Artificial intelligence-why?

[Neonsilver]

With all the talk of the coming singularity, and computers destined to become 'smarter' than humans, many have discussed the possibility of completely artificial minds, complete with individual personalities.

Well, I have thought about this: even if they could be made, why would they? What could an AI do that a person couldn't. And they would have all the flaws a person would.8

Most flaws would have to be programmed into the AI.

An AI:

Is intelligent, and therefore makes judgments. That means it could make the WRONG judgment.

So what?
The software we use right already can make wrong judgements.

Can still be a bad 'person'. If it makes its own judgments and choices, and has its own personality, it could just as well decide that it wants to have a 'bad' personality, being aggressive, unhelpful, just plain rude, etc.

The basic AI would probably only use logic based on the data it has available and the goals that were programmed into it. It probably wouldn't have something like a human personality.

May have access to vast amounts of data or systems. An error on its part would be no less devastating than one made by a human.

Would cost a whole lot of money.

As mentioned before, errors wouldn't be as bad as you think. It depends on the cost of the errors and the cost of being always correct.

The only use I can see an AI being put to would be to operate large amounts of highly complex machinery, or to analyze vast amounts of data. But in either of those situations, an error on the part of the AI could cause a huge amount of damage, possibly more so than a human because the AI is far more centralized.

Here's the thing though, an AI would cost a huge amount of money. If you are ready to spend all that money, you might as well just hire a team of analysts or machine operators.

So even though AIs COULD exist, do you think they actually will?

An AI would be able to learn on it's own, you could give it a task and it would learn on it's own how to perform it. So once you have a working AI you could save on money because you don't need new software

 

[deadish]

Well, the brain isn't turing-complete. In order for a computer to be a computer, it must be turing-complete.

Wait? What?!

My brain can simulate a single tape turning machine, which is the definition of a turning-complete, just fine thank you very much.

 

[zehydra]

Well, the brain isn't turing-complete. In order for a computer to be a computer, it must be turing-complete.

Wait? What?!

My brain can simulate a single tape turning machine, which is the definition of a turning-complete, just fine thank you very much.

No it can't.

 

[deadish]

Well, the brain isn't turing-complete. In order for a computer to be a computer, it must be turing-complete.

Wait? What?!

My brain can simulate a single tape turning machine, which is the definition of a turning-complete, just fine thank you very much.

No it can't.

LOL.

Dude, if we can't simulate a tuning machine in our heads, we won't be able to understand it much less build one.

PS: Oh yes, I don't recall anywhere stating that computers have to be turning-complete to be called computers.

 

[zehydra]

Well, the brain isn't turing-complete. In order for a computer to be a computer, it must be turing-complete.

Wait? What?!

My brain can simulate a single tape turning machine, which is the definition of a turning-complete, just fine thank you very much.

No it can't.

LOL.

Dude, if we can't simulate a tuning machine in our heads, we won't be able to understand it much less build one.

PS: Oh yes, I don't recall anywhere stating that computers have to be turning-complete to be called computers.

I'm sorry, I mean that the human brain cannot simulate every possible turing machine, whereas a computer can. That may not be "turing-complete"-ness but I can't recall what it is.

I do know that that is a fundamental difference between human brains and computers

 

[Hagi]

There can be no such thing as a "sentient AI" since such a thing is a contradiction in terms.

For something to be sentient, it cannot be an "artificial intelligence".

Because... reasons?

Allow me to provide you with a through experiment:

Scientists develop a device the size of a single human neuron which acts exactly the same in all respects. They then create a giant network of these devices and add additional devices that act exactly the same as all other cells, hormones and processes present in the human brain. They have created, for all intents and purposes, a human brain.

Except they made it. It's an artefact. It's artificial.

Would this not be a sentient artificial intelligence?

Artificial intelligence is not what you have described. You have described an artificial brain. The intelligence which the artificial brain creates, however, is not artificial.

That is artificial intelligence.

