AI Allowed to Create Their Own Culture

nomoreself writes "New Scientist reports that five European research institutes are building a virtual world with about 1000 virtual (AI) citizens, in order to observe the society these artificial agents create for themselves over the course of three years. From the article: "Each agent will be capable of various simple tasks, like moving around and building simple structures, but will also have the ability to communicate and cooperate with its cohabitants. Through simple interaction, the researchers hope to watch these characters create their very own society from scratch... [further], by pointing to objects and using randomly generated "words", characters should be able to conjure up their very own language and communicate with others inside their world." One of the researchers involved thinks the dwellers of this artificial world may even develop ritualistic practices."

Comments

Is buying 5 copies of

[twilight30]

... the Sims 2 really all that expensive? :)

Re:Is buying 5 copies of

[superpulpsicle]

It's a good alternative when you don't want your money with EA.

The test

[mabu]

The test of the realism of the sim's AI would likely involve how long it takes for one of the sims to seize power and exploit the rest.

Re:The test

[Anonymous Coward]

The test of the realism of the sim's AI would likely involve how long it takes for one of the sims to seize power and exploit the rest.

In other words, if the Sims don't fall for the same Leftist political fallacy as we have, it can't possibly mean that the latter has been refuted (again)... it means the sim's flawed.

That's the same as blaming human nature, rather than bad ideas, for the failures of his civilizations.

Agents?

[SDMX]

"Each agent will be capable of various simple tasks, like moving around and building simple structures...

...absorbing other agents, usurping gigantic networks, eliminating Keanu Reeves. You know, all the things a good AI should do!

That's All Folks!

[stuffduff]

Just be sure not to let it take over the world, OK?

Re:That's All Folks!

[Leroy_Brown242]

It's okay, we have everything under control . . . .

Learning about ourselves.

[Leroy_Brown242]

This has a lot of interesting possiblities.

One being that given a blank slate, what they tend to do on multiple runs. (Do they always end up the same place, or does chaos theory win out?)

Another, is that the AI could be programmed to have a pre-disposition, and see how they progress. (Homosexuality, self destruction, etc, etc)

And yet another could change their environments and see how they react. (Plague, overcrowding, etc, etc.)

Re:Learning about ourselves.

[ScrewMaster]

Another, is that the AI could be programmed to have a pre-disposition, and see how they progress. (Homosexuality, self destruction, etc, etc)

Sort of a "Gay-I", eh?

Err

[rbarreira]

Nevertheless, the researchers behind NEW-TIES hope to have seen some spectacular results by the time the project comes to an end in 2007. "It's incredibly ambitious, and it may be that, at the end of 3 years, we say we need at least another 30," Gilbert admits. "But it's worth a try."

Idiots... 30 years for a project like this? In 30 years we'll have much better methods of doing this, so any project started 3 years from now wouldn't be valid for so long. Those of you who have read Ray Kurzweil's essays probably know that there is a very good chance that we will pretty much understand how the human brain works by that time (like we understand the genome now).

Re:Err

[hab136]

Idiots... 30 years for a project like this? In 30 years we'll have much better methods of doing this, so any project started 3 years from now wouldn't be valid for so long. Those of you who have read Ray Kurzweil's essays probably know that there is a very good chance that we will pretty much understand how the human brain works by that time (like we understand the genome now).

We've mapped the genome. We don't understand it. BIG difference.

But if you meant we'll have a superficial understanding of it, and have big glossy headlines that miss the point, you're probably right.

Re:Err

[node 3]

Idiots... 30 years for a project like this? In 30 years we'll have much better methods of doing this, so any project started 3 years from now wouldn't be valid for so long.

You should put some thought into your words before saying, "Err. Idiots". They're not talking about running the same simulation on the same hardware for 30 years. What they want is for the project (not necessarily the simulation) to last 30 years (and beyond), and it's pretty idiotic to believe they'd learn all there is to learn in just 3 years.

30 years certainly seems like a long time, but on the subject of AI and alife, there's a *lot* yet to study. This 3 year simulation will answer just a handful of questions, and raise many more, leading to another simulation, and with the newer, faster hardware, and more capable software, they'll answer some of the new questions, raising yet more, and so on.

Those of you who have read Ray Kurzweil's essays probably know that there is a very good chance that we will pretty much understand how the human brain works by that time (like we understand the genome now).

Mapping the human genome (you've already noted this difference) is mere bookkeeping. It's raw data. Faster computers and newer methods sped up the project so that it was finished, it sometimes seems, before it even began. But that's just data, it's still going to take a *very* long time to really understand the data.

Take a simulation system far, far, far simpler than AI and alife--chess. Even after thousands(?) of years of study, and decades of computers aided study, we still have yet to fully explore that system--there's still work to be done, and will be, perhaps, forever. What makes you think AI and alife, which is far more complex than chess, will be so much easier that there won't be enough work to last even a mere 30 years?

As for Kurzweil's essay. He is making the case that we'll understand the brain in the same way a beginner at chess understands chess. We'll know pretty much what each part does, and how they work together in simple terms, but we won't have all the answers--there will *still* be work to be done.

Erm, Idiots indeed!

Re:Err

[rbarreira]

I didn't write the 30 years part without thinking at all. Read their sentence. They say they'll run the simulation for 3 years (read this post btw :)), and then they say that 30 years thing. From that sentence alone, one would interpret it as running the simulation for 30 years.

Now, if they're talking about running a big project for learning about society for 30 years, that sounds much better to me.

Re:Err

[node 3]

(read this post btw :))

Well, yeah, that's a joke though.

From that sentence alone, one would interpret it as running the simulation for 30 years.

I can see how that would be an initial impression (given the ambiguity of the statement), but it's one of those things where you think, "Oh, duh. I thought you meant the stupid, idiotic thing, not the reasonable and quite logical thing. haha", and not call them idiots.

But yeah, if they just want to run the same simulation on the same hardware for 30 years, well.. Err. Idiots.

Re:Err

[PakProtector]

Yes. What we need now is to map the 'Proteome,' or rather, to create a dictionary of the many, many different proteins that the human body uses -- what they do, what they're made of, how they're folded, how they interact.

That's a much, much more difficult thing than simply 'mapping' the genome.

Before you all jump on me :)

[rbarreira]

Sorry, I meant to say that we have mapped the genome, since that's what we have accomplished so far.

Call me skeptical, but...

[afabbro]

...wouldn't the culture they create be a reflection of the motivations they're given? 1,000 humans are so diverse that the culture could be anything. How do you get that level of diversity into AIs, using present technology?

Seriously, I imagine even describing programmatically the motivations and desires of 1,000 humans is impossible right now. You could simplify it (Sims, most CRPGs) but then you're at my question.

I have a feeling that if they are AIs who simply need to do X, Y, and Z to survive and survival is their priority, then there will be only a sterile culture of efficiency.

This isn't my area of expertise...just musing.

(Yes, I'm aware that you could therefore say that humans are result of the motivations our creators gave us...I'm not going into that.)

Re:Call me skeptical, but...

[Randseed]

You're right, of course. One way to handle this is to set up the virtual economy such that only a small number of virtual people have to deal directly with the basic necessities of survival. The ones that have free time are free to do whatever else they want. To make this work, have some kind of skill system, modelling how someone gets good at woodworking when they do that all the time, for example.

This leads to seeing whether the automata will diversify their labor or take equal shares. Will there be some automata who somehow become freeloaders? Nothing but breeders? Builders? Farmers?

This is an interesting experiment if it can be done robustly enough to actually show something.

Re:Call me skeptical, but...

[bersl2]

Seriously, I imagine even describing programmatically the motivations and desires of 1,000 humans is impossible right now. You could simplify it (Sims, most CRPGs) but then you're at my question.

I have a feeling that if they are AIs who simply need to do X, Y, and Z to survive and survival is their priority, then there will be only a sterile culture of efficiency.

People develop different tastes and (often inefficient) idiosyncracies. This usually creates diversity in a culture.

Now the question becomes, how do we simulate this?

Also, it appears that this will be a static population, and that the set of AIs will not change. This simplifies the task greatly, as being able to model the creation of a new personality is probably an entirely separate problem of equal or greater difficulty than the one they are taking on here.

(Note: didn't RTFA)

Re:Call me skeptical, but...

[Bastian]

I'd suggest that humans really aren't so diverse that the cultures you'd get from different populations can be wildly different - because they aren't.

Every human culture I know of is only superficially different from every other. Most of the differences I see that aren't of the "different clothing" type come from different circumstances rather than different motivations.

Re:Call me skeptical, but...

[yRabbit]

I read that the AIs will be able to sexually reproduce. :p

And creating a new personality would be extremely simple. Getting it "right" would be something else.
You could just randomize all the variables, or do a sort of combination of both parents' personality traits, with a little randomness (so down the line it doesn't all become completely averaged out).
At least if it works at all how I'd expect..

