Can People Please Just Stop With The AI Threads!: An Open Competition

reply to post by daskakik


To be honest, I thought it was quite clear at the start of the thread I'm aiming this towards people who seem to believe that in the next ten years we will have computer systems/AI constructs that have the ability to surpass human intelligence

reply to post by Death_Kron

In fact I'd even go as far to say that any AI based creation will still never be more intelligent than human beings, how could that be the case if it wasn't programmed to explore subjects beyond our comprehension?

It seems to me like you can't comprehend TRUE/STRONG AI. What happens when we can create machines that think for themselves, with a sense of self and identity. The Human brain is simply a complex biological machine, and we will one day design artificial brains with the capacity of a Human brain.

reply to post by Death_Kron


reread your original post
start at ETA
SPELL CHECK OR JUST LAZY
lets not nit pick
by the way i cant spell i use a dictionary

reply to post by ChaoticOrder


Maybe because it doesn't exist?

I think some people assume it's possible to become sentient using only logical thinking. There are things that cannot be made using algorithms and data.

Originally posted by philware
reply to post by Death_Kron

 

reread your original post
start at ETA
SPELL CHECK OR JUST LAZY
lets not nit pick

Well you are nit picking, I can't edit the OP now, so again if you also have nothing constructive to add then please refrain from posting in my thread.

Originally posted by 547000
I think some people assume it's possible to become sentient using only logical thinking. There are things that cannot be made using algorithms and data.

Exactly my point, some people however find this extremely difficult to understand...

reply to post by Death_Kron

And for those that disagree, I'm still looking for a sample of code that can think for itself, put it in any high level language you want, I'll even compile and test it myself. Prove to me that it's possible to create a truly self thinking set of code instructions ....

Well, do you really think humans have true free will, or is it simply an illusion? How are Humans able to be self-aware if you don't think it's possible for any system to spawn self-awareness? Don't tell me it's a divine human trait that can't be understood or reproduced in an artificial manner. We will understand sentience.

Originally posted by ChaoticOrder
reply to post by Death_Kron

 

And for those that disagree, I'm still looking for a sample of code that can think for itself, put it in any high level language you want, I'll even compile and test it myself. Prove to me that it's possible to create a truly self thinking set of code instructions ....

Well, do you really think humans have true free will, or is it simply an illusion? How are Humans able to be self-aware if you don't think it's possible for any system to spawn self-awareness? Don't tell me it's a divine human trait that can't be understood or reproduced in an artificial manner. We will understand sentience.

I'm not sure anyone can answer that question decisively. Self awareness on a physical, electrical level and self aware in terms of the human mind are two different things however...

reply to post by Death_Kron


i was only jokeing
but i do agree with most of what you say
but who knows what is in secret
they dont let on till some 50 years later on
when they find something better
technology evolves quicker than anything else
on this planet

reply to post by Death_Kron

Self awareness on a physical, electrical level and self aware in terms of the human mind are two different things however...

The human mind is a physical biological machine controlled by electrical impulses, those electrical impulses work through a neural networks and we can design our own Artificial Neural Networks.

secure.wikimedia.org...

An artificial neural network (ANN), usually called neural network (NN), is a mathematical model or computational model that is inspired by the structure and/or functional aspects of biological neural networks. A neural network consists of an interconnected group of artificial neurons, and it processes information using a connectionist approach to computation.

Self-learning systems using similar concepts power the most advanced AI of present day. By studying organic brains to see how they compute information, we can then recreate those systems and design hardware and software to create artificial brains capable of computing information in a very dynamic and adaptable way; hopefully one day so dynamic and adaptable we will begin to see signs of self-awareness.

You need to start seeing the Human brain as the machine it is.

Not that I think it will happen anytime soon but.

thesis:
1 programmer has knowledge 1
another programmer has knowledge 1 but it is different. (draw two circles a little offset and you have a bigger surface then one circle)
If we would compare these it would overlap big time but it is not the same.
(for this thesis lets assume they both work in the same language and other solvable technical stuff so it is no problem for them both to program a part.)
AI would have knowledge 1.1
1.1>1 AI>1human

Also worth considering, is self-learning programs that use principles of evolution to build complex AI systems. These complex programs can develop on their own without any help from a programmer along the way. So you say a program can only be as smart as the programmer, but it's a fact that some programs can learn to do things all on their own, things they were never programmed to do, and the final result can be something programmers barely understand.

