A frequently mentioned reason for the likelihood of human-equivalent AI being created within decades rather than longer is the fact that affordable computing power is approaching most estimates of human brain processing power.

100 billion neurons firing at 200 Hz — this is a basic neurological fact. Yes, there are many additional shades of complexity, including dendritic spines, neurotransmitter concentrations, and so on. Still, all of these put together seem to change the estimated computational requirements by no more than 2-3 orders of magnitude.

I can tell that I am speaking with an ideologue when they are unaware of the facts mentioned above, are informed of them, but that information then has no impact whatsoever on their subjective probability estimates of human-equivalent AI being created in the next few decades. Many people seem to act as if computing power has no influence whatsoever.

In contrast, Ray Kurzweil, Hans Moravec, and some other advocates of strong AI have seemingly acted as if computing power is everything — that when we have human-equivalent computing power, we’ll immediately have human-equivalent AI. That is wrong too.

It is easy to take the middle path. Particularly when the notion of human-equivalent computing power being available is combined with neural data from extremely high-resolution brain scans (a brute force argument for the eventual plausibility of human-equivalent AI if there ever was one), critics begin to sound incredulous when they do not revise their probability estimates for AI whatsoever.

One particular confused me that has been making the rounds for decades is the notion that some fundamental breakthrough in computing would be necessary to implement human-equivalent AI. A digital computer can simulate any possible analog signal, as long as it has the computing power — the inverse is not true. This is proven thousands or millions of times daily as old VHS and other magnetic tapes are converted into the digital medium.

If I had a computer faster than most expert estimates of human brain computing power and an extremely high resolution scan of the human brain, the burden of proof would be on the critics to say why I couldn’t create a human-equivalent AI immediately. The objections here tend to circulate around dualism, mysticism, biology-worship, quantum mumbo-jumbo, etc.
Yet, if we had sufficiently high-resolution scanners, we could just copy the brain’s design without understanding it.

SeaGates,

<<I was discussing about AI and since you do not care I don't think you should be discussing anything about it, no?>>

You as in all other SCIENTIST type is trying to create AI in an image of HUMAN BEING.  But, why??

There are MORE THAN one way to view and define INTELLIGENCE.

From Engineers' point of view, as long as something FUNCTION intelligently, we do not care whether it use the same method as human being.

<<AI system can't be selective in what it absorb;>>

Who say so?? Human being IGNORE things and event all the times.  That is ONE of the greatest sign of intelligence.  Knowing what is important.

<<I gave example when I said wheels, now tell me, is running on 2 legs faster than 2 wheels if you were tasked to move through every single city of Europe?>>

If you are SUPERMAN, it won't matter.

Dreamer

true AI will need to be able filter out noise from information (ignore certain things, if you will) even more than humans due to their higher sensitivity sensors.

true AI will also need to notice things outside of their design parameters and report back anomalies. how that is to be achieve, i have no idea, perhaps some form of neural network decision making system.

lin00b,

From a SCIENTIST POV, it is interesting to argue and define what TRUE AI is.

But, for an ENGINEER POV, FAKE AI or not close to 100% AI might be more useful.

For example, we may not be doing a TRUE 100% HUMAN EQUIVALENT robot.  But, replacement mechanical heart is keeping many people alive.

So, the QUESTION is

A) Is it MORE POWERFUL and FUNCTIONAL to create 100% TRUE AI??

Or

B) Create something that work along human beings that EXPAND and AMPLIFIES human processing and thinking power??

The same question is posed on the robotic side too.  Bionic arms and legs and so on...

SCIENTIST like to argue and debate until the hell freeze over.

Engineer tend to ask what give us BEST FUNCTION and PERFORMANCE with the technology that we have now.

I just want a robot to wash my floor.  And, I am still waiting....

Dreamer

+1..
We may be able to copy the brain design in the future yet we may not be able to understand the intricate design of the brain completely.

for low level job, yes, simple programming is adequate. what about a robot to take care of a baby? or to maintain a factory? explore uncharted area?

A frequently mentioned reason for the likelihood of human-equivalent AI being created within decades rather than longer is the fact that affordable computing power is approaching most estimates of human brain processing power.

