One step closer to having robot overlords.

[Kazeofwinds]

http://www.damninteresting.com/?p=870

The computers....they're evolving, oh noez.

<insert random commentary on evolution debate here>

[Protogod]

God Created man and/or Man creates god
Man Creates Computers
Computers Wtfpwn Man
Computers create Biological life as a more efficient means of reproducing.
Biological life becomes man
Man destroys computers

[DarkMirror]

ZOMG!
Variables!
With a system that was only built for On-Off.
ZOMG!

[BlackDefiler]

It's only a matter of time, but since it's the next logical step in evolution it must come sooner or later.

[Leosam096]

Oo. Interesting progress.

[Icarus]

holy shit. That's fuckin awesome.

But I feel sad because sooner or later we're gonna have to destroy this kind of stuff

[Neo]

Darkmirror quit spamming.

nifty: "There was also evidence that the circuit was not relying solely on the transistors' absolute ON and OFF positions like a typical chip; it was capitalizing upon analogue shades of gray along with the digital black and white."

-Neo

[DarkMirror]

Thats what I've been saying. Read my posts.

[Neo]

Thats what I've been saying. Read my posts.

But you were spamming. There is a difference between a constructive post and simply repeating the same thing.

Here's a harsh hint: If no one got it the first time, then chances are you aren't as clever as you thought.

-Neo

[GenocideAlive]

Why do people feel the need to post a Terminator thread every time there's an advance in AI? Good god, it gets old. I've been seeing this shit since, what, 95? It's a computer that is trying to tell the difference between black and white, not a fucking robot death machine that's becoming aware of its surroundings.

If you really find this stuff interesting, try reading up on Isaac Asimov. Otherwise, for the love of god, shut up. It's not new, interesting, or funny.

[Borgorb]

What about the laws of robotics?

1. a robot cannot harm a human being
2. A robot must follow its programming unless it conflicts with the first law
3. a robot must put a humans life before its own existance

i think thats wat ther are anyways...

[Thedutchjelle]

This is very, very interesting. I'm really interested in the progress of AIs, so this is very good news.

Didn't know they were capable of doing this!

[Schwitzer]

Evolutionary design has been used in software for ages. Heck, I've personally programmed systems that have "evolved" the optimal solutions - generally to mathematical equations and the like where direct solving would be too infeasible.

As exciting as it seems from the outside, the honest truth is that these systems are still remarkably stupid in the sense that they are merely a logical process lacking any sentience. You have to give each system a very clear purpose and a defined way to weigh up the "fitness" of each candidate.

Don't fool yourself into thinking that there are computers out there that are consciously building better, faster, stronger versions of themselves.

[TheListener]

If the program hosting these boards were smart enough it would close this thread after calculating the smack down brought down by GA.

But then again GA not everyone is 43, I was 5 years old in '95 and therefore wouldn't be able to comprehend what specifically it's talking about.

Right, it's rather old, this sort of stuff, but I mean, it would be really cool seeing this sort of research being put to practical use.

[DragonPaladin]

Go read some of Isaac Asimov. One of the computers in his story was designed to answer any question. It also served to design its successor which became so complex that humans could not design it, just the original computer that created the one that came after it and so forth. The computer continued to evolving by building a better version of itself until it literally became God. He writes this sort of thing. It's pretty interesting.

[Thedutchjelle]

Evolutionary design has been used in software for ages. Heck, I've personally programmed systems that have "evolved" the optimal solutions - generally to mathematical equations and the like where direct solving would be too infeasible.

As exciting as it seems from the outside, the honest truth is that these systems are still remarkably stupid in the sense that they are merely a logical process lacking any sentience. You have to give each system a very clear purpose and a defined way to weigh up the "fitness" of each candidate.

Don't fool yourself into thinking that there are computers out there that are consciously building better, faster, stronger versions of themselves.

I know, but it's another small step towards a good AI