Quote: Original post by WeirdoFu
I don't really feel strongly about flocking being an emergent behavior either.

I agree that the actual cohesion of the flock is very weak emergent behaviour, because that's what your coding for with the three usual rules of flocking. What is usually considered emergent behaviour is how the flock handles flock level behaviours, such as turning the flock to a new direction and flowing cohesively around obstacles. These are 'global' behaviours not coded within the rule set, but which emerge simply because the chosen rules generate such behaviour under the correct environmental conditions.

I agree that the definition of emergent behaviour is ill-defined. Indeed, there have been some heated arguments over this very topic in Gamedev's past. Personally, I think it is much easier to identify something that isn't emergent behaviour (because we have another explanation for it) than something that is. In the end, it may just be that emergent behaviour is nothing exceptional and we just lack the computational language to explain those behaviours in terms of local rules. This is what Wolfram has tried to do at least... hence the suggestion that anyone interested in this topic read his book. It's not the only view of automata hat one should consider, but at least it's an easily accessible one.

Cheers,

Timkin

I said before that AI needs better definition! AI's like religion, there is an infinity of schools and views, and no certainty. At any rate, thank you all for the help. They have the book at my national library, I'll try to get it. I read the introduction on the web, but you dont learn much, except that Wolfram certainly have a good opinion of his work...

In the while I'll just continue my own experiences! See if I can make my AI to surprise me...

I haven't had time to read it all yet so can't comment on it, but here is a link to a draft version of a paper about emergence, called Emergence Explained: Getting epiphenomena to do real work.

Link: Emergence Explained