really the only thing after the source engine is...

[smilez]

As for the robot laws, what if a robot gives another robot an order? Or forces it to carry out an order with the threat of termination?

 

[Direwolf]

Thats covered. It cannot hurt a human whether ordered by another robot or a human (1 and 2 respectively). And 3 covers the threat of termination.

 

[Reaper978]

robots taking over the world, in my opinion, could happen. We have made so many mistakes, and this one could be fatal. What we have to do is not be stupid and form an ai that has no access to weapons, nor hacking tools to get access to them. Also, see 2001. Classic.

 

[jayte]

Yeah but how do robots interpret the asimov rules?
How would you? How can one know for sure that ones actions
may or may not harm someone, sometimes a action to prevent harm, triggers a chain reaction which results in the way you tried to avoid. "What a mind job."

 

[SiNizTer]

Sry to change the topic a bit...but the Half Life 3 engine could possibly be better than source...but then again I think they'll use a modified version of source anyways. But you never know, since HL3 probably wont b out for maybe (guessin) 5 years? Any opinions?

 

[Demon-Dragon]

Yes, those Asimov rules are a very good guidline.
But the truth of the matter is that not everyone will make robots that follow those guidelines.
Even now the U.S. military is developing robotic landmines using a program called "Swarm" which allows the robots to communicate to eachother and work in teams. These landmines will be able to remove themselves from the ground, relocate themselves, and then plant themselves into the ground again.
Although not hugely based around A.I. robots are already being developed for the purpose or intent to kill.
As long as there is war robots will be used to kill either other robots, or human beings.

Not the greatest thing I know, but it's pretty much true.
:cheers:
Cheers!

 

[PainLord]

Originally posted by marksmanHL2
My brother is doing a computer science course soon, and has just about learned C++ and he knows alot about AI. He was talking about it to me the other day. He was saying that one day people hope to make a game AI that Learns from the very beggining.

I think he said it would work by giveing a human model as many restrictions as possible such as haveing so much strengh in their legs and bone joint restrictions. Then you would set up a system of algorithmic sumthin loops. Or sumthin technogablaly and it would eventualy learn to walk its self.
It would also learn from its mistakes and others and basicly would eventualy form a realistic AI behavor. Once the process is complete you simply copy the code to make endless baddies.

cool eh (even if i explained it crap)

i'll ask him again as soon as he gets bk from paris!

That would be cool. You wouldnt influence their lives in any way. Just wait and see what happens. Maybe it could be the next sims. You could change the environment around them and they would adapt.

 

[SenorDingDong]

Will defintly have robots like the T-800, in Terminator one day. Only thing we must remeber is not too make them as smart or smarter then humans.