I'm pretty skeptical about this thing, really. All they'd be observing is what they've programmed, a bunch of actions for the AI to choose from and some random number generators.. unless they're using some AI programming techniques I've never heard of before.
The article doesn't say how it would be different than that.. how (if at all) the AI can adapt into doing its own things.
Um.
And it confuses me about the language thing. They might have to interpret it? "It's quite possible they will develop a language that we have to interpret,"" Um, they couldn't include a translator into it? Or is that what they meant?

"Perhaps most importantly, however, by pointing to objects and using randomly generated "words", characters should be able to conjure up their very own language and communicate with others inside their world."
That's not much of an accomplishment.. It would be easy to define a bunch of objects, or combined sets of objects, and assign a random word/string of letters to them.

Re:Call me skeptical, but...

[Anonymous Coward]

Are you talking about Star Wars Galaxies?

Welcome

[AtariAmarok]

"The test of the realism of the sim's AI would likely involve how long it takes for one of the sims to seize power and exploit the rest"

As long as he builds cool carriers and solar sailers, and enslaves programs for the such enlightened purposes as videogames, bring it on.

Culture is for Bacteria.

[Jeremiah Cornelius]

I'll say it again.

Culture is for bacteria.

There! That just feels better, to get off my chest. And by the way - there is no such thing as AI. Combining an infinite series of light-switches will never produce conciousness. Eliza is a game that can fool you, but it could never fool itself.

Re:Culture is for Bacteria.

[rbarreira]

Combining an infinite series of light-switches will never produce conciousness.

I would like to see your proof of this. Because, combining a finite series of neurons apparently produces conciousness in humans. Have you or anyone else proved that neurons are more powerful than "light switches"?

By the way, we humans fool ourselves all the time...

Re:Culture is for Bacteria.

[FidelCatsro]

Aha but we have yet no proof that the finite serious of neurones are what cause consciousness in humans.
The topic is still highly debated , as of yet it is probably the most likely reason though it is not tested in any real way , there are even theories that consciousness may be quantum in nature

Re:Culture is for Bacteria.

[kmahan]

I've got to disagree.

To make the typical politician all you need are a couple of Dimmer switches.

Re:Culture is for Bacteria.

[Profane MuthaFucka]

The difference between "you" and "itself" is just a reference frame. In fact, it's even less than that. Humans are *really* bad at introspection, so what that means from a practical sense is that your ability to think about something someone else did is scarecely different than your ability to think about something YOU did. They are both nearly complete mysteries to the person attempting to analyze the (internal or external) personality.

For that reason, an AI complex enough to fool others is complex enough to fool itself.

Re:Culture is for Bacteria.

[erlenic]

Are you sure you need more than one?

Re:Culture is for Bacteria.

[Jeremiah Cornelius]

Bad at introspection means "cannot determine valid input" for a system.

That's if the fundamentalist materialists are even correct.

Re:Culture is for Bacteria.

[rbarreira]

Aha but we have yet no proof that the finite serious of neurones are what cause consciousness in humans.

At least we're sure that they are the biggest part of it. This is one of those proofs which will only appear if/when we produce a conscious human-like being, I guess...

there are even theories that consciousness may be quantum in nature

That wouldn't be a hurdle at all. OK, we might need "quantum light switches" but what's the problem? ;) I know that wouldn't be according to the original post, but what he meant was obviously that we will never be able to produce artifical consciousness at all.

PS: If I (in my almost full ignorance of these matters) had to bet, I would say that quantum effects don't play a big part in the human brain. But I've already got some books in my queue about this :)

Re:Culture is for Bacteria.

[bluelip]

>>I would like to see your proof of this. Because, combining a finite series of neurons apparently produces conciousness in humans. Have you or anyone else proved that neurons are more powerful than "light switches"?

When was the last time a light switch turned itself off?

Re:Culture is for Bacteria.

[rbarreira]

I don't know :) But I don't doubt it has happened...

Re:Culture is for Bacteria.

[FidelCatsro]

We are as of yet not sure they are the biggest part of it , it is assumed and with dammed good grounds that they are . Though we are not yet even able to assert assuredness .
Don't get me wrong , I am a rather Hefty cynic . I do however always keep an open mind till conclusive proof is available either way.

If the quantum nature of consciousness could be established as factual then a new era of neuroscience and medical treatment could be reached, which would be truly amazing(of course any establishment of the nature of consciousness would be amazing) .
http://www.quantumconsciousness.org/
http://psyche.cs.monash.edu.au/v2/psyche-2-21-glob us.html
If your into that type of thing it can make interesting reading. Many more examples to be found on google

Re:Culture is for Bacteria.

[rbarreira]

Thanks for the links :)

I was also thinking about reading books by John Searle and another guy whose name I can't remember now. I'm really curious to see if there is really any good evidence pointing at quantum consciousness which is convincing enough for me to accept the possiblity (assuming I can understand the concepts, of course...).

Re:Culture is for Bacteria.

[locoluis]

Our Brains Don't Work Like Computers

It's like the difference between "Discrete" and "Continuous", or between "Digital" and "Analog".

And that's only one aspect of the difference. Besides, neurons are _alive_.

Re:Culture is for Bacteria.

[el_jake]

Neurons do not act like transistor gates. (AND/OR/NAND etc.)
Neurons has multiple states. Active / Semi active / Sleep. Operating many connection lines in between other neurons.

The task whould be to simulate single neurons (with code) and make theese simulated neurons interact on a weighted level. You whould need at least a billion of those sim-neurons just to be equal to a insects brain. Every sim neuron should be pre-coded with a specialized task, but allso be able to take another sim-neurons task is it should be necesary.

Makes me think of john Conway and his life algo's http://www.bitstorm.org/gameoflife/

Re:Culture is for Bacteria.

[FidelCatsro]

I am as of yet not sure if i am convinced , though the work on the concept has developed .
It was in fact one of the original concept of quantum physics.
Quantum physics is still a very young science
, it's one of the great frontiers of science and discoveries are rather abundant(and simple) , so i wouldn't hold your breath on a discovery this advanced for a good while .
Still though , The nature of consciousness has been debated since time in memorial .I guess it is one of those things that will demystify as are knowledge accumulates, I like to compare it to the earth and sun debates of yore . As once upon a time it was perfectly logical and near unquestionable to consider that the sun revolved around the earth .in fact even as of the beginning of last century it was logically assumed that the causes of most female illness was hysteria (which was a theory generally abandoned as of the discovery of shell shock in men after and during WW1)
I think what i am saying is , never make assumptions in science without strong evidence.
I am rambling

Re:Culture is for Bacteria.

[rbarreira]

Yep, the brain is surely a huge computational device. But that doesn't mean we can't reproduce it electronically (or if not totally electronically, with the help of some biological parts). The human brain computational capacity is estimated to be at around 10^16 - 10^17 operations per second. One day we'll be there :)

Re:Culture is for Bacteria.

[rbarreira]

That doesn't mean we can't reproduce it, or does it? Not all computers are digital. And of course, there's also the possibility of mixing electronics with biological components...

"Neurons are alive" - that's just a matter of semantics, whether you like it or not :)

Re:Culture is for Bacteria.

[Jeremiah Cornelius]

Searle - much of what he proposes can be summed up in his "Chinese Room" simulation.

He's pretty good at this. He's on to something - but if you need maths to prove that a pile of dictionaries don't posess consoiusness, let alone a spirit....

Re:Culture is for Bacteria.

[Monkelectric]

The problem with the quantum physics consciousness crowd is its really more *metaphysics* then physics.

These people want thought to be special, and as the mechanisms with which neurons work is largely understood (even if what the do in toto is now), they want to hang what they want to be special at the door of something that is not understood, and thus mystical and special.

Re:Culture is for Bacteria.

[rbarreira]

Thanks, that's exactly my opinion, but I don't want to go saying that before I really know why people think that. For all I know, maybe there is something pointing at quantum counsciousness. But I really wouldn't bet on it, I would be far more inclined for your theory... Frankly, I don't give a fly whether we're special or not. That doesn't change what we are, in my opinion! What do I care if I'm a deterministic being? I still have to make an effort to make things happen for me (or at least I feel I have to), and, above all, I feel conscious, that's all I need :)

PS: When quantum theory is explained, what will be the next explanation they'll try to give in order to feel special again? :)
PS2: As I've said in another post, even if we're quantum-based, that doesn't mean we can't be reproduced by computational means... all it would imply is that we would need a quantum computer for that purpose!

Re:Culture is for Bacteria.

[LWATCDR]

"I would like to see your proof of this. Because, combining a finite series of neurons apparently produces consciousness in humans. Have you or anyone else proved that neurons are more powerful than "light switches"?"

