The busyness of going (day 263)

Getting ready for the Oboro residency.  The thought of being away for almost three weeks is both exciting and frightening.  I've resigned myself to not being ready but being ready enough.  What I most look forward to is the spark of creativity coming back, from sputter to more like sparklers you put on a birthday cake.  Something happens when the day to day is interrupted by something other than tragedy or sickness.  Threads are lost and others find each other.  That's interesting and unpredictable.  It's too bad the physical constraints are such that these kinds of interruptions are necessary.  Don't think the irony of leaving such an amazing research lab to do creative work escapes me.

Meanwhile, in whatever pockets of calm I can find, and in whatever interesting company I find myself, I'm bouncing off ideas about crowds, the topic I will speak about at Smart Graphics.   The management of crowds is especially interesting to me right now and I realize I've been thinking about this for a long time.  I was even thinking about this during my PhD. But back then I was into local rules influencing morphology.   That's still interesting but there's the other way around too - the context influencing the local behaviour.  I'll be looking for a way to map the ideas and sources I've collected in the past few months into some sort of map. 

A song for this post.

If your friends jump off a bridge (day 228)

Jaron Lanier says that every advertisement is a failure of the hive mind.  He doesn't like the hive mind or advertising so I don't think either one wins in this statement.  It is interesting that so much faith has been put in the hive lately when we know that the hive slices both ways.  Crowds can topple walls and can loot stores.  Thinking as a unit does not equal wisdom.  It's just a stable state that could be any stable state.  There's no particular reason to assume it's the optimal solution.  And often we don't even know the question we're asking of the hive anyway.   When the hive speaks the assumption is that it's uttering the answer to something.  Once again Douglas Adams was prescient.

It's pretty clear that complex problems are probably left to good old analysis and experiment by domain experts.  Even the example of the NetFlix prize is not a good example of hive computation.  A gladiator style winner take all model is not optimal in terms of evolution of a solution.  It's an example of a crowd of smart and hungry researchers, that's all.   As far as I can tell we are far from anything resembling optimized hive computation.  In fact, research shows that we have great influence over each other when trying to reach consensus and local influences can prevail, preventing a convergence to the correct solution.  Still the idea of hive computation is so attractive that we want to believe it.  We want to believe that magically a crowd not only could but will come up with solutions to hard problems like global warming, and unstable economic systems.   Like Douglas Adams says, the simple existence of the human race expresses solutions all the time.   The problem is that as a species we're short-sighted and we want to be liked.  And I'm pretty sure the former doesn't get better at the crowd level.

I don't think we should entirely give up on hive computation at all.  I agree with Jaron that we should study it with a clear focus - what is it good for and how to we set it up for success.  Undoubtedly, there are certain problems that stated in a certain way might be ideal  for the hive but without a clear study we'll keep fumbling and believing a false god.

A song for this post.