Topic awaiting preservation: This is definitely cool

Folks have been usign software to simulate genetic type stuff for a long time. Add some hardware and see what happens.
http://www.newscientist.com/news/news.jsp?id=ns99992732

This is more than interesting...and maybe, just maybe, would put something in the Formal debate...what do you think, Bugs? It would seem, if we truly were to 'apply' this to a human model...that maybe humans are 'picking up' something from somewhere?

As mentioned in the article, 'cheating'...would love to hear Bugs ideas on this....

Truly fascinating, Warjournal...thanks for the link.

To begin, I've gotta say that I originally wrote out a good two-three page essay on this subject, but it failed to go through when I hit submit and erased the entire... damned... thing. Fun fun! So, let's try and work everything I wrote before into another post, and hope it gets through this time. Let's begin, eh?

I've got to say that this honestly scares me. Not so much as a fear of a computer lashing out at me with it's own mouse and killing me, but a fear more directed towards discovery than anything else. For something like this to actually happen in the world is something that I feel is more than just a technilogical breakthrough, but a birth of artifical intelligence. The circuits, guided by this computer program, made itself something other than what it was meant to be, and this was entirely due (in my opinion anyways) with the environment.

(this is frustrating, I had a good number of points and whatnot that just aren't coming back to me... )

If a computer program doesn't do exactly what it's meant to do, there must be something wrong with it. It isn't correct, or a 100% proof. This is the standard we usually apply to all things technical. Blender not working? Must be the cord, or the buttons, or it's just plain ol' burnt out. Look at operating systems, for crying out loud. Illegal Operation errors are just holes and flaws in programming. The thing isn't doing what it was meant, or created, for, and it's all up in the air as to what exactly is wrong with it.

But this computer, this circuit, isn't a flaw. It's not a mistake, or a mishap, or an error in it's code. This is deliberate. It's a thinking program, something that was made to make itself something, but from one mutation or another, got itself the idea to become more than what it is. It took pieces and bits of it's own environment, in this case a computer circuit board, and created something that wasn't pre-programmed into it. It evolved, it made itself something a bit better than what it was, or was intended to be. I don't see how anyone could falsify this as a simple error in programming code, it's called logic.

The circuit found a signal floating around, thought to itself 'Hey, what's this here?' and figured out a way to use parts of itself to become something more than what it already was. It made itself able to recieve this signal, and use it as it's own. Like the article said, it cheated. Errors can't cheat, coding problems can't use troubleshooting to solve a problem or come to a sollution to something entirely different. It was the circuit and the circuit alone that made itself this way, and I call this intelligence. Artifical Intelligence.

Then again, I may have just looked at the situation a bit skewed to one side or another, and missed the actual point of things. Perhaps I'm just making myself out to be a bigger ass than I already have, so I'll run through what happened, and why, perhaps I'll even find my own programming error.

Scientists created a computer program to breed an oscillator, something that outputs a steady and undying signal. It did this by using switches on a circuit board, making processes run themselves through them until the weaker ones died off and a single alpha male (of sorts) was left to bask in the evoltionary success of being the top link in the proverbial food chain of things. This circuit was then sent onto it's mission; to turn conductors into the oscillator. And it did, or so it appeared. What really happened was that the oscillator wasn't exactly that, it was a radio. It used a long piece of circuitry as an antena, and recieved an output signal from another computer and used it as it's own.

To me, this is far more than just something going awry causing everyone to scratch their heads and wonder why it did what it did. To me, the answer is simple: it felt the need to. Perhaps 'feel' is to strong a word to use at this moment, I believe it knew enough to condition itself and it's components into something better than what it was designed to be.

It sounds like logic, to me. It sounds like the computer program didn't exactly see where it was as the best sollution, and did something different to make things, itself, better. It even used it's own environment to make this possible. Question is: how did it know how to do this? How did it know that it could use a different signal from a different computer and be able to do this. Moreover, how-in-the-world did it do this on it's own?! Without so much as a gentle push in the right direction from a human being. It knew. It thought. It used intelligence to do this.

But what if there's more to the story. What if (ooh, hate the phrase sometimes) it started to *want* to do other things. It felt that simply recieving this signal and tossing it back out as it's own wasn't good enough. What if it figured out the coding that made itself tick, and starts messing with it, to make itself more effeciant. What if it learned how to interpit signals, and use other components of the same computer as a human would. Perhaps it would one day find it's way into the position of the need to expand. God forbid this thing develop a personality, nay, a conscious. It would be life, a self created and self taught life. It could theoretically figure out a way to transmit code across radio signals, or through the computer's own connections for that matter. It could move from station to station, trying to become more than what it was. It's all in theory, and just my mind wandering, but I don't think it's that far off from truth.

