Artificial Intelligence

[FireFox]

I've used the search function and can't find anything regarding this so I thought I'd create a topic, if a thread already exists about this then I apologise in advance.

 

What are your opinions/thoughts or whatever on AI?

 

My views are that we shouldn't (though as someone who likes technology it is bloody cool) be pursueing it because a machine that can think for themselves could be a bad thing (hello Terminator), I was reading an article about scientist trying to create AI and one of the guys wrote that it'll be easier to create bad/potentially dangerous AI then it would be to create good/friendly AI, I'll have a look for the article but in the mean time I came across this which I found interesting.

 

http://www.nature.com/news/quantum-boost-for-artificial-intelligence-1.13453

[M-ursu]

Interesting indeed.  :D

[legopc]

it would be cool to have artificial intelligence but i dont think that that terminator stuff will happen if we get that far 

[FireFox]

it would be cool to have artificial intelligence but i dont think that that terminator stuff will happen if we get that far 

 

I sure hope so lol

[Samdb]

I really hope this doesn't turn in to another Skynet

[rashdanml]

Unknown to us, SkyNet already exists in the form of Google. It's simply dormant right now. They keep adding more and more features to it, so that when they do unleash it on the world, we're screwed.

 

Being serious, it's a nice thing to think about, but the potential pitfalls make me apprehensive. We'd need to consider how it'll be controlled before we advance AI in leaps and bounds.

[mozoa]

don't give it production facilities.

or guns.

[Emperor_Piehead]

I would think the VI that mass effect had would be a better alternative to AI or like in Halo give the AI a feeling of humanity so that it wouldn't try to kill all humans.

[Hyydrah]

We already have AI that can kill us. Look at the Phalanx CIWS for example.

[FireFox]

We already have AI that can kill us. Look at the Phalanx CIWS for example.

 

But it doesn't think or itself, we've programmed it.

 

That is no more intelligent than a household computer

[beebskadoo]

Ai is going to be a game changer. It will change the way we do things and open the doors to many new things.

[FireFox]

Ai is going to be a game changer. It will change the way we do things and open the doors to many new things.

 

Just hopefully one of those doors won't be our armageddon lol

[Hyydrah]

But it doesn't think or itself, we've programmed it.

 

That is no more intelligent than a household computer

Not sure what you mean, but it does make decisions for itself for the most part (judging whether an unidentified flying object is a threat, deciding how to engage the target, etc.) I've been told that the only human intervention required is for maintenance and to authorize fire (the latter of which doesn't have to be done).

[FireFox]

Not sure what you mean, but it does make decisions for itself for the most part (judging whether an unidentified flying object is a threat, deciding how to engage the target, etc.) I've been told that the only human intervention required is for maintenance and to authorize fire (the latter of which doesn't have to be done).

 

Still not actually thinking for itself though is it, it may make decisions about potential targets through whatever was programmed into it

[lorekeeper]

The thing is that all these AI`s are just that, artificial- they do nothing but follow strict instruction sets and as such any malicious actions they take are solely a reprecution of human actions. In time a situation where the idea of a ``singularity`` may arise (though lets be clear we are nowhere near this point, the problem lies in the fact that computers function in a very literal fashion) then whatever being is created is no longer artificial and thus should be treated as any other sentient being. The idea that they would be out to kill us and that we should thus avoid this thecnological path all together is irrational. Anyhow in truth should we create sentient beings it stands to reason that we could also define their nature and thus make them into peacefull beings who are neither pridefull, greedy, ignorant nor agressive - in other words we could remove the flaws in our own nature. Now theres always the argument that some malevoyant government or corporation would create killing machines to serve their own purposes however once again this is a purely human issue with any consequences arising as a result of our own actions. In summation i doubt these sentient beings- if properly managed- could ever be as dangerous to humans as we are to ourselves. Though this is meerly an oppinion and i would be intersted to see this discution continue.

[Hyydrah]

Still not actually thinking for itself though is it, it may make decisions about potential targets through whatever was programmed into it

That's why it's called artificial intelligence. :P

 

"The theory and development of computer systems able to perform tasks that normally require human intelligence, such as visual perception...". Precisely what the Phalanx CIWS is doing.

[FireFox]

That's why it's called artificial intelligence. :P

 

"The theory and development of computer systems able to perform tasks that normally require human intelligence, such as visual perception...". Precisely what the Phalanx CIWS is doing.

 

Fair enough :P

 

I've not read in depth with it yet

[Hyydrah]

Fair enough :P

 

I've not read in depth with it yet

You might find this interesting.

 

 

He also has other great videos. He's one of the best science "celebrities" in my opinion.

[T.Vengeance]

 

At 5min 20 sec. Quantum computers are only gonna help if the AI needs some number crunching. Beyond that, it shouldn't help much.

[T.Vengeance]

You might find this interesting.

 

 

He also has other great videos. He's one of the best science "celebrities" in my opinion.

Nope that prize belongs to Neil deGrasse Tyson :P

[Hyydrah]

Nope that prize belongs to Neil deGrasse Tyson :P

I said "one of the best".  :P He's definitely up there as well.

[WanderingFool]

The thing is that all these AI`s are just that, artificial- they do nothing but follow strict instruction sets and as such any malicious actions they take are solely a reprecution of human actions. In time a situation where the idea of a ``singularity`` may arise (though lets be clear we are nowhere near this point, the problem lies in the fact that computers function in a very literal fashion) then whatever being is created is no longer artificial and thus should be treated as any other sentient being. The idea that they would be out to kill us and that we should thus avoid this thecnological path all together is irrational. Anyhow in truth should we create sentient beings it stands to reason that we could also define their nature and thus make them into peacefull beings who are neither pridefull, greedy, ignorant nor agressive - in other words we could remove the flaws in our own nature. Now theres always the argument that some malevoyant government or corporation would create killing machines to serve their own purposes however once again this is a purely human issue with any consequences arising as a result of our own actions. In summation i doubt these sentient beings- if properly managed- could ever be as dangerous to humans as we are to ourselves. Though this is meerly an oppinion and i would be intersted to see this discution continue.

 

The singularity is very far away yes, but in terms of "artificial" the question becomes when does a digital computer become more than just "logic".  In some aspects you could say we, as humans, are just following logic (logic built within the neurons).  With modern computers, the thing that hurts them most is the limited parallelness of their processing.  A human can take in input, and within a tenth of a second, millions of neurons have received the signal and are in turn firing to other neurons to process the data.  So if a practical quantum computer were to be made then the singularity might not be as far away as one might think.  (I know there are quantum computers out there, but in all practicallity they have to be hand crafted to fit certain problems...unlike modern day processors which can run more arbitrary problems).

 

For tackling the logic, I would like to provide you with some information on Neural Networks.  Neural Networks usually simulate the neurons in your brain.  This is very much just a large simulation, but there are computers that have neurons that mimic the memory portion of your brain and can actually "remember" lists.  The memory is "fuzzy", which happens in our brains as well, being able to remember words at the beginning and end easier than ones in the middle....I wish I bookmarked the article at the time, but it is out there somewhere.

 

Remember we are just machines, neurons control what we do and think, as more accurate simulations exist we might be able to create a "living" creature in a digital world that is capable of learning in a non-classical sense of computer learning.  With that said, here is a good article about a new Neural Network (http://www.sciencedaily.com/releases/2013/08/130802080237.htm), quantum computing could make an even more powerful computer brain :p