Sunday, March 23, 2014

Making Good Characters Rationally Self Destructive



(Picture from here.)

About years ago, back when I was in college and dinosaurs ruled the earth, I ran across Garret Hardin's lovely book, The Tragedy of the Commons. If you want to go out there and read it quickly, go ahead. It's not long. I'll wait.

The metaphor was the British Commons where the townsfolk would share grazing land for their cattle. However, there was an incentive for an individual to sneak in and graze more than the others if he could get away with it. His cattle would get a little more food and he'd do better economically. Of course, if one person could do it, anybody could do it and ultimately the Commons is destroyed.

To summarize the book in a single sentence: short term individual gain trumps long term common good.

It's a beautiful metaphor for a lot of things. For one thing it shows why public space, be it capitalist infrastructure or water and sewer, has to be protected. Once I became aware of the process, it showed up everywhere from food safety enforcement to securities fraud.

It explains why governments are necessary.

But it does not explain irrational behavior.

Adam Smith's "Invisible Hand" is a purely rational construction. It has no moral conscience or presumption of anything other than the iron rules of supply and demand. Prices go up as demand goes up or supply is restricted.

But as we all know we are anything but rational.

Let's be very clear. I am not talking about how people accept and cling to irrational beliefs. That actually makes an odd sort of rational sense. Think of it in terms of our ability to build serviceable models on incomplete data. Whether we're trying to figure out or in-laws or determine if our children love us, we're always building models composed of incomplete data. We then cling to the inherent phenomenology underlying them. It's the end result of our natural selection. After all, the cost of thinking a leopard is under that bush is enormously less than thinking there's no leopard under the bush when their is.

Nor am I talking about addiction, though we may end up talking about that as well. Addiction is a well characterized neurological phenomenon. It's not pleasant or rational but we (sort of) understand it.

No, I'm talking about  people can continue in a given course of action when it clearly, demonstrably, not being successful. You've all seen it. Somebody gets into a sticky situation and instead of stopping the behavior that got them there, they double down on that very same behavior. I'm thinking of this clinically. As a writer. After all, good characters need a healthy dose of self destruction.

How can people continue to do that?

Enter Prospect Theory.

Prospect Theory is the invention of Daniel Kahneman, who won the Nobel Prize for inventing it. It reformulates the decision making process in the form of risk analysis when the probabilities of outcomes were known. People tend to make decisions based on the relative weights of the risks of the wins and losses rather than the wins and losses themselves. The wights do not have to be equal. And they tend to be relative to a reference point where the balance of risk is considered equivalent.

One would think human beings would biased towards chance of gain rather than risk of loss. After all, we are the invisible hand, right?

Actually, it turns out that we are weighted against loss rather than towards gain. Consequently, we often aggressively increase the same behavior in the face of loss. Imagine a gambler who hits a losing streak. Which does he do? Try to break the streak-- in effect, gamble out of the hole-- or accept the loss and walk away from the table? Okay, guess on the count: 1! 2! 3! Go!

We see this all the time. Workers who have embraced a profession have trouble giving it up when the profession is lost or moves overseas. People cling to racial stereotypes ferociously when they think implied privileges or cultural heritage is threatened. People in marriages cling to the trappings representing when things were good. If I just do this, like I used to, the marriage will be saved. It is too terrible to face the risk of loss even though the possibility of gain might be negligible.

We see it in the political arena. One party or the other suffers a defeat and goes through soul searching. Inevitably, they come back to their central message: We must be exactly the same but even more! More conservative! More liberal!

I worked at a Very Large Computer Company back in the 80s. VLCC never really understood why it's products sold. One colleague suggested our salesmen were like cheerleaders around a rocket, saying "Go Rocket! Go!" and the rocket took off and their products sold and everything was right with the world. But when the microcomputer revolution happened, the stock crashed, the products didn't sell and the poor rocket lay shattered and broken in a crater. The salesmen kept trying their old techniques: "Go rocket! Go!"

I had a friend, Phil, when I worked at Small Aircraft Company. One week, when the value of the lottery was a couple of hundred million, there was a movement to buy a bunch of tickets together. It was sort of like a bunch of kids deciding to buy a share in an imaginary airplane. Phil was agonized over this. He was a mathematician. He knew the odds were as close to zero as to be negligible. Why, then, did he feel so compelled to buy a ticket. Eventually, he did buy one. He said to me, "It's not that I think we'll win. We won't. It's because I'd feel so bad if you all did win and I hadn't bought a ticket."