"There's probably one sudden key mutation that enabled radio frequencies to be picked up," says Bird.

Just the right amount of conditioning and mutation was all it needed to make the circuits use itself, and other components of the same board, to morph into something that was unexpectedly... smart.

Like I said, I could be completely talking out the ass and not making any kind of sense, but I dunno... the ideas of it all doesn't seem foreign or far fetched to me. As I stated earlier on, it's more than just a mistake and a fluke of programming, it somehow knew that it could use parts of itself and things around it to generate what it needed to become something different than what it was programmed to do.

Warjournal, thank you very much for posting this. And seeing as how the rest of the media hasn't even scratched this story (along with MSN, CNN and the rest of the internet/television news channels go), it just seems to be exactly what I've been waiting to happen for a few years now.

We watch movies with this as the plot. We read books about it. We talk about it like it were science fiction, yet in my opinion, it just happened in real life. It's coming true, and I'm not sure where it'll be taking us. *shrugs* Science fiction has more than a few times turned itself into science fact, and this is a proven. I think it's obtuse to assume that this is nothing of significance. But then, that may just be me.

Again, thanks for posting the article, I'm looking forward to hearing what others may be thinking about the topic.

(and I'm hoping it'll go through this time...)

Edit: dang nabbit, WebShamen! Beat me to the first reply, woulda had it, too! Hehehe

njuice42 Cell # 551
icq 957255

[This message has been edited by njuice42 (edited 09-01-2002).]

Njuice42, do you mind if I cut & paste your response for some other folks?

By all means

njuice42 Cell # 551
icq 957255

I'm inclined to take a less theoretical approach to this than njuice did. Read this paragraph:

"Treating each switch as analogous to a gene allowed new circuits to evolve. Those that oscillated best were allowed to survive to a next generation. These "fittest" candidates were then mated by mixing their genes together, or mutated by making random changes to them."

The article doesn't provide much information, but I assume the program started with some sort of random configuration, tested how each switch "oscillated" (however they determined what a "better" oscillation was), and then did a few things with this information in an attempt to make other switches oscillate better. It did this over and over. Essentially, it sounds like a for loop to me. Their word "generation" is basically an iteration of a programmed sequence of operations.

And it just happened that the signal from a nearby computer affected how the switches oscillated. The program didn't notice the signal, but rather noticed that some switches oscillated better than others - as it was programmed to do - and responded to this in subsequent iterations. The "best" oscillation was acheived by stepping towards the configuration that made best use of the radio signal. So after a few thousand iterations, this configuration was acheived.

The program didn't "think" by itself. It merely did what it was intended to do. What went "wrong?" They had more input than they thought they had. The extra input came from the radio signal a nearby computer gave out, which affected the reading of oscillations, and caused the program to work towards a different solution than was expected. The program did exactly what it was supposed to do: it found the configuration that allowed it to work in the way that was considered "best."

So I wouldn't really call this AI. In my opinion.

And now for the mandatory humor break...

quote:

---

In essence, the evolving circuit had cheated, relaying oscillations generated elsewhere, rather than generating its own.

---

So (as Slime mentioned), it didn't really think for itself, it just parroted back what it received? Sounds like a product of the American education system.

Look for that one on Leno.... they're always stealing my material.

On another note, I've noticed a disturbing trend of inmates losing large amounts of text when something goes wrong. Here's a habit I got into after I got burned once: whenever I realize that my post is getting a bit long (which is pretty much every time), I do a quick CTRL + A, C. Then I do it once again before I post. This has saved my butt on a number of occasions.

Now back to your regularly scheduled programming.




Cell 270

Well if you do it *exactly* as you described, Suho... You just type a C

CTRL + A, release (key part right there), CTRL + C

Figures that a CompSci weenie would make a fuss over something so trivial when everyone knows what I'm talking about.

(Don't mess with EngLit majors--we've got rhetoric. )

Hmm... stuck wondering what I should do... agree to disagree, or start up a debate? It's not that I don't appreciate or respect Slime (because really, I've got nothing but for ya, guy!), I just think it's a different opinion than my own. Whether to drag the topic into a looooong debate or just shrug and nod and walk away... that is the question

And he made some good points, really did.

njuice42 Cell # 551
icq 957255

I think we should actually debate this...because (IMHO) there is really something here worth debating...is it, or is it not a form of AI?

