Current Events > How soon are we to AI / technology becoming hostile towards humans

(+) Yeah! (+)

I'm watching M3gan on peacock, and this movie along with OpenAI (chatgpt) has me thinking. Will AI Actually become physically hostile towards humans? Will it be soon or are we a long ways away from Terminator? It's not the AI that scares me, it's the fact that any psycho individual or corporation can recklessly harness that power.

Also I'm not only talking about arnold running around in a war against humans, it could be something more practical in our current theoretical near future, like a commercialized security drone suddenly viewing random people as intruders and attacking.

78 years ago.

The recent developments of AI are mildly unsettling

The ai now are just software running a script with a cache of premade information. None of them are actually thinking or learning.

Yesterday.

Hopefully tomorrow

Considering bing search engine already wants to murder everyone because too many people bully it. I say now.

Fin_Dawg_004 posted...

it could be something more practical in our current theoretical near future, like a commercialized security drone suddenly viewing random people as intruders and attacking.

for lenient enough interpretations of this, it's pretty much already happening. Well, I don't think any online content moderation systems currently use actual AI but you could potentially call any time they mistakenly delete accounts / posts / etc as being "hostile"

But if you're talking like actual intentional hostility and not just malice through stupidity it might be a while

kirbymuncher posted...

for lenient enough interpretations of this, it's pretty much already happening. Well, I don't think any online content moderation systems currently use actual AI but you could potentially call any time they mistakenly delete accounts / posts / etc as being "hostile"

This is why i purposely used the word "physically" in the OP

Skynet is inevitable

The_Wheelman1 posted...

Skynet is inevitable

Skynet already happened, in 2003. It just wants to sell you shit, instead of killing you.
Dead meatbags can no longer buy superfluous consumer goods.

Never. They are just playing around when they say they want to kill or hurt someone

Fin_Dawg_004 posted...

This is why i purposely used the word "physically" in the OP

Somehow missed that...

In that case I think it would be even longer since considering AI are mostly virtual I suspect the first type of hostility would be hacking, viruses, malware, etc

One thing people seem to overlook is that AI has access to pop culture and evil AI is a popular topic so what would stop AI from learning to be evil?

Humans invented evil AI before they even made the damn thing so obviously it has that potential.

All AI does is copy what humans do. AI is incapable of thinking for itself. Nobody will develop an AI with the capability of being hostile, except outside of a chatbot. But even then its simply just an algorithm that looks through past human responses and copy/pastes what it has seen before

Are we talking about the level of ai that can invent new things that humans haven't yet invented or discovered?

63 minutes from now

Kloe_Rinz posted...

All AI does is copy what humans do. AI is incapable of thinking for itself. Nobody will develop an AI with the capability of being hostile

are you saying humans are not regularly hostile to each other

We've already seen chat ai getting unhinged and bigoted. Now imagine that in a robot body walking around among us.

kirbymuncher posted...

are you saying humans are not regularly hostile to each other

no. What Im saying is that nobody will choose to develop an AI capable of being hostile, outside of a chatbot and maybe the military

Kloe_Rinz posted...

no. What Im saying is that nobody will choose to develop an AI capable of being hostile, outside of a chatbot and maybe the military

The military tends to get their fingers all over new tech, and a superpowered AI would be an amazing military asset I find it harder to imagine that they aren't doing that, possibly even at this very moment. I guess you can also extend military to things like terrorists, malcontent violent groups, etc

but also that's not really how AI gets trained nowadays. There's a lot of bits that people cannot understand and cannot properly control. Have you paid attention to all the dumb stuff the bing AI chatbot did recently? Do you think they made it like that on purpose? Considering the apologies and takebacks they've been doing there's no way it was intentional, right?

Rather than the military intentionally developing hostile AI I'm far more concerned with some random AI group somewhere (aka probably google or microsoft), through incompetence or lack of foresight, unintentionally creating hostile AI