100 billion neurons firing at 200 Hz — this is a basic neurological fact. Yes, there are many additional shades of complexity, including dendritic spines, neurotransmitter concentrations, and so on. Still, all of these put together seem to change the estimated computational requirements by no more than 2-3 orders of magnitude.

I can tell that I am speaking with an ideologue when they are unaware of the facts mentioned above, are informed of them, but that information then has no impact whatsoever on their subjective probability estimates of human-equivalent AI being created in the next few decades. Many people seem to act as if computing power has no influence whatsoever.

In contrast, Ray Kurzweil, Hans Moravec, and some other advocates of strong AI have seemingly acted as if computing power is everything — that when we have human-equivalent computing power, we’ll immediately have human-equivalent AI. That is wrong too.

It is easy to take the middle path. Particularly when the notion of human-equivalent computing power being available is combined with neural data from extremely high-resolution brain scans (a brute force argument for the eventual plausibility of human-equivalent AI if there ever was one), critics begin to sound incredulous when they do not revise their probability estimates for AI whatsoever.

If I had a computer faster than most expert estimates of human brain computing power and an extremely high resolution scan of the human brain, the burden of proof would be on the critics to say why I couldn’t create a human-equivalent AI immediately. The objections here tend to circulate around dualism, mysticism, biology-worship, quantum mumbo-jumbo, etc.
Yet, if we had sufficiently high-resolution scanners, we could just copy the brain’s design without understanding it.

SeaGates,

<<I was discussing about AI and since you do not care I don't think you should be discussing anything about it, no?>>

You as in all other SCIENTIST type is trying to create AI in an image of HUMAN BEING.  But, why??

There are MORE THAN one way to view and define INTELLIGENCE.

From Engineers' point of view, as long as something FUNCTION intelligently, we do not care whether it use the same method as human being.

<<AI system can't be selective in what it absorb;>>

Who say so?? Human being IGNORE things and event all the times.  That is ONE of the greatest sign of intelligence.  Knowing what is important.

<<I gave example when I said wheels, now tell me, is running on 2 legs faster than 2 wheels if you were tasked to move through every single city of Europe?>>

If you are SUPERMAN, it won't matter.

Dreamer

Guys. Dreamer 101 has a point.

The question here is simple:

Just because you're a human, what makes you think that you have a soul, compared to an A.I. built robot?

lin00b,

From a SCIENTIST POV, it is interesting to argue and define what TRUE AI is.

But, for an ENGINEER POV, FAKE AI or not close to 100% AI might be more useful.

For example, we may not be doing a TRUE 100% HUMAN EQUIVALENT robot.  But, replacement mechanical heart is keeping many people alive.

So, the QUESTION is

A) Is it MORE POWERFUL and FUNCTIONAL to create 100% TRUE AI??

Or

B) Create something that work along human beings that EXPAND and AMPLIFIES human processing and thinking power??

The same question is posed on the robotic side too.  Bionic arms and legs and so on...

SCIENTIST like to argue and debate until the hell freeze over.

Engineer tend to ask what give us BEST FUNCTION and PERFORMANCE with the technology that we have now.

I just want a robot to wash my floor.  And, I am still waiting....

Dreamer

Assuming we managed to do that, doesn't that mean we just cloned the human brain? Would you call that artificial intelligence or just an artificially created human brain?
IMHO I think the fundamental advantage of AI is the ability to learn and adapt to new task/environment.
Application wise this means the same floor washing robot can also be taught to mow the lawn or a variety of other tasks by some simple instructions/examples similar to how you'd perhaps teach a human child. Of course from the Engineering POV, Dreamer's right, AI hasn't reach such level of technology yet to be applicable.

---

Added on July 8, 2009, 3:55 pm
So are you taking the middle path?
A lot of computing power appears to be great but is it the essential ingredient in order to have intelligence?
As a programmer I do not see how by just having much more computing power is considered as intelligent since the computer only executes what it is programmed to do.
Some have suggested it's not the amount of teraflops of computing power but the amount of information that can be stored, retrieved and reconstructed that give rise to intelligence i.e. the memory prediction hypothesis.
If such is really the case, then the amount of computing power becomes irrelevant as to when we'll be able to develop human-equivalent AI. We may just need the right neural structure or perhaps the right language to write an AI program.
No, the burden of proof would be on the proponent of the idea to proof that it has the intelligence equivalent of a human, or it's just another faster computer or fancy machinery. That's what the Turing Test is for - to test for machine intelligence.