The first video is the most important.
ps- when they talk about "brains" they are referring to artificial neural networks

In short they cant, sorry.

In long, how do you know it cannot happen? Just because it hasn't happened? What about learning algorithms?(as someone else said)
or what about self evolving AI networks? What about a synthetic brain? Scientists don't know how the brain works.... yet, but nothing says they wont know in the next 500 or 1000 years, or even the next 5 years.

You have got yourself into a logical fallacy with this one.

Please answer this question.

Since we cannot create anything "more intelligent" than us, yet we are here, that then means that whatever made us had to be more intelligent than us, so what made us?

If you answer : evolution, then sorry, evolution is not "more intelligent" than humans, or even alive for that matter so no go there.

If aliens, then what stops those same aliens from making an AI more intelligent than humans but less then themselves and basically give it to us?

If GOD then read above, replace alien with GOD.

edit:
P.S. Be careful what you say, someone just may take you up on your offer. (unless you like eating shoes that is)

Amazing. Op you do realize that the challenge you have laid out is impossible.

you, literally, have asked someone to program a "self aware" script without coding it at all. What kind of self indulgent thread is this, really?

I don't think we will ever be at the stage that we have "self aware" terminators or what have you.

But we are tantalizingly close to "robots" that could fool you. Robots that LEARN from their encounters with you and REMEMBER these things. A Robot that smiles at you, a robot that can *almost* carry on a human conversation.

And this is technology of today, in 10 years, unless we hit a wall, we will have much more power than a server farm today, in our laptop then.

What is your definition of AI? A true living machine? You, my friend, are the one who has seen too many movies.

Have you not played a game? For example, Brink.

Sure, the programmers created the "ai" code to make the NPCs work, but they didn't code YOU into it. These AI scripts, within the bounds of their coding, are able to react and function as a team, or enemy, and learn from your tactics.

This is just a half assed game with maybe an 8th of the cpu dedicated to that AI.

Computer models of self replicating, and "evolving" machines are already running right now.

We don't need to fully understand the human brain for AI as we're not trying to build a human brain, we are trying to CREATE a thinking machine. Obviously this has to start with a human hand, but that isn't to say we can't give the machine the technical knowledge that let us design it, so he can redesign it better.

See, the human brain is amazing, and we don't even use a lot of it day to day. A good portion of your brains processing power is spent in the background, making sure your organs work, that you breath, tracking your limbs for you, converting the stereoscopic image into 3d giving the illusion of depth, there is a ton of stuff going on in the background before you even get to the thinking stage.

AI doesn't need that. AI can dedicate 100% of it's cpu cycles to whatever task. Beyond that, we can link processors together to add to the power available, we can't do that with the human brain.

We, as humans, are pretty full of ourselves, and assume our way is the best way.

But once we give a machine out knowledge and ability, it will perfect the designs, make them more efficient in ways we can't even fathom.

10 years? We'll have some major breakthroughs and I fully expect to be able to have a natural flowing conversation with a machine. True sci-fi AI? Probably not in my life time.

But how arrogant of you do say it's impossible. I'm an IT guy myself and have been in this field for a long time. Plenty of stuff I have sitting around my desk is "impossible".

Using a cisco router as your example of AI not only shows your lack of knowledge in the AI field, but in your career field as well.

Originally posted by Death_Kron
All,

Nice and quick thread this, but something that has been annoying more than slightly the last few days. I'm sick and tired of seeing people talking about advancements in the computing industry as if we will be seeing sentient, artificial beings walking around the planet in the next decade.

I keep repeating the same point over and over; at this point in time a computer system can only be as smart as it's programmer, not smarter.

In fact I'd even go as far to say that any AI based creation will still never be more intelligent than human beings, how could that be the case if it wasn't programmed to explore subjects beyond our comprehension?

