intrust

We understand how facts and figures, as computed by programming code make up computer systems, but the incorporation of norms is still rather foreign to us. Though we should not confuse computer simulation with what goes on in our everyday lives, let it be pointed out that the strategies employed in Axelrod’s studies are not essentially different from strategies used by system architects in real world environments. Google, for example, uses social proof, dominance, and reputation mechanisms to sort out the ranking order of results for your queries. They also use dominance and membership algorithms to promote their Gmail webmail offering.

Evolutionary approaches to program and system design are a reality of today’s information technology. In loosely-coupled dW3 landscapes agents are given the ability to discover interacting partners amongst a variety of alternatives, they are authorized to negotiate, enter into contract, and trade.

Axelrod’s games, or rather their more complex modern day equivalents are very much a part of our computer-human verosphere. They offer a connection between traditionally formal and informal systems. But they work best in environments with a large population of agents. A human dealing on a one-to-one basis with a computer that was acting autonomously on a strategy of statistically derived norms, might find the experience quite exasperating

// posted by Greg FitzPatrick @ 00:21