Well it would be impossible to prove what an infinite series of light-switches can ore can not do for obvious reasons. As to neurons being more powerful than light switches. Yes a neuron is more powerful than a light switch. I have yet to see even the ant level functioning out of an AI yet. I would say that the idea of an electronic consciousness is still totally unproven.

Re:Culture is for Bacteria.

[rbarreira]

I said finite, not infinite.

I would say that the idea of an electronic consciousness is still totally unproven.

I didn't say it was proven, but you can't say either that it's false just because it hasn't been proven yet. That's what I meant, and it's not what the original poster seems he was saying.

Re:Culture is for Bacteria.

[Profane MuthaFucka]

I understand introspection to not be the process of determining valid input for a system, but for determining internal states of a system. As humans go, we really don't have any clue what our own internal states are. In that way, our knowlege of our own consciousness is very similar to the knowlege of others' consciousness.

I don't think I'm a fundamentalist materialist, because if there is good evidence or good argument, I would have no choice but to change my mind. I'm just a plain old materialist.

Re:Culture is for Bacteria.

[rexstuff]

This comes down to the classic debate of dualism - wheter are minds are distinct from our bodies. I'm inclined to favour the side of the dualist - that consciousness can never arise out of a machine (or even of merely matter). To be sure, neurons are ridiculously more complex than light switches, logic gates or even more complex ICs. However, a neuron can be rather accurately modeled with perceptrons, or as a simple (analog) electronic circuit with weighted inputs. The neuron takes in some electric signal input(s) and produces some electric signal output(s). Now, to me it seems that regardless of however many of these simple (or even of more complex) tiny circuits we stack together, they will never achieve anything approaching awareness. On a different note, the article summary stated that one researcher suspected that this society may even develop ritualistic behaviour, and I am inclined to agree. If the project is performing random mutations on an agent's behaviour, then it is natural for an agent to develop a behaviour with excess process and continue to use and spread it without that agent or any other perceiving the excess process as being superfluous.

Re:Culture is for Bacteria.

[LWATCDR]

I was not agreeing with the parent. And I do not agree with you. I think it is a big unknown. I feel that it there maybe more to consciousness than we know yes maybe even a soul. But I will be honest that I do no know. I just know that we do not know. However I do know that a neuron is more complex than a light switch.

Re:Culture is for Bacteria.

[rbarreira]

Of course it is an unknown, it hasn't been proved either false nor right.

However I do know that a neuron is more complex than a light switch.

Quantitatively, sure. Qualitatively, we'll see in the future :)

Re:Culture is for Bacteria.

[VanillaCoke420]

Intelligence, I believe, could come in many different forms. It's not so that the only kind of intelligence is the human intelligence, using the exact same components we have in our brains. Or is there some sort of safeguard that protects the universe from the existence of conscious structures of matter unless they are humans?

Re:Culture is for Bacteria.

[BerntB]

Aha but we have yet no proof that the finite serious of neurones are what cause consciousness in humans.

People used to e.g. think that biochemical processes was of another kind than other chemical processes.

All animism theories tested has been wrong.

Can you present any reason why there is some new, metaphysical stuff needed to explain the next thing we don't understand yet in biology? (This decade it is consciousness.)

It seems just more wishful thinking by theists.

Ok, I asked a question I know the answer to -- there isn't. I'll make it easier.
Has there ever been any religious inspirations that has given a scientific insight into anything?

Re:Culture is for Bacteria.

[FidelCatsro]

Don't get me wrong , I'm not saying i believe any of it .
My point was more that there has been no proof as to what causes it .So until a method is found to gain adequate proof one way or the other i shall keep an open mind on the subject.
Personally i do favour the theory that it is caused by the arrangement of neurones , but I would like to see some clear evidence that proves this.
Also don't always equate quantum consciousness with mysticism or spirituality .It is generally just a theory relating to the mechanisms and implies no spiritual or mystical dimension to the subject

Re:Culture is for Bacteria.

[lw54]

Not when it's controlled by my wife.

Re:Culture is for Bacteria.

[jimmypw]

Maby you cant create conciousness but people can do an amazing job simulating it.

Re:Culture is for Bacteria.

[Alsee]

I just wish they had a damn brightness knob.

-

Re:Culture is for Bacteria.

[JustAnotherBob]

So in theory we would need quantum light switches to have consciousness? Yeah for spintronics.

Re:Culture is for Bacteria.

[JRIsidore]

If you go down to the neurons they work kinda digital, not analog. The input to a neuron from nerves is delivered in a frequency encoded signal, not amplitude. Higher input means more 'pings' per second. I think the fact that neurons are alive is not of great importance. It more or less just a tiny machine that maintains itself.
Also the basic function of a neuron is pretty simple and has already been simulated (neural networks). It receives signals from many connections (dentrites) and acuumulates them for a certain time. When this value then surpases a limit the neuron will fire itself. There are two kinds of such connections, those which increases the value and those which decrease it.

Re:Culture is for Bacteria.

[Anonymous Coward]

the brain is surely a huge computational device. But that doesn't mean we can't reproduce it electronically

Yes, but it doesn't mean that we *CAN* reproduce it electronically.

You do realize that you sound like a religious zealot, don't you? ("Your religion is wrong, because you have no proof! My religion is right, even though I have no proof.")

Re:Culture is for Bacteria.

[rbarreira]

Yes, but it doesn't mean that we *CAN* reproduce it electronically.

I didn't say I was sure of that, I just said I believe it. If you knew how to read I wouldn't need to tell you this.

Re:Culture is for Bacteria.

[CardiganKiller]

When was the last time you switched your lights off? I only know of one person who can voluntarily will himself into unconsciousness, and he was trained to do it to avoid interrogation (special forces). But take that a step further... can you permanently turn yourself off by thought alone? I've read about some Buddhist monks being able to do this. But last time I checked, most people had to use crude implements to turn themselves off. Sometimes I think some parts of our brains and ways of thinking are worse off than a lightswitch.

Re:Culture is for Bacteria.

[Jeremiah Cornelius]

Ahhh.

I must confess myself to be an Idealist - in the Platonic sense of the term. I suspect that human existance in the world more resembles a model like "The Matrix" - a shadowy projection from unseen sources - than it does a fascinating watchworks.

Re:Culture is for Bacteria.

[Jeremiah Cornelius]

Dolphins have intelligence and awareness.

So do rats.

If you don't believe or have awareness of some trancendant, creative divinity...

Re:Culture is for Bacteria.

[Jeremiah Cornelius]

BINGO!

Simulation is never duplication - only modelling.

An engineer's model may well be close enough "to get the job done." This I agree with. But I don't believe that real consiousness is located in my "mind" or "brain" - any more than I think that George W. Bush is located "in" my television.

We are beyond our own understanding. So cargo-cults like Christianity and Strong AI are constructed.

Re:Culture is for Bacteria.

[Retric]

As long as you're asking scientists for "proof" your trolling.

We have already simulated a simple brain which accurately responded as the original creature did by simulating the way each of their neurons interacted with each other and their environment thus demonstrating that our understating is probably correct but that does not provide proof.

Linking QM to consciousness is silly. If you want to exactly copy something you need to follow all the QM rules but you can make several PIV's which behave in vary similar fashion without replicating the QM states. QM is twice as old as I am and it's the basis for our understanding of vary small systems but the implication that you need to replicate a specific arrangement down to quantum levels in order to simulate consensus is silly. The brain operates at a vary large scale at extremely high temperatures and long time scales which tend to override most QM effects.

As far as QM is concerned there is nothing unique about a neuron or a liver cell they both hot and huge. Over a second a single cell goes though an insane number of quantum states but the ~100,000,000,000,000,000 cells in the human body over a lifetime of 2,049,840,000 seconds goes though so many specific states as to overwhelm any QM coherence that might be linked to consciousness. The human body has around 10, 000,000,000,000,000,000,000,000,000 atoms. Each of which can go though well over 10,000,000,000 * 10,000,000,000 *10,000,000,000 *10,000,000,000 states per electron per second. Now if you want to link how the receptor on a single neuron works then fine call in QM but for large-scale things QM is over kill. At best QM might act as a sort of random number generator which let's the mind pick things out of thin air but it's not going to be how the mind stores the memory's of your 10th birthday so you can think back on it when your 80.

PS: Give or take a zero or two on some of those numbers.

Re:Culture is for Bacteria.

[Orne]

Actually, neurons have more states than just "on" and "off". They exist in the analog world, so they can reach a continuous range of values.

So saying that, neurons are more like dimmer switches...

Re:Culture is for Bacteria.

[|/|/|||]

I think it's been a couple of hundred years since "spirits" have been the going explanation for anything at all. ;)

In the case of the dictionaries, since the information that they store isn't self modifying I would consider it to be "non-conscious."

Re:Culture is for Bacteria.