One of the reasons I find Prospect Theory so compelling is that it sounds so rational. But what it is really doing is evolving a mathematical model around irrationality. While the theory itself is perfectly rational, the weighting of risk found in observation is not. We tend to overvalue low probabilities and undervalue high probabilities. Which is another way of saying the lottery is designed for the mathematically challenged.

It gives an interesting development to character development in fiction. Writers always have to decide what the character wants and will risk to get it-- the princess must overcome obstacles to save the prince and the nature of the obstacles show the depth her bravery and strength of her commitment. Now, we need to look at this more closely. Now we also want to know if the loss or failure of not going might outweigh the risk of failure of going.

After all, Gimli said: "Certainty of death, Small chance of success... What are we waiting for?"

More on Prospect Theory:
Putin and the Crimea
Kahneman's original paper
Applications of Prospect Theory to Political Science
Evolutionary Origin of Prospect Theory

Sunday, March 2, 2014

Topics of Interest


(Picture from here.)

A lot of good stuff coming over the transom from Io9 lately. The picture left was derived from Googles partnership with organizations to create a deforestation monitoring tool: Global Forest Watch. The blue tends towards reforestation. Red tends towards deforestation. Norway, status quo. Sweden, much change.

Here's the first ever geological map of Ganymede.

These tunnels are beautiful. I especially want to visit the Guoliang Tunnel in China. I wonder what the accident rate is.

There's a good discussion here on how Kepler has changed our view of the universe. Just a few years ago there was no evidence for any other planets other than hope and optimism. But Kepler has given us an estimate there are on average 1.6 planets around every star in our galaxy-- or 160 billion planets. And that doesn't include free moving planetary objects.

Kepler also suggests that 22% of the sunlike stars in the Milky Way are Earth sized. That's about 2 billion possible planets.

There's company out there. We just haven't met them yet.

Original articles herehere and here.

Finally, here's a really good article on the similarity between the brains of dogs and humans. First, the study examined the area of the brain in both humans and dogs that responds to the human voice. Turns out they are the same. Emotionally charged sounds such as whimpering dogs or crying children caused similar responses in both species.

The study suggests that dogs are as conscious as human children. (Original NYTimes article here.)

There are two distinct possibilities, both of which are discussed by the authors. The first is that in the 100k years we've been associated with dogs and the 20k years we've been actively working with them, we've modified their brains to respond to us. This is, no doubt true. However, the authors think the data shows something different. They believe that these associations are much, much older. They speculate that the structures are common between humans and dogs and existed when primates and canids ancestry split-- about 100 million years ago.

If this is true and dogs are as conscious as children then it brings right to the forefront the whole "is conscious necessary?" argument. This discussion centers around is consciousness an artifact of general cognition, an unnecessary emergent property of the mammalian brain or a positive attribute selected for by evolution? The cons of consciousness are its metabolic expense and relative slowness to unconscious activities. The pros of consciousness selection are exactly the same: if it's so counter productive why do we have it? Natural selection would select against such extravagant expense if it wasn't useful.

But if it shows up all over in vertebrates the discussion narrows to: it's either a necessary but useless byproduct of mammalian brain or it has positive attributes that outweigh its expense. It seems that some of the bird studies of intelligence might be able to at least broaden argument to byproducts of vertebrate brains.

Which, to my mind, makes the argument fairly thin. I suspect that though some birds might have a cognitive level sufficient to support consciousness, I suspect that many do not. Whereas I suspect that there are many more mammalian species that can support consciousness. This is not purely chauvinism on my part. Birds and mammals parted ways long before the evolution of the neocortex and its equivalent in avians. The neural mechanism used for intelligence in the two groups is markedly different.

Add to that the tremendous limitations imposed by flight. Whatever birds have to do in their brains it must be in a small package. The expense of cognition and conscious must be justified by its utility and the expense for birds is much, much higher than for mammals. So if crows, parrots, etc., are conscious it must be advantageous for them.

Original article here.