Peter Small of UK has contacted me about including WebSage into a demo of a stigmetric system he is building for search engine optimization based on the way insects collaborate in hunting for food. It is an intriguing idea, particularly for me, since my dad is a beekeeper. My dad and I have often discussed the fascinating kingdom of bees with their detailed hierarchies, job roles, inter-relationships, etc. While I am not a biology expert, here are some preliminary arguments, based upon my limited knowledge of bees and bee hives, explaining why I am not as enthusiastic as Peter Small about the role of software agents replacing human experts and in particular about the system he proposes…
My initial thought about the system Peter proposes is that it is based upon collaboration. What about the situation when systems or agents compete?
Bees, very much like the contributors to a SEO forum, cooperate only until there is an incentive, whether it is the presence of a queen bee, or the presence but not overabundance of food.
If two conditions are present at the same time:
1) The queen bee is weakened
2) There is too much food is available and the working bees fear that the queen bee is not “generating” enough new working bees, a group of them (experts? agents?) might split into a separate hive (forum?)
What happens next is that the former and the future hive-residents might compete quite aggressively… I am not a biologist, of course, but I am sharing this to indicate a possible flow to the architecture envisioned by Peter. Do you agree? Do you disagree? Feel free to share your thoughts!