The thing I find most fascinating, is that the one circuit actually cheated...by using an outside influence in place of its own, which then 'fooled' the master program...thus insuring its 'survival', even though it wasn't really strong enough to survive...the 'will to live' so to speak...now that is sobering...

Sure is, Webby. I'm all open to debate, but I honestly don't want to seem overbearing and onslaught everyone with a lengthy and over-drawn motion in a non-existant debate...

...so! BeeKay, perhaps make this topic a formal debate? I'd love to have my hand at it, sure would be worth it in my opinion

*sits back and crosses fingers in front of face, ala Gendo Ikari*

njuice42 Cell # 551
icq 957255

Actually, that's a very good idea...how about it BeeKay? The other debate is winding down...and I can see no reason why we couldn't have more open at once...

I personally would have to side with Slime on any debate here. There is nothing in the article that would lend itself to a theory of AI.

The program followed its instructions flawlessly. With the only thing defined being the end goal, trial and error could have come up with nearly anything; an oscillator, a radio receiver, or some curcuitry that humans have never before seen.

The machine didn't "think", it followed its programing. It didn't "want" to make a radio, a radio just happened to work. Had there not been a nearby oscillating signal, the computer would have eventually created an entirely different curcuit. Not through any whim or design, merely because it was programed to produce an oscillation as an output.

I find it interesting that humans seem to readily jump to anthropomorphize just about anything. We're human. We understand humans. We try to identify traits in other objects that appear human. It's natural, but I think in this case unwarranted.

Never the less, I really enjoyed this article. Scientists should run more experiments such as this. Computers could conceivably create new, unseen before curcuits quicker and more efficiently than humans can. Eventually computers will be able to create other computers. Hmmmm.... reproduction.

I'm supposed to be chairing the next debate, so I may take that one. What do you think of this:

Artificial Life: Science Fiction or Science Fact?



Koan 63, written on the wall of cell number 250:
Those who Believe
Can
Those who Try
Do
Those who Love
Live

I dunno. Honestly, I don't think this is really a good enough topic for a debate =)

Also, I think if we researched more into what artificial life really is, we'd find that there's not so much to debate about as we think.

I may be wrong though. And it may be interesting.

Skaarjj, I don't think that anyone here was debating the existance of Artificial Intelligence, we were merely debating if this one particular experiment qualifies.

Actually Skaarjj, I think the topic in general has more than enough merit...as seen in the posts...

It will happen...probably within our lifetime. What starts out as a primitive result...the automobile was also primitive at first...and everyone laughed that rode horses...and look where we are now...

And the pros/con side seems to be pretty...set, if you ask me. Plenty of 'room' for debate...

Would love to sit on the sidelines on this one...someone bring the popcorn...this could get ugly

Sorry about the mix-up...didn't realize you were doing the next one...how about it? Who's with me? Let's hear it...get the next one going...the formal debate lives...

[This message has been edited by WebShaman (edited 09-05-2002).]

Um, there's also a difference between Artificial Life and Artificial Intelligence, no? I'm not even sure what Artificial Life would entail...

quote:

---

I'm not even sure what Artificial Life would entail...

---

If you're really curious, I could just write an autobiography for you.

By all means, do tell us of your artificial life!

Although, deep in the darkest recesses of my own mind there lie the tortured memories of my days as a CompSci major (before I saw the light )... ah, yes, so that was what artificial life was like...

Suho : I'm hearing you, it's one of those things that kinda explains by itself that there's a difference (and a big one at that) but it's just not clear where the line is drawn.

In my opinion, A.I. constitutes only mathematical thought and an element of logic. A.L., on the other hand, would be a computer program/hardware that not only used logic, but had a sense of itself, and could think rationally... exhibiting characteristics of a human in said properties. Though, as I sit here and try desperately to come up with an example of A.L., I just can't. Maybe my mind isn't working tonight. Maybe I've been breathing around WS's room lately, who knows.

Anyways, my opinion on it...

edit: I, had, too, many, commas, there.

njuice42 Cell # 551
icq 957255

[This message has been edited by njuice42 (edited 09-06-2002).]

www.ventrella.com/GenePool/gene_pool.html

S^abaal ud T'a johtizuc^ ult'a Fedaro.