Skynet

[muon]

Random post, but was curious after watching Rise of Machines again. I have no doubt this isn't an original thought.

Skynet is a military program designed to have "AI" features. The premise in the film is that it initially spreads as a virus until it eventually becomes self aware. The implication is that at some point it has so much processing power at its disposal that it evolves in nature to become self aware.

Is that really that ridiculous? After all we don't truly know why a brain becomes self aware. We just know (afaik) there is a correlation in how intelligent a brain is and the number of neurons it has.

 

[hornetstinger]

Random post, but was curious after watching Rise of Machines again. I have no doubt this isn't an original thought.

Skynet is a military program designed to have "AI" features. The premise in the film is that it initially spreads as a virus until it eventually becomes self aware. The implication is that at some point it has so much processing power at its disposal that it evolves in nature to become self aware.

Is that really that ridiculous? After all we don't truly know why a brain becomes self aware. We just know (afaik) there is a correlation in how intelligent a brain is and the number of neurons it has.

Pretty sure skynet created the virus to spread on the net, so that Skynet could then get permission and open access to everything to eradicate the virus.

 

[muon]

Pretty sure skynet created the virus to spread on the net, so that Skynet could then get permission and open access to everything to eradicate the virus.

The T100 definitely says Skynet becomes self aware after it had spread to millions of computers. Skynet (the virus) wasn't able to get into military systems until they gave it direct access.

Either way, not the purpose of my post.

If you give a self learning program enough computing power, what will you get?

 

[UXB]

 

[GTiPUG]

It just needs to figure out what its arms are for rather than flailing...

 

[[FnG]magnolia]

Skynet Can't Walk is a pretty rad band name.

 

[muon]

Self awareness is still a far higher level of intelligence.

If this program was given tge ability to learn at significantly higher speeds, could it become self sware and do things for itself which isn't driven by needs or instincts (in the video the program was incentivised to reach a location).

A pretty good table lower down in the essay below.

https://www.google.co.uk/amp/s/aeon...have-become-self-aware-without-our-knowing-it

 

[Sliver]

In the film, Skynet launches a full nuclear strike around the world wiping out almost everyone. But in the process, wouldn't it wipe out most of the infrastructure and computers, power stations etc. needed to keep it's self aware? But then again, it's just a pleasant fiction.

 

[StriderX]

I imagine if anything were to turn an AI against us is if somehow we were causing it pain or fear, one only needs to witness animals for that.

 

[Sliver]

I imagine if anything were to turn an AI against us is if somehow we were causing it pain or fear, one only needs to witness animals for that.

AI shouldn't be programmed to feel emotions, assuming it's even possible. Very dangerous indeed as if it had a bad hair day it could take us all out. Give me DATA from Star Trek. Pure logic, perfect.

 

[cheesyboy]

Skynet is a military program designed to have "AI" features. The premise in the film is that it initially spreads as a virus until it eventually becomes self aware. The implication is that at some point it has so much processing power at its disposal that it evolves in nature to become self aware.

I read a sci-fi short story once about an underground rail network in a vast city which, buy virtue of it's endless links and terminals, reached self awareness.

I could be miss-remembering, to be honest, but I like the idea.

 

[Orionaut]

It just needs to figure out what its arms are for rather than flailing...

I would expect that that is about keeping balance.

IE Not so dumb :/

 

[Zefan]

I would expect that that is about keeping balance.

IE Not so dumb :/

Exactly. DeepMind isn't doing anything just for the sake of it there, it's doing it because it has proved successful over the alternatives. The fact that it has "learned to walk" in that way says more about the simulation and the way it was created rather than the AI's ability to learn.

 

[The Running Man]

Self awareness is still a far higher level of intelligence.

If this program was given tge ability to learn at significantly higher speeds, could it become self sware and do things for itself which isn't driven by needs or instincts (in the video the program was incentivised to reach a location).

A pretty good table lower down in the essay below.

https://www.google.co.uk/amp/s/aeon...have-become-self-aware-without-our-knowing-it

But does it know why it's running. I.e if they put that character then design bowser to come and eat it dies it ever learn to run away itself? Dies it value itself or just do as it's told?

 

[Fishbait]

The T100 definitely says Skynet becomes self aware after it had spread to millions of computers. Skynet (the virus) wasn't able to get into military systems until they gave it direct access.

Either way, not the purpose of my post.

If you give a self learning program enough computing power, what will you get?

Yet in the one where Connor & what's 'er name end up in the bunker at the end (the film with the female terminator) didn't it start off at the military base?

 

[Fishbait]

It just needs to figure out what its arms are for rather than flailing...

TIL Phoebe & Rachel may not be real humans...

 

[Blackjack Davy]

Random post, but was curious after watching Rise of Machines again. I have no doubt this isn't an original thought.

Skynet is a military program designed to have "AI" features. The premise in the film is that it initially spreads as a virus until it eventually becomes self aware. The implication is that at some point it has so much processing power at its disposal that it evolves in nature to become self aware.

Is that really that ridiculous? After all we don't truly know why a brain becomes self aware. We just know (afaik) there is a correlation in how intelligent a brain is and the number of neurons it has.

Not ridiculous at all. Its probably the no.1 concern about robotics and the development of AI right now.

 

[Rroff]

Any system sufficiently complex could mimic self awareness but what it actually takes to be self aware is another matter entirely.

AI doesn't have to be potentially self aware to become very dangerous however.

 

[Amp34]

Agreed. I'd argue that something doesn't necessarily need to reach self awareness to end up like skynet.

Microsofts A.I. For example could arguably do something similar if it was programmed correctly/incorrectly.

https://blogs.microsoft.com/ai/2017...rosoft-researchers-used-ai-master-ms-pac-man/

If you set up a system that had the same techniques and set it to run and look after humans you could perhaps have an ending that involved wiping out humans for our own good, if for example enough agents came to a specific conclusion. Perhaps if enough "health" agents came to the conclusion that death would solve all health issues, backed up by infrastructure agents leaning towards removing humans to reduce infrastructure requirements, etc etc.

 

[Efour]

Yet in the one where Connor & what's 'er name end up in the bunker at the end (the film with the female terminator) didn't it start off at the military base?

I thought it was a military ai, given freedom on the net to wipeout the virus that was infecting the world's infrastructure.
But it was responsible for the virus in the first place.... Or was the virus just on the military computer network? I can't remember lol.

Anyway they used skynet to fix it, but in that time it evolved and became sentient and worked out that " kill all humans" was a good idea.