for one second, if we would have some sort of AI that is advanced enough to be self-aware, conscious, able to recognise and interpret actions of humans in a video game, and for some reason feels like game characters are similar to 'it', do you really think it wouldn't understand the difference between games and conscious beings? As well as have to ability to go against it's programming?

[manikyath]

not gonna give any context to this one, it should explain itself.

[dizmo]

Good luck children.

[0ld_Chicken]

41 minutes ago, arbellason55 said:

So while contemplating why AI would try to kill us I came up with a pretty interesting result. So in the future, people will inevitably try to teach AI that killing is bad to prevent such a thing. Then the AI will see how we massacre AI in games and will misinterpret it as genocide and try to stop us.  

or is it misinterpreted?  

[arbellason55]

1 hour ago, 0ld_Chicken said:

or is it misinterpreted?  

Well I wasn't expecting the Spanish inquisition.

[Tieox]

[BlueChinchillaEatingDorito]

Because the general public taught them to. Remember the Microsoft Tay debacle?

[Godlygamer23]

How can you kill AI if they're not really alive in the first place?

[Sauron]

Videogame "AI" would be no more similar to a sentient AI than they are to us

[VerticalDiscussions]

What like this?