The AI Engine

I don't think evolution is a process of random mutation, but I don't think organisms change by design or purpose either -- I think circumstances change and species try to adapt to those changes, and some of those adaptations are encoded genetically.

But not all evolved traits are "by design". Human-level intelligence could easily have been a happy mistake or coincidence -- excessive redundancy in brain tissue perhaps protected us from fatal heat-induced brain damage but then got used in a new and different way -- thus enabling us to be the most successful (for now) species around. And then that kind of unforseen adaptation changes circumstances for everything else, and evolution continues...

But since it's probably not true that our kind of intelligence/self-awareness is a necessary result of evolution, there may be little hope in "evolving" an AI. Still, if one *did* evolve, it would probably be closer to our kind of intelligence than anything we might create from the ground up.

If there was a fossil record for computers, and we lived in a society that knew nothing of computers, we might be fooled into beleiving that the computers were rapidly evolving, that they started out as mere single circuits, roaming the office, then joined with others, grew plastic and metal coverings, and then roamed the world as such, "evolving" more with each passing year.

But we know better. We know there was a design, a maker. So why think that something a billion billion times more complex and advanced such as ourselves could have been some sort of accident?

A few unorganized points I'd like to make- we're heroes in our own minds. No virus or predator has done as much ecological damage to our planet as ourselves. In that respect, we are the most selfish and unevolved beings on Earth.

And why string things together in sequence when there might not be one. Without the continuation of single-celled beings, the pyramid of life on earth would collapse. Perhaps humans were the first thing temporally and through millions of years of science, developed the single-celled organisms that are necessary to make life on earth such that it is.

In a way, computers are evolving. Take the demise of Apple, for instance . . . <grin>. But seriously, the complexity of organic life is an argument *against* a creator, not in favor of one. There is also the fact that computers are built in an observable way and for an obvious purpose. The fact that human beings are destructive of their environment doesn't mean they are unevolved, it just means they aren't as well adapted as they'll need to be to survive long-term as a species. Elephants are incredibly destructive of their habitat, but they survive as a species (prior to human intervention, anyway) because they are so few and nomadic (destroyed areas have time to recuperate before the elephants need them again). Disagree with the idea that we're the most selfish and unevolved... just probably the most destructive (for the time being).

Of course there's nothing in evolution that refutes the possibility of external guidance or purpose -- there's just no evidence for it. And such a guide would clearly be far beyond our abilities -- hence it's a problematic solution for AI.

Explain this temporal idea. Are you saying maybe non-organic humans used science to create single-celled organisms, infused themselves into them, and then promptly forgot all about it? Would that have any meaning?

The complexity of organic matter is only an argument against a quite limited creator, but imagine a creator who is everywhere and within all of its creations at all times simultaneously. In such a way it could be accurately stated that each being is evolving itself, but likely this development is from a far more general and intelligent field that includes knowledge of every other form of life in its environment. Also, looking across time, evolution could be goal-oriented. There may be a perfected end result with creatures reaching toward that through time.

I dont at all believe in the notion that some white-bearded cosmic (and generally insecure) god thought.. "Hmm, dogs," then snapped his fingers and there were dogs. But this doesnt mean there wasnt a different kind of plan.

Strangely enough, science denies the possibility of evolution with the 2nd law of thermodynamics- the running down of the universe, the increase of entropy. Whatever caused the proverbial two cells to hook up together and form a new life form must have been grossly ignorant of this 2nd law.

You mentioned external guidance, but I think it's internal guidance. There's actually no evidence for randomness. No one has seen a non-cancerous random mutation. Sure genetic mixes appear that play up or down certain characteristics, but when has something entirely new been found in an offspring that wasnt there before?

The temporal idea- I'm just saying, what if humans were first, living in some sort of void, then said "Let's create a planet where we can live" and spent millenia developing the sort of self-sustaining single-celled organisms that are necessary for life on earth to exist. Then the single-celled organisms would be the more advanced (more recent, harder to design) life form. And of course, they would deny our existence.