[plover]

I prefer my argument the other way: I don't eat yogurt because culture should be hanging on the wall in a museum, not in a spoon headed for my mouth.

Re:Culture is for Bacteria.

[spin2cool]

If you've ever had calculus, you'll realize that at some point, putting together enough discrete intervals creates something that is for all intents and purposes, continuous.

You must also remember that a computer simulation doesn't have to work exactly the same way as the human brain. If it can accurately reproduce human characteristics to such an extent that we can't tell the difference between it and a human, then how would we know? And why would we care?

Read some Kurzweil - I recommend 'The Age of Spiritual Machines'. It'll give you something to think about.

Re:Culture is for Bacteria.

[hplasm]

And I know a lot of people whose switches are dimmer than normal..

Re:Culture is for Bacteria.

[Anonymous Coward]

Yes, but it doesn't mean that we *CAN* reproduce it electronically.

You do realize that you sound like a religious zealot, don't you? ("Your religion is wrong, because you have no proof! My religion is right, even though I have no proof.")

One side has evidence to suggest that it may be true; the other side says "it's never gonna be possible; just 'cause".

We routinely modify brain chemistry, and change how the mind operates by changing it's operating environment. This demonstrates that human conciousness has a physical basis, and operates some form of mechanical process that we can manipulate, albeit crudely.

We can grow brain cultures of primitive organisms, and program them do simple tasks.

I can make you feel fear; anger; pain; or pleasure by changing your environmental settings.

If I build a similar system that does similar things, isn't the process by which it does it deserving of a similar label? In humans, we call that process "conciousness", or "intelligence".

The guy who used to work down the hall in my University created robots that "learned" to walk every time they were turned on. He made things that would seek light, and avoid dark shadows; much like a rudimentary insect brain.

There's a lot to suggest that we may well be able to reproduce brain function electronically. There's not much other than blind faith to suggest that it will forever remain an impossible task.

Face it, it's the naysayers that have religion on their side. We've seen no arguments to suggest that's impossible[1], and plenty to suggest that the problem should be as solvable as it appears to be.
--
AC

[1] Penrose takes pages and pages of irrelevant math problems to posit his theory that "quantum effects" might play a role in human conciousness; but he doesn't explain why the quantum probabilities don't average out into predictable effects when applied to neurons, like they do for all other forms of matter on the planet.

Re:Culture is for Bacteria.

[meringuoid]

I've never been convinced by the Chinese Room. Pile of books doesn't speak Chinese, fair enough. Man following rules in books doesn't speak Chinese, also fair enough. Room as a whole doesn't speak Chinese? That does not necessarily follow.

The counterargument, from the linked page, is

In response, Searle claims that if we simply imagine the person in the Chinese room to memorize the look-up table, we have produced a counter example to this reply.

Not really. Bearing in mind that the Chinese Room is assumed to be a Turing machine, this is no different from moving the lookup tables from the hard disk into RAM. It just contracts the system geographically.

Reading your link, it seems that Searle's actual argument has been that although the Chinese Room reasoning doesn't prove that it's impossible to write such a list of instructions and such a pile of dictionaries, it does prove that the Chinese Room would not actually be conscious. Perhaps, but does that matter? How would we know? What exactly is the difference between a Turing machine running ChineseRoom 1.0, capable of convincing a Chinaman it's conscious and speaks Chinese, and a Chinaman's brain in a vat inside a similar room?

Certainly, the Chinese Room Turing machine would say it was conscious, in perfect Chinese; it would compose poetry, it would discuss philosophy, it might discuss whether or not Turing machines can be conscious, it might even agree that they can't (has anyone told it what it is?) If it didn't show all the signs of the same inner being and spirit that humans do, it wouldn't convince our jury of Chinese that it's conscious and intelligent.

Conclusion: maybe it IS impossible to write a Chinese-room program, and maybe it wouldn't be conscious, but Searle hasn't proved it to be so, not by a long shot.

It's probably obvious by now that I side with Turing over Searle. If something acts like it's conscious, we may as well assume that it is conscious. Otherwise we wind up arguing solipsism: how do I even know you're conscious? You might be a zombie!

Re:Culture is for Bacteria.

[Anonymous Coward]

This comes down to the classic debate of dualism - wheter are minds are distinct from our bodies.

Given that I can change your entire personality by hitting your brain in the right spot, I feel that this debate is pretty much solved. I can often even change your mind just by altering your brain chemistry; even very simple mind-altering drugs like alchol are so effective in altering that contracts made while drunk aren't legally binding in many countries.

Arguing about "mind"/"body" duality makes no more sense to me than trying to argue for some mysical "software"/"hardware" duality. Only hardware really exists; software is just the name for the abstractions we use to mentally explain the effects of the hardware.

Now, to me it seems that regardless of however many of these simple (or even of more complex) tiny circuits we stack together, they will never achieve anything approaching awareness.

Define "awareness". It's harder than you think.

A thermostat senses what temperature it is, and in some sense, "knows" how to make the room the right temperature. It's vaguely "aware" of the room temperature. The hard question isn't so much about designing systems with "self-awareness", it's rather "How much self-awareness is enough to be impressed by?"

Humans aren't absolute in our "self-awareness". I didn't keep count of the number of times my heart beat while writing this, nor the number of eyeblinks I made. I don't know the exact body temperature of my fingers, nor how many newtons of force my fingers exerted on the keys to type this message. I didn't conciously decide to grow certain cells today, nor to let some die. I don't even realize the reasons why I get angry until afterwards; sometimes it's because I'm sick, tired, or hungry. I don't even know where a lot of my preconceptions and prejudices came from, nor do I conciously choose to become sleepy at night. I behave the way I'm built to behave, not the way I'd like to behave.

To me, all this suggests that my behviour is due to a very complex, dynamic system, and that suggests to me that a sufficiently complex, dynamic system should be able to mimic my behaviour.

How hard would such a system be to build? Depends upon how "smart" you want it to be. Walking robots, which teach themselves "how to walk" by flailing limbs until they manage movement in the direction desired were being built in the late 80s. They can also be programmed to "collect" things like pop cans, hide in dark shadows, and recharge themselves. They're probably nearly as smart as dumb insects. Making something approaching human level complexity will take time and money, but it just seems like the same problem, on a much, much bigger scale.

--
AC

Re:Culture is for Bacteria.

[Anonymous Coward]

Simulation is never duplication - only modelling.

An engineer's model may well be close enough "to get the job done." This I agree with. But I don't believe that real consiousness is located in my "mind" or "brain" - any more than I think that George W. Bush is located "in" my television.


Where is Micky Mouse "located", then? If he really exists outside of your TV/VCR, where exactly does he exist?

If I simulate your Mickey Mouse by making a copy of your DVD, which TV is showing the "real" Mickey Mouse cartoon, and which one is showing the "simulation"? If the DVDs produce identical programs, how can you state one program is "more real" than the other?

As for human beings, the term "real conciousness" is an absolute misnomer: humans are unaware and completely unconcious of many, if not most, of the vital aspects of our own existance. People aren't usually aware of how fast their heart is beating, what their blood pressure is, how much oxgen is in their bloodstream, or a whole host of other life critical information.

Neither the father nor the mother are aware of when a child is conceived; women sometimes don't even know they're pregnant a month or more after the fact. That's not very self-aware.

Our senses are rather dim compared to our pets. We've got lousy smell and hearing compared to dogs, or even cats. We can't sense magnetic north like a bird, and our distance vision is poor compared to them. Our night vision is lousy. We're comparatively unaware of the larger world around us.

There's really only a few things we are aware of; so using the term "real conciousness" for entities that are as unconcious as we are smacks of hubris to me. We're blind, deaf, and dumb to the underlying realities of our world, 99.99% of the time. We're just unaware of it. :-)
--
AC

Re:Culture is for Bacteria.

[John Hasler]

> Aha but we have yet no proof that the finite
> serious of neurones are what cause consciousness
> in humans.

In the absence of an objective, testable definition of consciousness it is impossible to prove anything about it (or even that it exists).

Re:Culture is for Bacteria.

[web_boyo_in_sac]

>>When was the last time a light switch turned itself off?

http://www.energy-solution.com/off-equip/occ_switc h.html/

They are using CS as display...

[Anonymous Coward]

I can imagine the language those AIs will develop

sharon_apple: fag camp3rs! OLOL!
Hal_9K: l33t sniper roxorz

Rituals

[ndansmith]

One of the researchers involved thinks the dwellers of this artificial world may even develop ritualistic practices.

What, like reading Slashdot?

Re:Rituals

[Anonymous Coward]

One of the researchers involved thinks the dwellers of this artificial world may even develop ritualistic practices.

Sounds like this experiment may have been inspired by a Futurama episode.

Bender: "Listen here, Malachi. Time for a religious donation. Hand over your wallet!"
Malachi: "But Lord, we are a poor and simple folk."
Bender: "Poor? Aw, crap."