Why you asked? It's nature for human to explore every possibility. Why don't you ask the scientist who are pursuing the same thing too?

Yes you're right and you're just bringing back all my points that artificial intelligence is different from human intelligence(I already said something along that line in my 1st(2nd in thread) post).

Human being IGNORE things and event, but can an AI system do that, give enough confusing/clashing input to an AI system and it'll crash. Human simply takes a time out and slowly figure out everything. If I were to bet on whether an AI system or a human brain is able to come out with the theory of everything, I put my bet on the human brain. If you feed an AI system both law of General Relativity and Quantum Mechanics, it won't come up with anything like String Theory.

Scientist don't obey the rule of engineering that 'if it works, why fix it?' Science QUESTION everything. What is there to discuss here if everybody just say forget it since everything is working, why fix it?

And superman is the most irrelevant thing I've read in this thread so far. It's being totally ridiculous.
We're not discussing soul here, but human intelligence and comparison of AI to human intelligence.
Already answered you on the first quote.
Computing power is the shortcut to human like intelligence, the computer generates all the possible outcome of an event and choose the best. Very basic AI system still uses this pure brute force method.

SeaGates,

<< Yes you're right and you're just bringing back all my points that artificial intelligence is different from human intelligence(I already said something along that line in my 1st(2nd in thread) post).

Human being IGNORE things and event, but can an AI system do that, give enough confusing/clashing input to an AI system and it'll crash. >>

You are using HUMAN INTELLIGENCE view point to judge AI.  Is that the RIGHT WAY to do it??

Human has PROCESSING POWER running at low speed at much less than 1KHz.  Computer is running at above 1GHz aka 1 million times faster.  Human is good at ignore things and focus on key stuff.  Computer is good at look at and processing multiple things very very fast.  Those are just the basic differences.

So, if we want to use computer to function INTELLIGENTLY, we would not let computer work the same way as human.  Aka emulating human being.

Are we going about the WRONG WAY to begin with?? Aka, this is NOT the best way to approach computer based INTELLIGENCE.  We need to think outside the box of HUMAN INTELLIGENCE.

This has been going for 20+ years.  And, we have seen very little break through in this area.  We might be going the wrong way all this times.

I do not have the RIGHT answer.  Intuitively, I feel that computer based intelligence is SUBSTANTIALLY different from any kind of human based intelligence in term of processing.

Dreamer

thats why its ARTIFICIAL INTELLIGENCE. ie, to emulate human intelligence. we already have computer intelligence.

first huddle is to find some algorithm to assign importance to stuff so the AI will know what to ignore, what to focus, what to KIV, and what to forward to seek advise from others.

lin00b,

It is ARTIFICIAL INTELLIGENCE.  It is NOT ARTIFICIAL HUMAN INTELLIGENCE.  So, it does not have to emulate HUMAN INTELLIGENCE.  So, some people are studying INSECT INTELLIGENCE and try to emulate that.

Dreamer

No I think insect intelligence or swarm intelligence is a bunch of independent simple units that interacts with each other and the environment and producing a self-organizing and seemingly intelligent behaviour. It's like the main antagonists in the "Prey" by the late Micheal Crichton. For example some ant species can build monumental and very architecturally challenged structure without having the same kind of human intelligence.
Perhaps the simplest account of self-organizing behaviour is prey flocking, where simple-minded organism will flock together in the presence of predator and apparently confuse predator from isolating out and attacking individual prey.

But if they are pure trial and error, there would be many failures before they get one that's works and nature is seldom that inefficient. Like for example those 8 meters tall termite monoliths built by 1cm size termites.
The interesting thing is, even if they had the whole blueprint in their DNA, they still need to coordinate their building effort, or a loop sided structure can easily go tumbling down.

But if they are pure trial and error, there would be many failures before they get one that's works and nature is seldom that inefficient. Like for example those 8 meters tall termite monoliths built by 1cm size termites.
The interesting thing is, even if they had the whole blueprint in their DNA, they still need to coordinate their building effort, or a loop sided structure can easily go tumbling down.