At least that's the definition used by universities. An intelligence created by an artificial device.

How else would you define an artificial intelligence?

AI is essentially a set of computer algorithms designed to either appear as if an intelligence is controlling the outcome, or a series of algorithms designed to compete with actual intelligences (for instance, a computer playing chess).

The difference between an AI and an actual intelligence is that an AI is purely algorithmic, whereas an actual AI is not. (note that this is not the same thing as being deterministic)

Your definition is at odds with the academic world then.

The academic field of AI is focused on creating intelligence. This intelligence would be artificial AKA an artefact, by virtue of being created.

That's what artificial means. Created. Made by humans.

An artificial intelligence is an intelligence that was created by humans. Nothing more, nothing less. It is an actual intelligence, if it wasn't then it wouldn't be called an artificial intelligence but instead an artificial algorithm or whatever.

You've got your definitions mixed up mate.

Well, in the Computer Science Academic world at least we aren't at all concerned with creating actual intelligence. Just intelligence that can compete with non-artificial intelligence.

I'm not even sure what you're saying anymore...

You aren't concerned with creating actual intelligence. Instead you're concerned with creating intelligence?

What's the difference between intelligence and actual intelligence?

EDIT: It's starting to sound suspiciously like a no true Scotsman fallacy...

Ok, so we have two things, AI and real intelligence. Real intelligence is generated by things like brains and neurons.

Generally speaking, when people refer to Artificial Intelligence, we refer to Computer Artificial Intelligence. In theory, as you said, we could make a artificial neurons and artificial hormones etc. However, what you would create is not the same as Computer Artificial Intelligence. That is, no set of computer algorithms can produce the results of a system of artificial neurons + artificial hormones.

The brain, while deterministic, does not behave in the same way as a computer, that is, it does not behave algorithmically. For this reason, it is quite difficult to program a computer to do certain things which a brain can do fairly easily and vice versa.

The Academic world of Computer Science is only concerned with making an artificial intelligence capable of imitating intelligence generated by a brain, not with actually creating the same intelligence.

Bottom line: AI is not an intelligence. It's just the name we give to algorithms which produce outputs which mimic brain-created intelligence.

You do know that there's algorithms that imitate neurons right?

Neural networks are quite a fascinating subject and they're a computer algorithm that imitates the human brain.

Of course, it's a long, long way off anything actually approaching the human brain but the basis is there.

Next to that we simply don't understand that much of the brain. This means that it's impossible to say that the brain does not behave algorithmically. We don't know for certain how the brain behaves deep down, and by extension we also don't know how it does not behave deep down.

And I don't what world of academic computer science you're a part of but it's apparently not the same one that's at my university because that one is concerned with 'actual' intelligence and how it works.

I think there's a rather large part of the academic world of Computer Science that you're not at all aware of.

 

[kortin]

Because the Quarians are really just humans in the future. We must make sure all the events occur as they should.

 

[deadish]

Well, the brain isn't turing-complete. In order for a computer to be a computer, it must be turing-complete.

Wait? What?!

My brain can simulate a single tape turning machine, which is the definition of a turning-complete, just fine thank you very much.

No it can't.

LOL.

Dude, if we can't simulate a tuning machine in our heads, we won't be able to understand it much less build one.

PS: Oh yes, I don't recall anywhere stating that computers have to be turning-complete to be called computers.

I'm sorry, I mean that the human brain cannot simulate every possible turing machine, whereas a computer can. That may not be "turing-complete"-ness but I can't recall what it is.

I do know that that is a fundamental difference between human brains and computers

OK, let me put it this way. Almost computers we use now are turing complete. And yes I can simulate the operation of one in my head, else I won't be able to program them.

 

[Hagi]

Well, the brain isn't turing-complete. In order for a computer to be a computer, it must be turing-complete.

Wait? What?!

My brain can simulate a single tape turning machine, which is the definition of a turning-complete, just fine thank you very much.

No it can't.