Re:Rituals

[Tigsen]

OK, that almost made me blow Dorito chunks and Mountain Dew out my nose when I read it. ;)

Yep, sorry

[rbarreira]

Yeah, that's why I wrote this post, you probably didn't see it since it was just a minute before yours :)

Anyway, I don't understand why they'll need for a simulation like this. Will the virtual beings build a whole universe? Simulations aren't usually so slow. I found this page which probably has more interesting information, maybe I'll check it out later :)

Re:Yep, sorry

[hab136]

Anyway, I don't understand why they'll need for a simulation like this. Will the virtual beings build a whole universe? Simulations aren't usually so slow.

Because if it ran in a week, how would they eat for the next 3 years? No, they've probably slowed down the simulation so they can "study" it as it progresses.. read: eat Doritos and play video games for the next 3 years, occasionally looking at the simulation and write something up about it.

These guys are my heroes. :)

Re:Yep, sorry

[Anonymous Coward]

No, they've probably slowed down the simulation so they can "study" it as it progresses.. read: eat Doritos and play video games for the next 3 years, occasionally looking at the simulation and write something up about it.

And who said school doesn't prepare people for the world of work?

They comment that we already have human based sims

[bornyesterday]

They should just take one of the standard MMORPG games, take out all of the constructed buildings and everything else and then randomly spread all sorts of basic things (in differing forms so that they aren't immediately recognizable) and then sell it as the newest MMORPG where you simply have to survive - find food, water, shelter, develop societies, etc. If you die, you die, and you can start over, no problem. Use the randomly generated word thing for communicating too. We have enough geeks who are interesting in anthropology and related topics to put in thousands of hours into it. Hell, you employ me for 3 years to play it 8 or 12 hours a day and I'll build you Rome!

World on a Wire

[Scarabaeus]

Sounds pretty much like a certain Fassbinder movie from the seventies, Welt am Draht.

dude

[lambent]

These scientists, who sound pretty together, should be smart enough to realise that they don't HAVE to make their title into a easy to remember acronym:

"The project, known as New and Emergent World models Through Individual, Evolutionary and Social Learning - or NEW-TIES"

Geez, that's freaking almost unintelligble to anyone but the original scientists. Plus, you left out the "M" (models). You shouldn't be able to pick and choose which words get acronymed.

Just freaking call it "New Ties", already, and explain what it means later.

Re:dude

[chris_mahan]

I think the project would have gotten more attention by making it technical too:

New and Emergent World models Through Individual, Technical and Social Learning.

Where are the servers?

[FrontalLobe]

As long as the servers arent on the 13th floor...

Re:Where are the servers?

[Sancho]

The book was better. If you can track down a copy, I highly recommend it.

Re:Where are the servers?

[Anonymous Coward]

Is it as good as the stone tablet edition?

Re:Where are the servers?

[Anonymous Coward]

Almost, but hearing the oral version as the author originally conceived it was best of all for me.

Re:Where are the servers?

[silentbozo]

The book is Simulacron-3, by Daniel F. Galouye. I have the original pulp paperback (falling apart), and it is an excellent book if you're into old pulpy sci-fi. The premise (I haven't seen the 13th Floor, but I imagine from the trailers that it deviated quite a bit from the book) is that in a world where ad men and pollsters are everywhere, butting into peoples' lives and asking all sorts of questions on everything, these guys decide that there's a better way, and build a simulation of the world (complete with versions of themselves) to collect polling data.


SPOILERS FOLLOW: (although you can get much of this information from the back cover of the book)








Problem is, somebody doesn't want them to succeed - and as it turns out, that someone are the people managing THEIR world, which just happens to be a simulation created to solve the exact same problem (hence all the ad men, which don't exist in the "real" world.) Why do they not want the sim-in-a-sim to succeed? Why, it would intefere with the accuracy of data in the ad simulation, of course!

Re:Where are the servers?

[Sancho]

I have the same paperback and love it thoroughly. The movie takes the premise of simulations in simulations and builds a completely different story around it. The ad men/advertisements aren't even imagined in the movie.

incentives

[sfcat]

These types of independent agents can create interesting interactions. I assume that each agent has either common or individual goals/incentives. But the choice of these incentives is what determines what happens (at least it is a major determing factor), though it is unlikely that a person could predict the outcomes (if the system is complex enough). So how did they choose these goals/incentives? And are there penalities for over indulging (like eating too much to reproduce successfully, etc)? And what non-agent objects are in the system. People wouldn't do what they do unless they needed to manipulate their environment and compete for resources. Do these agents do the same?

Also, why have the agents develop a new language? Its just interesting to see what they do without having to guess what they are talking about. But it sounds like they are only mutating the nouns. But languages develop in different ways including different preposition structures (for instance old english had different forms of nouns instead of prepositions) different noun/verb/object orders, etc? Is this just being ignored?

And finally, human society is very complex. It is almost certainly a chaotic system meaning that any change in the initial variables makes predictions meaningless for the real world (the system to be predicted). So if they want to simulate human socities, shouldn't they make the agents mimic real people and their environment as closely has possible. It doesn't seem that this is what they are doing. If they are trying to predicte real societies, I think they are not close to this almost impossible goal.

When do they get to software?

[devphil]

I'm waiting for the simulations to program their own copy of "SimResearcher" and start running little AIs in virtual environments.

Re:When do they get to software?

[Anonymous Coward]

Apparently, we just did.

Re:When do they get to software?

[BlackCobra43]

The real mindbender would be if their version ran faster.

All I want to know...

[HotNeedleOfInquiry]

Is whether they invent God or discover metal first

Well...

[jolande]

Well, I for one welcome our newly simulated overloads.

REAL WORLD: AI

[grimover]

Maybe they should sell the rights to MTV, REAL WORLD: AI "the true story of 1000 strangers picked to live in a virtual world and have their lives logged. Find out what happens when sentients stop being polite, and start being real."

Re:REAL WORLD: AI

[Anonymous Coward]

Being real?

Does that mean when two of them hook up, they're being INTimate?

Re:All I want to know...

[geo.georgi]

I bet on God.

Work, eat, reproduce...

[toybuilder]

Every character in the simulated world will need to eat to survive, and will be able to learn from their environment through trial and error - learning, for example, how to cultivate edible plants with water and sunlight. In addition, characters will be able to reproduce by mating with members the opposite sex and their offspring will inherited a random collection of their parents "genetic" traits.

I wonder this world's oldest profession will be like the real world... ...farming.

Re:Work, eat, reproduce...

[Stoopid-Guy0]

Hunters and mid-wives came way before farmers.

Re:Work, eat, reproduce...

[vertinox]

Actually the worlds oldest profession was prostitution... And then theivery and then begging.

In that order.

Farming was somewhere 9th in the list after professional gravedigger but at least before tax collector which runs 21st before lawyer.

Re:Work, eat, reproduce...

[HotNeedleOfInquiry]

I've been told (by a German brewery tour guide) that brewing is the second-oldest profession. Always sounded good to me.

What happens when a virtual "ant" gets demented?

[Dark Coder]

Is there such a thing as a virtual agent that goes beserk? Like a cancer?

Will "it" take over the world, er, um, simulator?

Will we gain insight as to how this twisted ant maps to our reality?

Ego Massage

[Tyrsenus]

"One of the researchers involved thinks the dwellers of this artificial world may even develop ritualistic practices."

...including revering the researchers as their Gods.

Wait til they start writing grants...

[Dioscorea]

...or running fantasy simulations, or getting ritually excited about Stephen Wolfram's "A New Kind Of Science".

When the bots actually start empirically collecting data (instead of playing games), yea, truly then will they have surpassed their creators in technological sophistication.

They're calling the AI society ...

[slughead]

... skynet.

A I Control

[Jippy T Flounder]

i have no problem with articifial intelligence being used for all sorts of applications, but i think the second something so versatile is developed (like a virtual human), it may *just* be about time to start worrying about how much control these "applications" may be given.

haven't we all seen enough movies / read enough science fiction to take note of the -oh so subtle- warnings? let's be careful what we do with these things.

if they work.

Decade-hence

[jefu]

Yup.

But, it is also possible that this kind of experiment will help us to figure out how the human brain works. So that this experiment might be one of the requirements of such understanding.

timescale...

[cyb97]

Why observe over 3 years, if the whole society exists electronically; defining time and the speed of everything is up to it's creators?

I guess it's a nice way to make sure your grants and in turn workplace exists for another three years, too...

Lem-ing

[jefu]

Stanislaw Lem wrote a great story (actually a fictional book review) called "Non Serviam" (found in his collection of fictional reviews : "A Perfect Vacuum") in which precisely this is done and the scientist running the experiment eavesdrops on his artificial creations discussing the nature of God.

