Meta-Optimism Is Just Optimism

Thus don't feel the need to vociferously defend your broadly general "optimism."

A common critique of optimism is that one could, for example, delusionally believe the stock they just purchased was destined to grow by 1000% no matter what the signs were or what the rest of the stock market was going to do.

More generally, one could believe that whatever specific strategy one has already chosen from the set of all possible strategies is guaranteed to beat all other possible strategies.

One might decide to call such a mindset “optimism” or at least “excessive optimism.” As if optimism were a scale, and that if you increased the level of optimism you had to a high enough degree, it would result in a mindset like the one described.

But these critiques rely on a scenario in which the optimism is fixated on a single strategy, not the decision procedure itself. I might have a strategy, which is sort of like a level-0 decision procedure. Then I might have a level-1 decision procedure, which helps me pick strategies. Then I might have a level-2 decision procedure, which checks to see if my level-1 decision procedure needs to be updated, and so on.

I assume that once we get to recursive feedback loops - the “meta” level - I don’t really need to keep increasing the levels of my decision procedures ad infinitum.

What would “excessive optimism” in the meta-level decision procedure mean?

Well, I conclude that it would simply mean that I expected my highest-level, recursive decision-procedure to work, generally speaking. That is, that I expect doing it to be better than not doing it at all.

By the way, because it is at the top-level, if I believed it would cause me to be extremely successful, as opposed to being merely successful, that doesn’t change any of the lower-level mechanisms of the decision procedure. A shift in how successful I expect to be doesn’t affect which level-0 strategies I end up actually picking. Although it may affect my mood, if my expectations aren’t quite normalized properly. But note that the optimal mood is therefore achieved by having rational expectations of success, and I expect merely to achieve the best mood I can!

We might choose to call this top-level decision procedure meta-optimism, to distinguish it from regular optimism.

But I feel like meta-optimism deserves to just be called optimism, so that we don’t have to constantly defend our choice to call ourselves “optimists” in public.

I argue that it deserves to be called optimism, because the strawman version of optimism is actually excessively pessimistic to all other level-0 decision procedures (and level-1 and up) besides the one already chosen. It basically says that it is better to not use higher-level decision procedures, all of which include the ability to update from mistakes, even if you have them.

So let’s define optimism to mean whatever top-level recursive decision procedure actually has the highest faith in itself — and which minimizes meta-pessimism.