http://www.economist.com/science/displayStory.cfm?story_id=2121720
Article reporting on the first conference on agents in networking, displaying areas of success [principally financial markets] as well as areas where the application is more speculative, but has much broader payoff effects. This area of research is well worth constant attention of networking folks for two reasons:
1) It represents a major opportunity for more effective control of highly complex networks; and
2) If successful, it will cost most of us our jobs.
The agent concept is one which hitherto has not reached its considerable potential, and progress may be about as frustrating as with speech recognition, but it is nevertheless something which could become a dominant feature of our future lives.
Posted by jho at October 16, 2003 12:06 PM