Change Is Bad

Epistemic Status: Public service reminder (I want to be able to link to this in the future)

Leads to: Choices are Bad, Choices Are Really Bad, Complexity Is Bad

Almost all changes are bad.

People forget that. They say they want change. They say things like:

At the end of the day, I want to see change come about, whatever it would take. Whatever it would take to see change come about, I would welcome. - Chumbawamba, Be With You

What they actually want one of those rare, carefully chosen good, friendly changes. They do exist within change space.

Change space, like mind space, is deep and wide. Friendly change space isn't quite to change space what friendly mind space is to mind space, but before you apply any filters of common sense, it's remarkably close.

The more optimized things currently are, the less likely any given change is to be good.

The more time people have had to optimize other things around the current state of the thing you are trying to change, the less likely any given change is to be good.

The more effort people have put into optimizing other things, based on the thing you are looking to change, the less likely any given change is to be good. You could break a lot of things.

When you break those things, you cause harm. Since people hate losses more than they love gains, even a net improvement can make a person or group feel worse off.

If you do break things, often they stay broken. It is usually a lot harder to build or repair something than it is to break that thing.

The faster and bigger you make changes, the more other things you are likely to break, and the more critically you will break them. At a minimum, even when your change is strictly for the better, those things then must change to adapt. In many cases, they are broken entirely, beyond repair, and this goes on to break other things.

Modern life is highly optimized. It's far from perfect. Often it is optimizing for the wrong things. Nevertheless, it is highly, highly optimized. We certainly do not lack for hill climbers. In some ways it is too optimized!

If I currently have a highly optimized package of things, such as a house or apartment (which is implicitly a collection of things tied to my location), often I will have put a lot of work and sunk costs into the current package of goods I've assembled, so breaking up that package or adding new things to it is pretty bad if I still need to pay market price. You can easily make everything better, and still mostly make things worse until equilibrium establishes itself again.

Older people have had more time to optimize and have less time to optimize again, so it makes sense that they hate change even more than others. They should.

Seriously, though, before you go around talking about how everything is awful and we need to call for revolution and change everything right away because the world is a nightmare, I would like to remind you that things are really, really good right now relative to baseline. We have a lot of nice things, even if we're having trouble building new ones. More people than ever, in absolute and in percentage terms, have high technology, connection, information and entertainment. More of us have enough food and water and shelter and peace and prosperity, and we mostly have our freedom and mostly get along pretty well all things considered.

Respect that. Do not screw with this lightly.

By all means, get mad about the planetary death rate and the people who are still starving. Get mad about housing costs and lack of employment opportunities and infection rates. Rage at the irrationality of it all. Realize that unless we do something to prevent it, unfriendly artificial general intelligence will probably destroy all value in the universe. We might want to do something about that.

First, count your blessings.

Once that's done, yes, we need change. The only constant is change, other things are changing and breaking your things, forcing them too to change, and all that. We're all slowly dying of old age and forces move to wipe out all value in the universe, plus we're getting pretty tired of the same old restaurants and albums and TV shows.

So what are we to do?

First of all, we don't choose our change at random. No matter how bad a person is at deciding what to do, they're a lot better than random.

We hopefully don't fall for 'something must be done, this is something, therefore we must do it,' even if we do occasionally go for 'something should be done, doesn't matter as much what it is, this is something reasonable, so let's do that' since we also don't want to fall for 'something better could be done, so until we figure it out no one change anything.'

We try to explore and experiment, so that we have a better idea of what we are breaking before we implement things on too large a scale, and we think carefully about what the consequences might be.

Then we change things anyway, because we have to, but with our eyes open.

We don't capture all the upside of our changes; companies that provide valuable services, and people who do useful work, usually get only a small fraction of the value they create. In exchange, neither do they end up paying for most of the things they break and the harm they do. Which is how it must be. The policy question is where we can usefully change the rules about what people can change, and what they get in exchange, in a way that gets better results than letting events run their course.

On an individual level, you hope enough things roughly offset that the incentives are good enough, and you respond to them.

Them hopefully, as Raymond often advises, we make good choices.