I have a friend I always go to for new books. He reads a lot and has a sense of what will interest me. And I know there is a good chance whatever he recommends I will like. This time it was Infotopia: How Many Minds Produce Knowledge By Cass Sunstein. And while it seemed to have some political agenda, it was a very good book. But now that I have started my next book. Things have just gotten weird.
“Truths are more likely to have been discovered by one man than by nation.” – Rene Descartes
The book is an interesting deep dive into what group think, statistical groups, and predictive markets can do for the world of problem solving.
After I finished the book, and right before I decided to blog about it. Things started to get a little weird. The reason why is because of the book I started just after. Trust Me, I’m Lying: Confessions of a Media Manipulator By Ryan Holiday. I’m only 1 hour ( I listen to my books on Audible ) into the book and i’m seeing an obvious connection to what Crass was talking about in Ryan’s book. But not in a direct way. Even more ironically, which you will understand soon, is the fact that i’m blogging about both books.
Let me start with the two ways Infotopia was interesting. First was from the scientific aspects of crowd intelligence, and second from the psychological aspects. What I did not like was the political slant and the skew to legal data. But both are apart of the author so you have to expect these biases to be there.
It was not just any science that I liked. Once I started thinking about how predictive markets ( Crass introduced me to) work, and related how they are really a problem solving engine comprised of snippets of thoughts from a large body of individuals, who are unrelated. I saw a clear analogy to this and a type of machine-learning algorithm i’m quite obsessed with called Genetic Algorithms. Like predictive markets, genetic algorithms are massively parallel, and essentially identify local mins and max on a statistical basis. Their other similar trait is the fact that the participates in both, either the human being asked a simple question, or the AI critter also being asked a simple question, are entirely un-aware of the other participants and overall objective. Strangely enough making them ideal candidates for a statistically accurate engine. The big difference however is that in Genetic Algorithms ( GA ) you repeat the question many times over many generations, and you start to change the participants based on the ones that work well or not. So basically predictive markets with computer critters instead of humans repeated thousands of times per second and you are permitted to kill off those individuals who don’t offer any value. This revitalized my love for Genetic Algorithms, and I think hints are some interesting possibilities when combining this technology for solving problems with this human method for solving problems
The second part that drove me was the psychology. Part of me was reminded of how frustrating humans are, and my hatred of group think became even stronger. I’ve seen group think kill families and companies.
Crass talks about the flaws in deliberating groups especially, and how they usually do not arrive at good decisions because the polarized individuals in the group slant the entire groups decision in one direction or another. What makes deliberating groups of individuals even worse is the fact that because they believe their decisions are better because they were made in a group, the polarization of the group amplify every time they use the method. I see group think kill companies, where nay Sayers in the company are subtly silenced over time because they start to anchor their thoughts to the average thoughts of the group. And this usually happens in an abyss where the strongest motivator in the company becomes the central belief. But I would hazard to guess that the strongest motivator, for example, “hit this number at all cost”, is not usually the best for the long term growth of the company.
Even more frustrating is the anchoring effect and effect of group think on our two completely polarized political parties. Crass helps prove my point that republics and democrats are not really contemplating any real issues, what they are doing is commiserating with each other and thus polarizing beliefs with out actual thought. If a republican hangs out with a group of democrats his/her original views will become even stronger out of backlash of theirs, and vice versa. If a democrat hangs out with a democrat, their views become even stronger because of the effects of deliberating groups to take the average position and make it even stronger. Both phenomenons have the only effect to make republics more republican and democrats more democrat, and their divide even stronger. I joke, they really just love to hate each-other. Thus nullifying any actually individual thought processes that may occur from a member of either of these groups. The net net is that South Parks decision making manatees would make better political decisions with random methods than any “leader” in these groups.
Yet deliberating groups are the favorite method of society, and somehow they work in-spite of themselves. Progress is always happening in ways we will never expect or plan for. And because we under estimate what happens in a long period of time and over estimate what happens in a short period. Bottom line is those who think they are being super smart and deliberative, are usually missing something big, and often wrong.
Now why is this so weird when I add in the fact that I just started the book “Trust me i’m lying”, and the fact i’m blogging about both?
Trust me i’m lying is about Ryan Holiday, a master media manipulation, who, when you understand the manipulation he and his peers have done to your favorite online, print, or TV news sources you consume, will be shocked. Combine this cool, scary, and sometimes sad reality with Crass’s explanation of how information is produced on the web, more specifically how Wikis grow. And finally add to that the fact that I am blogging about both books, and blogs are the source of Ryan and his peers power. You see a perfect storm of miss-information and deceit on one of my favorite things in the world, the Dub Dub Dub.
In a way i’m awed by the power of Ryan and really like it, and I get the sense that if the world will be manipulated in this way, then maybe they ask for it and deserve it. But at the same time, and with respect to the flaws in deliberating groups, and the huge power of the cognitive bias of Anchoring. It makes what is going on very real and very scary. Especially when smart people get consumed by these systems in the wrong way.
I recommend reading Infotopia. And so far my post on Trust me i’m lying should be as intense and interesting as this one.