Poll: will machines ever be more intelligent than people?

[the random]

i dont think so because to become more intelligent than people machines have to be able to learn and i dont think machines will ever be able to learn. Because to make something learn you have to reward good behavior and punish bad behavior (like training a dog) and you cant reward or punish an emotionless machine (and i dont think machines will never be able to have emotions)

/edit machines cant learn by themselves, they can only learn what we teach them

 

[Redlin5_v1legacy]

God I hope not. It's one of my irrational fears that the toaster will start judging me because of all the toast I make...

 

[porpoise hork]

Yes and I see it going one of two ways.

Skynet

or

Matrix

 

[Dejawesp]

Highly unlikely. Not until we move on to some sort of biological computer that can handle all the calculations without the immense heat generation.

 

[The .50 Caliber Cow]

What makes you say they aren't more intelligent already? The things people do and say these days...

[sub][sub]Moo! [http://www.youtube.com/watch?v=f9iIgQN5uZE][/sub][/sub]

 

[Lionsfan]

Thank god for the Turing test...unless they figure out a way to beat that

 

[mad825]

Very unlikely, the biggest problems for AIs are the hardware resources. It's just isn't sufficient enough so far especially with transistor technology.

 

[Lazier Than Thou]

Not anytime soon. I'm pretty sure World War 3 is going to put a stop to almost all technological advancement, personally.

Then again, I might also just be completely insane.

 

[Vern5]

I'm pretty sure that my five year old laptop is already smarter than I am.

The difference is that I'm more creative. Also, I have limbs.

 

[Zantos]

Sometime in the far future, maybe we'll be able to create a computer that is actually intelligent. Heuristic algorithms running through a quantum processor will not just mimic intelligence, but replicate it.

But as Michio Kaku so nicely put it; "Computers now are as smart as a cockroach. A retarded, stupid cockroach". I'm not thinking much is going to change for a while. Quantum computers, maybe in the next few decades, proper AI, not in our lifetimes.

 

[Suicidejim]

Depends on your perspective. Your average computer probably can't write or truly understand Shakespeare, and possibly never will. On the other hand, nor will they do things like getting various objects stuck up their anus or try to poke a crocodile with a stick because they're bored.

 

[omega 616]

i dont think so because to become more intelligent than people machines have to be able to learn and i dont think machines will ever be able to learn. Because to make something learn you have to reward good behavior and punish bad behavior (like training a dog) and you cant reward or punish an emotionless machine (and i dont think machines will never be able to have emotions)

http://en.wikipedia.org/wiki/Deep_Blue_(chess_computer)

That was in '97, I am sure we could do a lot more now.

Isn't learning just remembering? That is all computers do, if you want to save a file then you make the computer remember the file.

Learning has nothing to do with emotion, you sit in a class and write down information that you will latter be tested on. You don't get angry and string theory pops into your head.

 

[Dejawesp]

Isn't learning just remembering? That is all computers do, if you want to save a file then you make the computer remember the file.

Learning is when the computer makes a conclusion based on observations. Not just accepting someone else's conclusion.

 

[omega 616]

Isn't learning just remembering? That is all computers do, if you want to save a file then you make the computer remember the file.

Learning is when the computer makes a conclusion based on observations. Not just accepting someone else's conclusion.

No it's not, that is making conclusions based on observations. That is like progressive thinking, looking at a problem and working out how to over come it.

You don't go to school/college/uni, sit at a desk, the teacher says "here is a problem, work it out". You haven't been taught how to work out that problem, take maths for example:

a ship is coming into a harbor on an unusually high tide. The ship has to pass under the harbor bridge but the captain doesn't know if the ship will fit. he uses a theodolite to measure the angle at an unknown distance from the bridge and then re-measures the angle when he is 300 meters closer. The first angle measured is 2.3 degrees from sea level and the second angle is 3.3 degrees from sea level. If the ships height is 35 meters out of the water, will it fit under the bridge

I was never taught trigonometry so I can't work that out, I don't have the knowledge to do it. To make it simpler imagine a teaching saying whats 2+2=? without ever explaining what the number 2 is, what + means or what = means .... I might as well write this in a dead language and ask you to understand it, you haven't remembered the dead language so you can't use it.

 

[Rowan93]

Yes. Undoubtedly, unarguably yes.

Unless something operating in the brain is actively defying the laws of physics and operating on magic, there's literally no way that working out how the brain ticks can be impossible, which means we must be able to work out how it works (and, eventually, we will) and then we can improve on it, if only by doubling the processing power given to it. The brain is way below what physical limits allow, and even if it's the most processing power density that's possible, an AI can be made bigger than a brain.

 

[teqrevisited]

They'll be as intelligent as we make them. Built up over time they could potentially know everything that is possible to be known and use learned "thought" processes to determine the unknown. When a computer has the true ability to think and the power of uninhibited locomotion we should consider stocking up on EMP grenades.

That time is far, far off at the moment though.

 

[Scarim Coral]

While it is possible but I'm saying no on the ground that we not capable of doing so at least not in our life time. To even more a sentient machine we would first need alot of funding/ money to build one and the technology has got to be advance/ powerful for the senitent mind to stay alive. Neither ofthese are possible in this current time.

 

[Rowan93]

Very unlikely, the biggest problems for AIs are the hardware resources. It's just isn't sufficient enough so far especially with transistor technology.

it's difficult for us to multiply four-digit numbers together without paper. Neurons operate at a pitiful 100Hz. There might be a lot in there, and a lot of parallel processing, but there's no way the issue is hardware. It's software.