IM going to be starting a Webcite soon that tells things as how I see them. I thing that you two will like it since it incoraptes some of both of your ideas. I will post it once I get done with it.

I don't think these pre-matter humans are like what I would call humans at all (and in any event, they're not US). Also I don't deny and can't refute that there might be an underlying or overt plan. But if there is, it's definitely not simple enough for us, limited creators, to grok. Since we are limited creators, we might face obstacles to creating an AI similar to our own Organic Intelligence that a hypothetical or real unlimited creator wouldn't/didn't.

THe second law of thermodynamics isn't violated, by the way, by any localized increase in complexity. As long as the total entropy in the universe increases, there can be any *localized* decrease in entropy you like. It happens all the time in the universe, though organic living systems are among the most dramatic examples.

Lastly, there are non-cancerous "random" (meaning apparently random) mutations all the time. Our schoolmate Mike Hammer is working on a project at U.Chicago right now studying an example of non-genetic DNA variation, trying to see if these variations are predictable (they seem to be based on the presence of certain proteins during critical phases of growth). Anyway...

I didn't think you believed in boundaries. How do you have this notion of everything being planned out without resorting to determinism? Our bots certainly lack anything approximating free will, and that's the biggest challenge for an AI, don't you think?

Holy smokes! I just caught up on the recent postings, these conversations are outta' control! This one'll have to be a little long I think...

It seems to me that us humans have a big problem saying "I don't know". People can go on and on about "god", "the beginning", "conscience" with great certainty and conviction in what their saying when in reality, we simply don't know. We've got impressions or hunches of things like "collective consciousness" or "inner guidance", but unless you were actually part of some master plan or in touch with some being that is, I don't believe we'll ever know the answers to those very complex questions that are brought up here. Our main problem being our incapacity to manipulate time. No matter what your IQ is, you can only explore the "present". Yet, we're constantly travelling through time so the present is constantly changing.

I also sometimes get the feeling that everything could actually be explained and is a lot simpler than it seems, but that's ridiculous! This notion that we're somehow the most successful beings or that intelligence only exists in humans is something I just can't bring myself to believe.

When you think that their are about 1 million insects for every human on this planet and that they've been living here at least since the dinosaur era...

Or how about the unlimited number of wave lengths that exist? We've got ears and eyes that can detect certain ranges of frequencies, but other animals could detect and use any of the infinite number of ranges, thus enabling them to communicate with the rest of some borg or somethin'.

There maybe no evidence of randomness, but when you combine that to infinite, it seems like there's some kind of infinite loop in the calculation required to foresee and predict something that we see now as random.

Anyways, if we ever do come up with all these answers... and step back and look at the sketch that describes the existence of everything, everywhere, "everytime" - we'll have this horrible feeling that the search will have ended. The party will be over. yuch!

Evolution is just a word invented by man to describe change that we'd like to consider good. Calling a difference between "now" and "then" "Evolution" just gives us the impression that the change in question is for the better. Grossly misused in human history to control the population.

On another note, how come no one has discussed the possibility of us humans "evolving" into the machines? We're already using the internet to connect to the borg and using hard disk space as an extension to our own memory banks. When are we going to get rid of these troublesome bodies? If I spend my whole life teaching a machine to be like me... and then I die off... but don't tell anyone and let the machine keep on taking care of business...

What if we all did this and then there were only machines left? Would they have taken over? Or would we have evolved?

resion I have not said anything about this is because this is an AI page. But agian I dont want to go in to a long speach about stuff. I will have my webcite done soon and then you can all go and see what i think for your serlfs.

Also Proof that humans are Evolving, Im taller than people before me in my family. My girlfriend has a lager chest than women before her. Growing and becomeing more complex is what its all about, and if you can sit there ans say that Humans are growing (in all forms) then your not looking at things I guess.