Careful. Our ancestors could not do that either. But at some point evolution resulted in the human brain having that ability. Now you either believe in the brains ability being a function of its "wiring" (like a computer) OR you believe in an external spiritual influence. I go for the former which means a machine should in theory be possible of achieving the same.

It seem's to me that too many people on here have watched a few too many Sci-Fi movies and have no real understanding of how the fundamentals of Electronics, Operating Systems and programming logic come together in the real world.

So as a challenge for those who believe, If someone can write some code that displays completely Autonomic behaviour, that isn't pseudo-random then I'll eat my shoe, any takers?

It can't be done.

ETA: Scientist's already admit they have a no where complete understand on how the human brain truly works, so how is anyone going to create a system smarter?.......

edit on 18/8/11 by Death_Kron because: (no reason given)

It is not a simple "algorithm" that results in the complexity of the human brain. Now, delving into philosophy and aspects of human nature as yet unresolved : It is more than likely that our behaviour is not truly "free will" but an illusion of free will due to the highly complex and MASSIVE number of interactions of learned behaviours, memories and experiences. From the moment we are born (even identical twins) our experiences are different and thus our expression of free will, being the sum of all we have experienced is different.

This means a machine CAN think if it is sufficiently large with enough experiences to draw on. No machine will "think" as soon as it is unwrapped. It will think if it has a lifetime of experiences. A certain level could be pre-loaded.

"Your" algorithm only provides the mechanism for linking and analysing experiences. You also need the experiences to think.

Just remember Einstein came up with his theory of relativity whilst remembering a train trip to Munich (I think it was Munich). There's the trigger for relativity!!!!!!!!!!!!!!!!!!!! you can't get much more "thought" than that can you.

reply to post by Death_Kron


Genetic algorithms have solved solutions that no human has in ways that even the programmers have no understanding of how and why the solution works. But other than that, I think that any form of artificial sentience is a long way off.

Kron,

First, S&F for you for creating a thought provoking thread. Whether people agree with you or not, they're engaged, and that, I think, is one of the pre-req's to make a good thread.

For the most part, I don't disagree with you on the premise of your post. Currently we cannot create a thinking machine that would be more intelligent than us. We CAN make machines that make decisions based on the same stimuli that we would, much more quickly, or "efficiently". Understood, that this would be based on code that tells the machine what to do. But, that's what the human machine does as well.

Although, when born, a child doesn't start with a totally blank slate, pretty much everything that they do to interact with the new word they're born into is based on trial and error. A baby cries and someone reacts to them. As a child grows, perhaps they touch a burning ember, again, a learned experience.

I certainly can envision someone coding sufficiently to do totally automonus things without human intervention. I can even envision code that would tell this mechanisim to do something else, if none of the preprogrammed variables can resolve the issue and note the outcome. Again, based on some kind of guidline by the coder, the machine can "decide" if the outcome is acceptable.

I know that your main point is that machines do nothing unless we tell them to, but I do foresee a day when code will be sufficient to need little, and eventually no human intervention.

One last thing, as I can see that there's a bit of confrontational banter going back and forth. I know a few programmers (I too am in IT), and the ones that I do know are "different", not good or bad different, just different. Their world can sometime be ONLY black or white. Something either is, or isn't, no in betweens. Be careful to not get stuck in the black in white my friend!! Sometimes when someone thinks and speaks in absolutes, they can sound condescending. I don't think you mean to and when we start going down that rat whole, the intelligence portion of the discussion goes south.

If the confrontation is part of what you're looking for, not a problem for me either. Just expect your licks and give as you take.

Thanks for the interesting thread!

Nice and quick reply this, but something that has been annoying me more than slightly the last few days. I'm sick and tired of seeing people complaining about threads as if they cannot simply move along and look at another thread. Especially when they create a thread saying "stop ALL the threads!" of a particular topic.

People will continue to be dumb no matter what you say, so theres always some people who think we'll suddenly see sentient AI, or people that think that gay marriage will suddenly lead to people marrying bridges and goats, or people who think that there is absolutely no other intelligent life in the universe.

reply to post by ChaoticOrder


Thanks for the videos, I'll watch them and get back to you, they seem interesting