On self-emergent behaviour and Summer Glau

My intellectual interests these days: artificial intelligence, economy and alternatives, a little bit of ecology, genetics and evolution, company structure and system-business.

What all of these have in common? It seems quite a lot. All these subjects are heavily based on, if not essentially on self-organization and emergence of behaviour.

Economy work so well, because it is a system set up and all the variables and flow of value are kind of emergent. You can’t rule the economy, you can influence it heavily, but that’s what all the others (who have enough influence on the system) do.

Evolutionary theory is about self-emergence of various living things based on natural selection. There is no god in evolutionary theory, it’s a self-organized system based on principle of the survival of the fittest. What beautiful things emerge from it. Like dogs and us and Summer Glau.

Who plays Terminator Cameron. I like the story, the hidden fun seen by me, when no one else is laughing. I’ve been telling about me liking this show before. My subject at school was artificial intelligence and I’m still really interested in that subject. I saw a talk by Jiří Wiedermann at one conference last week and he was talking a lot by artificial intelligence and nanotechnology. And technological singularity.

He had a nice note about the risk of us building a killer robot race, that will try to wipe us out. And his answer is comparison about us and our children. (Almost) none of us raises his/her children to be murderers. But it sometimes happens.

Where is self-emergent behaviour in artificial intelligence? Multi-agent systems are basis for emergence observation. Self-organized maps create topological representations of things using self organization. Neural networks degrade gradually when parts are destroyed. If you take a normal computer program and delete 10 lines of code, it will usually break down quite badly. But there’s redundancy in artificial neurons and slow degradation happens…

But I think, people tend to be afraid of self-organization. We don’t know the outcome. We can see it in economical crisis, yet we propose solution, that have much in common with centralized planning. If that happens, it will break down in one step and completely. It’s not slowly degrading system.

People are afraid of self-organization. And they should not be. Because it forgives mistakes. And it usually works better with glitches (think about evolution without mutation).


Written by Juraj Bednár //