Archive for October, 2015


Why changing your mind is (usually) bad

I’ve written before about how I don’t really see changing your mind as a good thing. To me, it usually means that you made the mistake of having an opinion before you had sufficient information. Rational people shouldn’t be changing their minds often, because about most (policy) things they shouldn’t have an opinion.

To make this more concrete, I think it’s useful to think further about what an opinion is or should be. You can imagine two dimensions: how confident you are and how big of an advantage one thing has over another. For example, if an event occurred 4 times in 5 experiments, my best guess might be that the true success probability is 0.8. If, instead, an event occurred 4000 times in 5000 experiments, my best guess might be that the true success probably is 0.8, but I’d be a lot more confident about it.

If you’re going to think about ‘Bayesian updating,’ you shouldn’t just think about a point estimate, you should also think about uncertainty (something like a credible interval). Updates would tend to look like “I am a little less uncertain now” rather than “I changed my mind!” Typically, changing your mind just means you either ignored or previously underestimated uncertainty.

Most things about which there is a lot of debate will tend to not have much very good data supporting one side over the other. And yet, most people seem to have strong opinions. This leads to my next point

Many light bulbs without dimmers

Most political issues tend to have pros and cons that are pretty easy to see if you really try. Generally, things like: compassion versus tough love; privacy versus security; inflexibility versus discretion; etc.

However, almost all debates involve people arguing passionately for one side against people arguing passionately for the other side, without really acknowledging the trade-offs with either. These debates, where no one admits there are tradeoffs, do tend to lead to compromise.

I sometimes picture an alternative world, where everyone acknowledges the tradeoffs and a compromise is reached by discussing the tradeoffs and making a judgment on how best to weigh them.

It’s like trying to achieve a certain degree of brightness. In one room we have many light bulbs, some at full brightness and some off. In another room we have the same number of light bulbs, but they are all dimmed. Both rooms might be equally bright. Humans never seem to achieve the desired brightness with a dimmer.

Read Full Post »


I love watching people take care of their lawns. Today I saw people with leaf blowers. Grass basically says “I don’t want these leaves on me” and the humans race over and groom it. When grass turns brownish it’s saying “I’m thirsty” and the humans water it. When it gets long people cut it.

In Sapiens, Harari asked “how did wheat convince homo sapiens to exchange a rather good life for a more miserable existence?” In that case, the explanation, while also awesome to think about, is pretty easy to understand. In the case of grass it seems more complicated.

I enjoyed the 99pi episode on lawns. Lawns were really great for signaling wealth because it showed that you were rich enough to own land that you didn’t put to food producing use. It would be like if today rich people bought or built big factories but didn’t have them make anything.

Once lawns became extremely popular (and industrialized, to an extent), it was no longer just for the elites. So their social status purpose morphed. Currently, lawns seem to signal how good of a neighbor you are. If lawns were easy to take care of, by having one we wouldn’t be signaling to our neighbors that we are the kind of people who care for things.

However, it’s not quite as simple as: if I care for my lawn it shows I care about things and I’m a good neighbor; if I don’t care for my lawn then I don’t have pride in myself or my neighborhood. Some people pay lawn care companies to take care of their lawn. By having others do the work, you aren’t quite as strongly showing that you will work hard and get your hands dirty to do your part to make the neighborhood beautiful. You do, however, show that you have enough money to hire people to care for your lawn. Thus, one could argue that you are signaling that you are so important that you don’t have the time to do the yard work yourself, but you still care enough about how the neighborhood looks that you’ll pay to have it done. On the other hand, a person who is rich enough to pay someone to take care of their yard, but instead does the work themself, shows their neighbors that they’re ‘down to earth’ (not above physical labor).

Getting back to how grass gets us to care for it. If it was extremely easy to keep green, it probably wouldn’t have had so much evolutionary success. Being a bit of a diva can be an advantage. Dogs used a totally different strategy to get humans to care for them. Dogs just flood people with flattery. They basically say “oh my god, you are so great! I can hardly contain myself. I can’t believe someone as great as you exists.” Every inch of their body is used to make the dog’s caretaker feel important (waging tail, happy dog sounds, rolling / running around, jumping up the owner’s legs, etc). I love that there can be such different strategies to make humans do reproductive work for other organisms.

Read Full Post »