Thursday, April 7, 2016

Hive attacks

I am currently reading a fascinating book that looks at decisionmaking by the hive mind.  The book is called Smart Swarm by Peter Miller.  But instead of overlaying a bureaucratic concept on an assumption of how hives work, what he has done is look to the researchers who study hive behavior to figure out how the decisions are made:

How do ants know which path to take in order to get good stuff?
How do bees know where to go when they move the hive (swarm)?

Getting ready for the big dance.  Image from link
The answers, so far as I have read in the book thus far, are fascinating.  There is a decentralized decisionmaking process.  It is not a matter of convincing the leaders which path is best.  It is a matter of convincing yourself that you have a good product.  And do a waggle dance. Or leave a pheremone trail.  By combining our knowledge of the landscape, and looking at the problem repeatedly, we can decide to agree with a waggle dance of our neighbor, or waggle dance on our own, better.

It should be obvious that I am working on processing the ideas in the book. But the possibilities become incredible, if it is applicable to the bureaucracy.  Because we certainly have enough bee-brains working in my office with me to where we could use the hive mind to make decisions quickly and easily.

The bureaucracy in general is very hierarchical.  That means that my boss approves the action I take, but that approval only lasts as long as her boss approves her action.  At each level of the burrocracy, the answer can turn from a Yes to a No.  (And sometimes, with the right political pressure, the reverse happens).

So the project manager has to convince the four levels of the supervisory chain - up to the Colonel - and two different panels of external reviewers that the decision is a good one.  Then the Colonel sends it to the Division, and it is reviewed by three levels there, and is approved by the General.  The General then signs off on it being a good product, and it goes to HQ.  Two or three layers of reviews later (HQ policy review, Office of Water Projects review, and probably a review from the Assistant Secretary of the Army's office) and the Chief of Engineers hosts a meeting where everyone says OK.  The Chief signs the report.

After the Chief's report, it goes to the Office of Management and Budget.  Their review approves it for release to Congress, who authorizes the project.  That's right.  An Act of Congress is what it takes to get the work done.

And maybe, just maybe, they also appropriate funds to do the work.

At each level, there are demands.  At each level, technical and policy review identifies weaknesses in the project, and finds reasons why it won't work. Ways to 'improve' the product.

The hive approach is counter to that level-by-level approval system.  The approval is not sequential, but corporate.  The hive votes with its feet, getting behind the best solution as it 'hears' the plan.  (I suppose you'd still need drones to write the report, but that is a different matter).  It is a little like the cluster of people at the coolest exhibit in a museum.  

But what that approach would do would be to streamline government.  (Side note: I have heard people all my life tell me that government inefficiency is at the root of a lot of society's problems, but I am not sure I want a government capable of spending money any more quickly than we do now.)  In a decision mode, though, it has got to be better than this.

And on a day that seems like there are enough barbs and stingers in every corner, I think maybe just doing a little waggle dance with my coworkers would be the bees knees.

Let's dance.

No comments: