The Warring States of NPF

The Warring States of NPF (http://www.nuklearforums.com/index.php)
-   Dead threads (http://www.nuklearforums.com/forumdisplay.php?f=91)
-   -   Say hello to your new robot masters (http://www.nuklearforums.com/showthread.php?t=26843)

BitVyper 01-20-2008 04:55 AM

Say hello to your new robot masters
 
The level of communication happening here seems really cool: http://discovermagazine.com/2008/jan...arn-how-to-lie

TDK 01-20-2008 12:28 PM

Is it really a good thing that robots can lie now?

TopHatAssassin 01-20-2008 04:13 PM

Not necessarily. Then again, they can only fool their robot peers, as far as we know. Perhaps listening to a robot lie is like listening to a 5 year old do the same, it's pretty easy to figure out it's all false.

adamark 01-20-2008 04:49 PM

This is extremely dangerous. Just the fact that robots are being set up to "evolve" is putting them on the same footing with humans. Right now we are above them on the food change, but their digital evolution could change that in the future.

Telephalsion 01-20-2008 05:10 PM

Theoretically, if they evolve like humans and end up wiping us out, isn't it jsut as likely for them to create a new breed which will wipe them out, creating a spiral of destructive evolution?
Either way, this is really cool stuff, hopefully it won't lead to anything malign.

The Argent Lord 01-20-2008 05:11 PM

Quote:

Originally Posted by adamark
This is extremely dangerous. Just the fact that robots are being set up to "evolve" is putting them on the same footing with humans. Right now we are above them on the food change, but their digital evolution could change that in the future.

But they only evolve with human intervention. Without us, they stay at the same level they're at.

Mobeus 01-20-2008 05:24 PM

Quote:

Originally Posted by The Argent Lord
But they only evolve with human intervention. Without us, they stay at the same level they're at.

Unless we bring them to a point where they're capable of doing it themselves. Find program, streamline program, merge programs, streamline merged programs, repeat ad nauseum.

Khael! 01-20-2008 09:11 PM

I keep hearing GLaDOS trying to convince me she'll actually let me have that cake.

I know she won't. It's certainly real, but be damned if she's sharing it.

Will robots be able to lie to us? At some point probably. I don't see how they can get better at it than we are though. Even if they are as good as us, they'd have to attune to our social tendencies to really be any good at it. That would require a lot of empathy and abstract thinking to judge the efficacy of their own lies on us.

Demetrius 01-20-2008 09:20 PM

Quote:

Originally Posted by Mobeus
Unless we bring them to a point where they're capable of doing it themselves. Find program, streamline program, merge programs, streamline merged programs, repeat ad nauseum.

Funny thing is, that already exists. I think its an 80 line program in LISP.

Khael! 01-20-2008 09:23 PM

I once saw a simplistic version using randomness that accomplished that. It was written using C++. There wasn't any lying, but the fittest ones survived until the swarm of little hungry dots consisted of only the best ones.


All times are GMT -5. The time now is 12:46 AM.

Powered by: vBulletin Version 3.8.5
Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.