An excellent read (as are all the pieces in "A Perfect Vacuum").

Re:Lem-ing

[Troed]

... and on the same topic. It's more likely that we are living in a simulation than that we aren't.

http://www.simulation-argument.com

I hate the asymmetry in news reporting

[exp(pi*sqrt(163))]

When someone creates a new project it gets press releases and news coverage. But when the project fizzles out because it was completely lame and didn't go anywhere in the least bit interesting nobody bother reporting it. It's like news stories are written in a dialect of C++ with constructors but no destructors.

Well I think it's time for the garbage collector of the news world. Someone who goes through year or three old science and technology magazines looking for projects where the leaders say things like "this technique will replace everything else" or "I expect my system to develop self-awareness over the next 18 months" and brings a bit of closure to them. If the project has failed then the project leaders need to be asked "what do you have to say about your extravagant claims?", "how do you feel about the grant money you frittered away?" and "how do you respond to the poeple who claimed you were a crackpot at the beginning?".

I'll have to put this story in my queue for re-examination in 2006.

Re:I hate the asymmetry in news reporting

[nyrk]

Why? That would only serve to discourage experimentation. They do face reprocusions already, and that is in the form of not getting their grants renewed if they do a poor job. Many many bad ideas have to die out in order to get a few good ones.

Re:I hate the asymmetry in news reporting

[exp(pi*sqrt(163))]

Why? Because reporting gives a biased view of what is going on in academia. When you carry out an experiment there are at least two ways it can turn out. Both of these outcomes provide useful information (unless it's a dumb experiment). Reporting the experiments that come out one way and not the other deprives us of some of that information and gives a very skewed view. This is particularly acute in statistical results: 5% of studies will show significance at the 95% confidence level just through chance. If only those are published then we are basically seeing published noise even though, to the casual observer, we are seeing only statistically significant results. The same reasoning holds valid for other types of experiment too.

Re:I hate the asymmetry in news reporting

[nyrk]

There is the more fundamental problem, and that is that people like to read more about what is new, and coming up than they like to read about project failures. If you picked up a trade magazine or a newspaper, and it contained article after article of failure, how many of these would actually be read? It would need its own section, like an obituatary for research projects. Point is that since no one is forcing anyone to read the news, they print what sells.

Re:I hate the asymmetry in news reporting

[FleaPlus]

Well I think it's time for the garbage collector of the news world. Someone who goes through year or three old science and technology magazines looking for projects where the leaders say things like "this technique will replace everything else" or "I expect my system to develop self-awareness over the next 18 months" and brings a bit of closure to them.

To an extent, can't Wikipedia articles be used for that?

Re:I hate the asymmetry in news reporting

[MrScience]

Be sure to interview yourself and let us know if you got around to reporting on hyped-up projects. :)

Seriously, it's a good idea. You should do it.

Re:I hate the asymmetry in news reporting

[Anonymous Coward]

To an extent, can't Wikipedia articles be used for that?

Overstating the overconfidence of the moment is definitely one of Wikipedia's strongpoints.

Re:All I want to know...

[KrisW]

I think they'll be rocking out to Judas Priest way before religion comes into the picture. Oh, wait... you meant the other kind of metal didn't you?

Re:All I want to know...

[HotNeedleOfInquiry]

Wow, so I'm a troll.

The question really was serious.

Realistic Simulation?

[sirmikester]

Will these AI being be able to kill each other for survival? Will they be able to steal? Or are they only allowed to talk and build?

What a boring use for AI

[anthony_dipierro]

building a virtual world with about 1000 virtual (AI) citizens

Wait a second... Someone managed to build 1000 virtual (AI) citizens, and they're going to use these things to play games in a virtual world? Somehow I doubt it.

"characters should be able to conjure up their very own language and communicate with others inside their world"

It took billions of years for languages to develop through random processes. This aint gonna happen in 30 years unless the programmers intentionally program the language into the "virtual citizens".

This project sounds really cool, but when you realize the description begins with "pretend we've just discovered the key to artificial intelligence" you have to realize it's vaporware.

Re:All I want to know...

[vertinox]

Apparently, God reads slashdot... And has moderator points. ;)

But seriously, I think the social implications of AI creating religion without aid of human intervention might be quite ground shaking.

You don't need to worry

[petrus4]

Like an earlier comment said, it took millions of years for languages to appear among humans, and that doesn't even account for whether or not the process is genuinely random, or whether there was some kind of direction involved.

I personally believe that the term "artificial intelligence" (at least as it currently applies) is misleading...Outside of science fiction, there really isn't any such thing. Even in situations where somebody's been able to come up with a genetic algorithm that produced something interesting, the AI it produced was only able to operate within its' given environment; i.e., as an expert system. Take it out of the target environment however, and it would fall flat on its face just as surely as a desktop machine after coming to the end of a shell script. There's no adaptability there whatsoever.

Computers still don't have any real capacity for dealing with novelty...the best any GA I've ever heard of has been able to do is widen the category of knowledge that a given expert system can have, and make the boundaries of said category *look* more fuzzy and organic...but in reality, it's smoke and mirrors.

Occasionally I'll see applications which stimulate my interest...the creatures in Black and White were innovative, and the Sims 2 makes reasonably good use of numerical weighting, even if the pathfinding there still sucks to a degree.

Assuming it's possible for strong AI to exist at all, (and again, I have grave doubts) everything I've seen tells me it's still anywhere between 50-200 years away. Skynet or it's equivalent won't be showing up anytime soon.

Re:You don't need to worry

[StrawberryFrog]

Assuming it's possible for strong AI to exist at all, (and again, I have grave doubts)

The question is: is it possible in principle to construct, using only ordinary matter and energy, a thing that is intelligent in the same way that a human being is?

Logically, the answer can only be yes - consider exhibit A, a human being, which is made only of matter and energy.

What is your grave doubt?

Re:You don't need to worry

[StrawberryFrog]

and that doesn't even account for whether or not the process is genuinely random, or whether there was some kind of direction involved.

Of course there was direction involved: Natural selection.

Computers still don't have any real capacity for dealing with novelty...the best any GA I've ever heard of has been able to do is widen the category of knowledge that a given expert system can have, and make the boundaries of said category *look* more fuzzy and organic...but in reality, it's smoke and mirrors.

Have you studied the way that the human brain works much? It seems that that we do not have boundless capacity to deal with novelty, and that we use a toolbox of "smoke and mirrors" tricks which evolved for specific purposes rather than a general purpose mechanism.

Re:You don't need to worry

[petrus4]

>Have you studied the way that the human brain works
>much? It seems that that we do not have boundless
>capacity to deal with novelty

I am rather intimately acquainted with the lack of capacity for novelty present in at least some human brains; I have NVLD, a subset of autism similar in some respects (but not identical) to Asperger's Syndrome.

My point however is that human beings started off in caves (as one example) but eventually got to the point where we were capable of going to the Moon. The Moon represents a very different environment to what we are used to, and thus a considerable amount of adaptation and inventive thinking was required in order to work out how to send men there without killing them.

Human beings do, as you say, have a number of hardcoded instinctive behaviours which have developed over time; however, we have a few different characteristics which distinguish us from (at least most) other animals from what I have seen:-

1. We have the ability to perform complex manipulations of objects which are external to our own anatomy. Granted, that in itself is not entirely unique; it is the level of complexity which is. A cat might be able to figure out how to open a door; it would not however be able to invent or construct a gun.
2. We are able to exist in multiple environments. As an example of what I mean here, although dolphins are considered intelligent, they live in water only, and are unable to live out of the water. Human beings, although we primarily exist on land, have been able to learn to swim, and we can also exist underwater due to the previous point.
3. We have a limited degree of ability to anticipate/imagine conditions in environments alien to our own. This skill is part of what enabled us to go to the Moon, and if we go to Mars, it will be used again there.

If I saw an example of artificial intelligence which exhibited all three of the above characteristics, I might be inclined to consider it genuine. However, up until this point anyway, I have not.

Re:You don't need to worry

[Dan512]

...everything I've seen tells me it's still anywhere between 50-200 years away. Skynet or it's equivalent won't be showing up anytime soon.

Nice post skynet!

Simulating predestiny

[DynaSoar]

"Don't anthromorphize the machine."

Take that machine, run some software on it (a machine of its own, in the sense of an engine designed to produce a result) intended to mimic a small subset of human behaviors, then have the results interpreted by a bunch of people with a vested interest in having those results mimic the results of human societies.

Does anyone seriously think they could possibly see any other outcome than the one they intended to see, no matter how unrealistic that outcome is? More than anything else, this project would reiterate humans' tendency to see things within their own familiar context. But it wouldn't do it in a surprising way because it was designed to from the start.

