Hive Mind Redux

April 23, 2007

I’m just tickled that my recent article on the hive mind has garnered some attention on Digg.  As a quick follow-up, I’ll  say that my talk on beekeeping and the hive mind at the Web 2.0 Expo was well-received and led, through random happenstance, to an introduction to a Carl Anderson, developer and co-founder of BatchTags (a web-app for sharing information), but more relevantly, he spent a decade and a half as a researcher on swarm intelligence, studying such phenomena has how an ant colony can determine which of two food sources is superior despite the fact that no single ant ever visits both choices.

I’ve just started to sift through his publication list, it’s impressive.  I was particularly interested in his article on worker allocation in bee hives, which discussed how bees use a host of different communication mechanisms, including the tremble dance, shaking signal, stop signal, abandomnent and, of course, the waggle dance, to adapt the worker distribution of workers based on need.  For example, many bees collecting nectar may produce a temporary glut of collected but unprocessed nectar, leading to a need for bees to shift from collection to processing.  Without any centralized decision-making, individual bees make this switch at the appropriate time by reacting to the communications and behaviors of the bees around them.

I was just discussing with someone I work with at my software job that it’s frustrating to see certain projects fail for lack of resources while other, seemingly less more important or wise projects, continue to consume cycles long past the point when it is apparent to many that time is being wasted.  Reading Carl’s article got me to wondering whether more fluid, dynamic systems for allocating human workers could be devised based on decentralized decision-making and simple intra-person communication. 

What if people within a company were free to work on whatever they wished and were provided with a set of communication tools that would foster the feedback loops necessary for self-leveling?  Would they naturally organize and adapt to the needs of the organization, or are humans too selfish, and too full ulterior motives to assign themselves based on priorities appropriate to the company, not necessarily themselves? And are the skills and knowledge necessary for humans to be effective in their jobs much harder to learn than those bees are born with?

I have heard that Google engineers work somewhat along these lines, and are encouraged to pick projects to work on that are interesting.  I don’t know believe, though, that they switch roles and projects dynamically based on need, though. Can anyone think of an example of a human self-organizing system like this?

Oh, and because I know that there are some people that read this blog for the beekeeping info, which is what it is actually focused on, a quick update:

The new colonies I bought a couple weeks ago are both doing well.  I had a bit of a conundrum on my hands because I had purchased a new queen for the ailing Hive 2, and then realized when I went to introduce her to the colony that there weren’t enough bees to support her until she got laying.  I bought a full 4 lbs package of bees to supplement the hive, but that package came with it’s own queen, leaving me with an extra queen. 

I had set them both up separately, and my plan was to combine the two hives and kill one of the queens. Fortunately or unfortunately, depending on whose perspective you’re taking, when I opened the hives today I discovered that the Hive 2 bees were unable to work the marshmallow cork out from the queen cage I had left in there a couple weeks ago, so the new queen was still in her cage, very dead.  Reminds me a bit of Poe’s Cask of Amontillado, without the revenge motive.

Anyway, I put the new colony on top of the old with just a couple sheets of newspaper between them.  They should chew through that in less than a day and then they’ll be one happy family with the new hive’s queen. The newspaper allows their introduction to one another to be slow enough that they can get used to each other, rather than simply dumping them in together, which might lead to some internecine warfare.

Related Articles

I’m excited to try a new experiment this year: top-bar beekeeping! In traditional “Langstroth” hives (or at least, traditional since the mid-1800’s), rectangular frames are placed into rectangular boxes, and the bees build their comb …

March 28, 2011

Based on comments from Rusty and some that I received through Facebook, I think the verdict on the Bee Kill question is clear: the girls found something they shouldn’t have. Rusty left this comment: The …

December 7, 2010

What do you get when you combine two hives of very active, healthy bees with one very active, healthy two-year-old boy? Nah, it’s not the set-up for a joke, just a setup for about a …

November 21, 2010