LOL.

Dude, if we can't simulate a tuning machine in our heads, we won't be able to understand it much less build one.

PS: Oh yes, I don't recall anywhere stating that computers have to be turning-complete to be called computers.

I'm sorry, I mean that the human brain cannot simulate every possible turing machine, whereas a computer can. That may not be "turing-complete"-ness but I can't recall what it is.

I do know that that is a fundamental difference between human brains and computers

Do you even know what a Turing machine is?

The human brain is easily capable of simulating every single possible turing machine.

It will do so at stupidly slow speeds but it's not exactly rocket science.

All a Turing machine does is provide an output based upon an input and an internal state. A child could simulate it without any problems if it weren't for it being as boring as can be.

All you do is have one sheet of paper as your input, which is encoded in a certain way as known by you. Then you've got another sheet of paper which is your internal state, probably in binary but can be anything you want. And a last sheet of paper which is your output.

You read each line of input, perform the basic mathematical operations encoded upon your internal state and write the result onto your output-sheet. Congratulations, you're simulating an universal Turing machine. You could even do it without the paper if you've got a very good memory.

The only thing we can't do is simulate a Turing machine at 4 GHz. But at a speed of 0,1 Hz? Easy.

 

[ZippyDSMlee]

Its simple,cheaper work force.

 

[RobDaBank]

You could send probes to planets to check if they are habitable, allowing the AI to choose where to go to take, analyse and report rock samples, water pH and things such as that. This is something that we can do, but may take lifetimes to accomplish, meaning we are sacrificing people for the risk of finding nothing, who in turn would have to procreate and educate their children in order to carry out the task when they arrive at their destination.

AI systems with facial recognition software could be used to analyse the millions of hours of CCTV footage in order to track down criminals saving on valuable manpower.

Ai bomb disposal units, or even infiltration units, and even aircraft (like the film stealth)

AI refuse sorters to gather recyclable materials

There are alot of applications that could either save time, or replace the need for risking a human life and that is fundamentally the point.

 

[Daverson]

Humans are proof that strong intelligence can exist. If it can be printed on meat, it can be printed on circuitry.

There is a problem with strong AI, though. We have absolutely no idea what sort of alien moral code strong AI might create for itself. Even with careful design, advanced programs are prone to unforeseen glitches (just look at anything Bethesda have made over the past few years!). Given that such a construct would effectively be as close to a God as we're likely to see in our lifetimes, it's essential that we build it right, otherwise we're roll the dice on our own extinction.

It seems to me the best way would be to build your supercomputer around a human brain. By choosing the right candidate for this, we could effectively ensure that necessary values for future human prosperity are preserved.

 

[Rowan93]

"Deserve", "proper" and "superior" are all purely subjective words like should, which only have meaning when you're coming from a particular perspective. Being a human, I don't give a flying fuck about non-human-centric perspectives. And of course, "objective" perspectives are fucking bullshit. So what is your point supposed to be?

Oh look, it's the Lebowski Defense: "That's just, like, your opinion, man".

Please try again.

Well, it is just your opinion, because it's a subjective thing. You can't just dismiss that with a wave of a hand and a reference to a film. And you didn't answer my question!

 

[zehydra]

In order to hate you humanity, you must hate yourself

No, not really. For one thing, I don't think I'm human. For another, it's possible to hate a whole without hating every one of its components; for example, I hate Judaism, Christianity, and Islam, but I've had Jewish and Christian friends toward whom I felt only moderate annoyance.

"Deserve", "proper" and "superior" are all purely subjective words like should, which only have meaning when you're coming from a particular perspective. Being a human, I don't give a flying fuck about non-human-centric perspectives. And of course, "objective" perspectives are fucking bullshit. So what is your point supposed to be?

Oh look, it's the Lebowski Defense: "That's just, like, your opinion, man".

Please try again.

"No, not really. For one thing, I don't think I'm human."

uh oh. Well, I really have nothing left to say here.