Crab & Rahz, I do beleive the greater plan is currently beyond our ken. Perhaps there are psychics or mystics who can observe it directly, and it would be an interesting AI that such people could come up with. But one thing we can be experts at is fooling ourselves- I wont deny that. And so I think that's our first goal in AI- to pass the Turing Test, to fool ourselves into believing it's real. We can fool ourselves into thinking an AI has free will until we know enough to know the difference between real and artificial free will.

On the second law, you're right Crab- it does make room for pockets of anti-entropy. My argument with it is that it states that in any closed system, the energy will run down and entropy will increase. But I dont think there's such a thing as a real closed system. And that's open to belief, as it cant be proven to not be the case. So you're right, I dont believe in boundaries, or determinism for that matter. But to get beyond such apparent paradoxes, you have to look to higher dimensions of reality where time comes closer to simultaneity.

I'd like to hear what Mike Hammer finds out. Keep me posted on what you hear.

Ender- I'm about to launch FreeSitesForAll.com. That might be a good idea for your site if you want low-cost hosting. Just a suggestion.

We're taller and bigger-busted than our predecessors for the simple fact that a lot more nutritous food is available to us now. It's not due to evolution since the cause isnt genetic.

Sir Rahz- You have some good ideas. I think that there is a very simple "engine" running the universe and all of reality. I think the things, people, viewpoints, times, and circumstances this engine can create are beyond count and measure. So discovering the engine I think would just open the door to exploration of areas we've barely dreamt of. And so I go on searching.

We assume that there is nothing beyond our senses, but clearly there is. What if we build a machine that can sense triple the range of frequencies than we can and crush them down into our range. Is it more intelligent than us? More aware? More conscious?

K Ender, I'll really try to stick to AI...

You know that "Random" feature in these BOTS that makes up for their spontaneous remarks and realness? Well eventually, it would be nice to key in some other factors that aren't so random.

The date, for instance, can affect us in more ways than are at first evident. If it's remembrance day, someone that remembers the war will react differently than someone who's never experienced it. If it's Christmas time, a youngster will be genuinely excited while it's parent will be rather depressed.

The weather, well the weather controls everything! Yet some of us aren't going outside that much any more... and others work the night shift in an underground parking lot and would prefer a quiet rainy night because everyone stays home.

I hate to compare our AI BOTs to a measly game, but surely you've all experienced the SIMs right? Or at least know what it is... Anywayz, they've set it up so you create you persona with certain degrees of playfulness, cleanliness, humor and a few other factors I can't remember now (can hardly remember anything since I found the SIMs nude patch). Actually, I haven't played in ages, but the idea was that your SIM would be in a pretty rough mood if it's the "Clean" type and no one's had time to do the dishes.

I didn't put this posting in the bug stomp forum because implementing this will not happen over night. This would take tremendous research and restructuring, I'm sure. But I think if we want to have true AI, it's decisions should be based on *knowledge*, *mood* and *something random* (until we prove that randomness doesn't really exist).

Then all of a sudden we've got "moody bots"... which is great! And even more real!

So, if we were to ask one of these moody entities why it was so silly, it would answer "I dunno, I was just built that way I guess"... Sort of how we explain our own moodiness and deal with it. If we asked it what it's f.... problem is today, it would also respond "I dunno, must be the weather or something". Or maybe it would have to be a strictly logical response and it would answer something like "Moods for AI beings are a subset of parameters and factors that are included within our programming to make us seem more real. My parameters are set to silly, yet today, nothing silly is happening. On the other hand, moods for humans are simply an unexplained royal pain in the ass... I can't say why you all went ahead and made BOTs moody!".

Timeline for us, Prof?

Seriously, couldn't it be something like this type of feature sparks "consciousness"? Or is it still just an imitation?

Mood we can do, and I plan on it. The little things that affect us in the outside world like weather and traffic, cleanliness and seasons can be simulated, but that's as far as it can go until we drop these Bots into some sort of Sim-World.

Attributing emotions to causes (and they will largely be social causes) wouldnt be too difficult to do. The very calculations that determined the emotion can be written out in terms of cause and effect.

That's it. We need a new forum. I suggest we call it "The moody AI" forum. (Just a suggestion though).