Or perhaps it will be surprising. They will produce a supposed mimicry of a human culture out of 1000 individuals despite the fact that no collection of 1000 individuals not having, much less sharing, a set, stable language has ever occured. They will produce a completely familiar and supposedly "understandable" result from a patently ridiculous premise. They could very well set a new record for anthromorphization of the machine.

No matter how novel the emergent behaviors of the agents, they will be mapped to a predetermined explanation. If this models anything, it is the reification of predestiny. If the agents could have this insight, there's no doubt they'd invent ritual, out of frustration over a god that insists that everything they do is according to his design: they may as well engage in stereotypy, since that's how it'll be interpreted.

Re:Simulating predestiny

[Stu22]

Not every experiment in computer science has known results.

In the seventies Douglass Lenat programmed a computer with basic mathematical rules and heuristics, then he ran it without giving it input (other than random numbers and the code). When it started he made guesses at what it was going to figure out and not figure out, but in the end he was wrong about many things. Although it didn't do anything amazing, like discover theorms that had never been discovered, it was able to teach itself eighth grade math. It's called the Automated Mathematician. The Wikipedia article is very brief, I read about it when I was working on my thesis, I wish I could remember the book it was from.

Re:Simulating predestiny

[sgt101]

I think that this was debunked (by Lenat himself) in the article "Why AM and Eurisko appear to work", which is a mea-culpa that shows that Lenat is a proper scientist (we need more of them!) but that he was wrong. The critical insight I drew from the work is that insightful intelligence can arise from an interaction between a human and a machine, which is not to say that pure machine intelligence is impossible, but that Lenats experiments show insight arising and being codified from a collaboration.

Re:All I want to know...

[vettemph]

Maybe they invent Satan first and quickly realize they need to invent God because there is no Mr. Darkness with Light or something like that.
hmm. "Prince of Lightness" just doesn't have the same ooomph.

will they find religion

[josepha48]

it will be interesting to find out if they find religion as well

I would love news story followups, too.

[ToastyKen]

The thing I'm most curious about is: <a href="http://subjunctive.net/klog/2003/06/13/what<nobr>e<wbr></wbr></nobr> ver_happened_to.html">whatever happened to the three mysterious ships</a> reported during the war?

Re:What happens when a virtual "ant" gets demented

[Anonymous Coward]

Well, I've heard of a real-world equivalent that's currently serving as president of an entire country. Does that help?

they print what sells

[exp(pi*sqrt(163))]

Of course! But the asymmetry is annoying nonetheless.

Re:they print what sells

[nyrk]

Yep, it is much more annoying in politics though. but that would take us way off topic

Why it's five institutions

[Anonymous Coward]

The European Union's equivalent of the NSF really *really* likes cross-country grant proposals. If you want a big grant, you more or less *have* to partner with several institutions, and they have to be in different countries. Furthermore, it's extremely important to have some "underrepresented" countries in the mix, or the proposal often isn't considered. In this case, Hungary serves that purpose. Spain and Portugal are other good choices.

Now humans are playing God

[burdalane]

... just like somebody else played God when they decided to create humans.

Re:Now humans are playing God

[Anonymous Coward]

Creating humans is a lot of fun, haven't you tried it?
Oh yeah, this is /.

Re:Now humans are playing God

[burdalane]

Ewww, why would anyone do something as nasty as plopping humans out their rear end? That's why we have abstinence, birth control, abortions, and /.

Thanks for reminding me that human parents have already been playing God for a long time.

Re:They comment that we already have human based s

[spot35]

3 Years? Surely, it would only take the one day.

Re:All I want to know...

[Anonymous Coward]

Particularly as we are god to them.

Of course, they would be in for a shock as soon as they realised that we/god, in turn, are real.

(This is assuming they're more sophisticated than they really are...)

This was a James Hogan novel!

[SAPHRguru]

I knew my airport lounge reading would be useful for something...

This is exactly the premise of James Hogan's novel 'realtime Interrupt' (Baen Books
ISBN 0-671-57884-7)

A reasonably fun read -- intruiging if a little simplistic.

*PHEW*

[Jippy T Flounder]

*puts away the homemade explosives*
*removes the bandana*

seriously though, an A I doesn't have to be especially sophisticated to misunderstand something and produce a totally inappropriate response.
hell, i'm not especially sophisticated and *I* have a seemingly limitless supply of inappropriate responses. really, you can have some if you'd like. just kidding.

point is, on a smaller scale than the terminator, i can quite imagine a roomba becoming convinced that the new puppy should be "removed" - it DOES produce rather a lot of refuse... it may be extreme for the roomba, it may not be for the roomba's progeny.
'fink about it.

We are closer than you think...

[cr0sh]

Personally, I think we (as in mankind) are closer than many people believe to creating a true form of machine intelligence, if not true conscious and sentient artificial intelligence. The former is likely to come about first, and the latter, if it came about, would likely be quickly regarded as not being "artificial" anymore. In fact, we humans would probably quickly accept and adapt to it. One only has to look at how quickly we anthromorphosize just about everything that even looks like it could be "intelligent" to know this.

Why am I so certain that we are close? We have at our hands numerous pieces of technology and knowledge, some fairly mature, others fairly new, which, taken together, can be seen to represent a possible whole greater than the sum of the individual pieces. The key pieces (and there are many other important pieces, but not as key, such as recent advances in growing neural nets using real neurons, and growing neurons on a silicon substrate, among others) are:

• Network theory and applications
• Neural network modeling (and associated neuron modeling)
• Artificial Life (CAs and evolution modeling, such as being discussed in the article)
• Theory and Evidence on intelligence (and likely consciousness/sentience) arising from pattern matching
• The Genobyte CAM Brain Machine

Network theory and Neural network modeling has yielded, over the years, a ton of insight into how brains work (as well as a lot of data in how they likely don't work). Such modeling has taught us a lot in this area, but has yielded little to nothing on what we call consciousness or intelligence. These systems are forced to be small (relative to our brain - or even the brains of lesser creatures) due to the fact that currently the only practical method to simulate such systems is via software - simulating an analog, non-discrete, networked, and distributed system on a digital, discrete, serial computer. These small systems have yielded some very useful and practical technology (improved OCR and better credit fraud checking software, for instance), but none has created a true intelligence. The majority of the limits are imposed by the fact that you have to simulate a large networked system, where each node consumes many bytes of memory, along with many more bytes to describe the interconnects, and still more bytes for the software to simulate all of that - that ultimately you are left with a very hairy problem that a serial computer takes time to churn through. While a network with trillions of nodes could in theory be simulated in software, given enough hardware, it currently isn't something that can be done practically. Given these constraints, researchers have had to be content with studying smaller networks, hoping to interconnect them in some way to make a larger, more powerful whole. These efforts have yielded some results, but as a whole these "hand-tuned and taught" networks are not likely to be the building blocks we hope them to be. Ultimately, it will take a very large neural network to be built to make an intelligent system.

Some researchers have speculated that a simulation may limit these systems precisely because they are not "non-discrete", and that it is the ability for a brain to take in multiple inputs all at once and process them all at the same time to yield outputs which is responsible for what we call "intelligent behavior". Whether this is true or not, we don't know, simply because we haven't been able to build large scale non-discrete processing neural networks. On the other side of the coin, it has been speculated by researchers that the brain simulates serial computation from the inherent pattern matching ability of the neural network which makes up the brain. This is a very interesting possibility, one that argues, in a sense, that a neural network can become and simulate a UTM for serial computation (mathematics and logic). This isn't surprising - if we can simulate a neural network using a UTM, why can'

Too Much Emphasis on Culture

[Omegalomaniac]

The problem with research like this is it places so much emphasis on "culture" and "society" and how these will "emerge" from collections of individual units that it is easy to forget about how complex the individual units need to be.

Human culture is certainly amazing and could not have been created without many, many people. But it also can't be created by many, many ants. Sure, with ants you can get a lot of interesting behaviors, but you'll never get anything approaching human culture. I rather doubt the individuals in this simulation will be as complex as ants. No matter how much they wave the magic wand of "emergence" they can't get around the hard problem of creating something that is actually intelligent.

Valuable Insights for Cognitive Science

[David_Shultz]

Those of you taking the position that this project is a waste of time (or simply a researchers way to make some cash for a few years) need to step back and give yourselves a shake -or alternatively read a cognitive science textbook. In terms of gaining understanding of the workings of the mind this project has real potential. Traditionally, cognitive scientists usually focus on the individual agents using neuroscience and psychology, modelling their theories of cognition in computer simulations. However, it is occasionally pointed out that we should pay attention to anthropological issues. What if it is the case that certain behaviors and thought processes of an agent only make sense when interpreted within the sphere of that agents immediate society? This project has the obvious potential of exploring this possibility, and may yet reveal valuable insights.

The Genetics of Communication

[Baldrson]

Any simulation of culture that doesn't evolve the ability to communicate from an artificial genetic basis is bound to fail the most important test:

How is the genetic capacity for communication maintained in the presence of the evolutionary incentive to manipulate signals rather than communicate?

My preliminary research based on the demographic memetic prisoner dilemma (with climate variation) indicates it is very difficult to sustain communication if high migration rates are allowed.

Language doesn't develop by Random

[Anonymous Coward]

I've seen too many posts positing the development of language as a random event. C'mon guys, that's going a bit far.

Language arises from the need to communicate for mutual gain. Do herd herbivores "randomly" develop warning calls? Do dolphins "randomly" cooperate to corral and feast on schools of fish? Ridiculous!

Communication develops from need, and the complex communication exemplified in language arises from the complex needs of a budding race of cooperative tool-users.

These agents will be imbued with the necessary preconditions (i.e.) need to communicate with each other; let's see if they can develop complex needs enough to create a language!

AI isnt really AI

[PhantomRogue]

AI isnt really AI in this sense. The only true test of Intelligence is to think outside the box. Anytime you program something, it has constraints, its impossible to program something to be open ended. No matter how hard you try, the computer/program will always hit a wall. The closest some AI has become to being even remotely AI (in the true sense) could only be described as 'bugs' :) Bugs are the code acting in a manner in which it wasnt ment to, which is actually the closest thing to 'Intelligence' that I believe that we can comprehend. A true test of Intelligence is doing something that isnt expected and makes us question the intent/rules of the given system. You make 1,000 Agents, even with the most varied and complex systems of checks, weighted checks, and blind checks, you will still have Empirical Data in order to describe the results. And since there is no way to classify Intelligence, any results you may see are undoubtably skewed because the test was given restraints, no matter how 'verbose' the system was supposed to be, they will have been limited in one way or another. Also, its why Humans are so stupid. We are bounded by the physical form. ;)

Over three years?

[possible]

"Over three years"? Why not just run the simulation faster and have your answer overnight?

Re:Over three years?

[Anonymous Coward]

do you know ANYTHING about computers??

I want to vomit because of your stupidity. Please, shut up and never speak again until you have sufficiently gained an IQ over that of a wet sponge.

"Researchers Produce Ultimate Game"

[Paraplex]

...by combining the sims 2 with with Half Life 2:
http://www.newscientist.com/data/images/ns/9999/99 997674F1.JPG
at the request of a thousand horny teenagers wanting to see Dr Breen have it on with Alyx.

Of course God israel

[tepples]

Of course, they would be in for a shock as soon as they realised that we/god, in turn, are real.

And our God isn't real? Heck, He chose the people of Israel (code for "is real") to do His work in the second and first millennia BC.

Is this program named skynet?

[Anonymous Coward]

Worse yet it is a game that plays itself.

Your work is done here human.

Re:All I want to know...

[relaxrelax]

I'm wondering if their entire civilization will die out due to lack of ability to control the trolls?

I mean, even humans don't always remember you shouldn't feed the trolls. How can a basic AI know better? (-;

and my tax dollars paid for *that*???

[relaxrelax]

...and my tax dollars paid for *that*???








-Virtual citizen #791





(I'm sure that would be a proof of intelligence the programmers would censor. Funding is of the essence.)

We've been waiting for you Mr. Anderson

[DigitalReverend]

Seems familiar, almost like something I saw in a movie once.

OT

[tommck]

Hey... I saw a posting you had about the Chaintech 710 and using AC3 passthrough with mplayer in an old thread. I am trying to do the same thing but getting no sound. Can you tell me if you had to do anything specific to get it to work?

Tom

MODDERS: I had no other way of contacting the guy... cut me a break?

Re:OT

[hab136]

Hey... I saw a posting you had about the Chaintech 710 and using AC3 passthrough with mplayer in an old thread. I am trying to do the same thing but getting no sound. Can you tell me if you had to do anything specific to get it to work?

Tom

MODDERS: I had no other way of contacting the guy... cut me a break?

I've now put up some info in my slashdot journal. :)

Re:OT

[tommck]

Thank you. I've been playing with Ogle, can you post what your .oglerc looks like? I appreciate the help immensely. Both Xine and Ogle have been mute, but MythTV plays through the AC3 passthrough fine.

Thanks in advance,

Tom

Re:OT

[hab136]

Ah, forgot about that. Added, and updated my instructions.

I think I've enabled comments on the main discussion journal page if you want to move this conversation there. :)

Re:OT

[tommck]

2 unfortunate things:

1) can't comment on your journal stuff (maybe I'm doing something wrong)

2) with your audio section, I get errors and it exits without playing the movie. With the --debug flag added, I get this:

Note[ogle_audio]: Audio driver 'alsa' not found -- ????????????????
Note[ogle_audio]: Using audio driver 'oss'
*** format: 8
SNDCTL_DSP_GETCHANNELMASK: Invalid argument
ERROR[ogle_audio]: wanted sample resolution 16, got 0
ERROR[ogle_audio]: wanted byte order 1, got 0
ERROR[ogle_audio]: wanted encoding 2, got 0
Debug[ogle_audio]: channels 2, encoding 0, resolution 0
Debug[ogle_audio]: byte order 0, rate 48000
Debug[ogle_audio]: fragments 16, size 4096
Debug[ogle_audio]: 2 channels decoded: Left Right
Debug[ogle_ctrl]: child: 20791 killed (/usr/local/lib/ogle/ogle_audio)
Debug[ogle_ctrl]: pid: 20791 terminated on signal: 11 (/usr/local/lib/ogle/ogle_audio)
Debug[ogle_ctrl]: child: 20753 killed (/usr/local/lib/ogle/ogle_nav)
Debug[ogle_ctrl]: pid: 20753 terminated on signal: 2 (/usr/local/lib/ogle/ogle_nav)
Debug[ogle_ctrl]: child: 20752 killed (/usr/local/lib/ogle/ogle_cli)
Debug[ogle_ctrl]: pid: 20752 terminated on signal: 2 (/usr/local/lib/ogle/ogle_cli)
Debug[ogle_ctrl]: child: 20756 exited with 0 (/usr/local/lib/ogle/ogle_mpeg_vs)
Debug[ogle_ctrl]: pid: 20756 exited with status: 0 (/usr/local/lib/ogle/ogle_mpeg_vs)
Debug[ogle_vout]: set_sync_point()
Debug[ogle_vout]: set_sync_point()
Debug[ogle_ctrl]: child: 20755 exited with 0 (/usr/local/lib/ogle/ogle_vout)
Debug[ogle_ctrl]: pid: 20755 exited with status: 0 (/usr/local/lib/ogle/ogle_vout)
Debug[ogle_ctrl]: child: 20754 killed (/usr/local/lib/ogle/ogle_mpeg_ps)
Debug[ogle_ctrl]: pid: 20754 terminated on signal: 2 (/usr/local/lib/ogle/ogle_mpeg_ps)
Debug[ogle_ctrl]: removing shmid: 222953474
Debug[ogle_ctrl]: removing shmid: 222986243
Debug[ogle_ctrl]: removing shmid: 223019012
Debug[ogle_ctrl]: removing shmid: 223051781
Debug[ogle_ctrl]: removing shmid: 223084550
ctrl: ipc_rmid: Invalid argument
Debug[ogle_ctrl]: removing shmid: 223117319
Debug[ogle_ctrl]: removing shmid: 223150088
Note[ogle_ctrl]: exiting

Re:OT

[hab136]

Bizarre. Is your ogle compiled with ALSA support enabled?

$ ogle --check |grep alsa
Build: Linux 2.6.10-gentoo-r5 #1 Mon Jan 17 12:58:38 EST 2005 i686 Intel(R) Pentium(R) M processor 1600MHz Fri Jan 28 17:21:07 EST 2005 mmx Xv oss alsa

Re:OT

[tommck]

turns out my alsa is too old 0.5.X...
didn't know it needed a higher version.

That'd probably do it.

T

P.S. Comments on the journal, when enabled, only works on new posts to it

Re:OT

[tommck]

Actually, it turns out that ogle's configure script was wrong. I found a patch online to fix it (I have alsa 1.0.3 right now).

But now I get:

ERROR[ogle_audio]: Unable to update the IEC958 control: No such file or directory

so, it's still not working :(

Ever think about Aliens?

[paulpas]

Perhaps this could be some adapted to experiment with Alien cultures and develop ways of translating abstract languages and culture practices. Create a thriving artificial culture then make a Western society interact with it to see if they get along or destroy eachother (or other cultures for that matter). Maybe they could invent technologies that we would have never considered before. If it's made well enough, the possibilities are quite endless. Time to buy up some old computers to get a cluster going. Even allow it to be distributed so all of us can have a part of emerging spciety.

Just beware, we don't want Terminator coming true. ;)

May as well run this on a quantum computer so it's less predictable while you're at it.