Perfection is not a straitjacket

I attack hard problems the same way that I assume everybody else does.

When I chose my major in college, I tried to imagine the most likely outcomes that would result from each choice, and then weighed them against each other on a basically linear scale of “total expected happiness”. It wasn’t really as formal as that — but I spent a lot of time thinking about what was most important to me in a job (time, money, risk, intellectual stimulation, social impact), and I tried to choose a major that would give me all those things, maximized in the right order of priority. Basically I made pro/con lists.

But these comparisons never seemed to reach a satisfying conclusion, and still haven’t.  I love the degree that I chose, and really do believe it was the best fit overall — but when I have a dull day at work, I think longingly about international relations or journalism, and when money’s tight, I daydream about computer science or accounting. (Seriously.)

I assumed, like Dr. Chang did, that this anxiety was rooted in my own ignorance: maybe I’m bad at predicting which choice will give me the things I care about most, or maybe I care about the wrong things. I assumed it was, fundamentally, an information problem; that there was one correct choice, and if I could have all the ramifications of all the possible options laid out in front of me, it would be possible (if not trivial) to spot the “best” outcome, and take it.

This kind of thinking also colors my intuitions about God.

If all our decisions are analytically solvable, a perfect being never faces any decisions at all. God is always perfectly aware of the most correct course, and he always takes it. Perfection might allow him a few aesthetic liberties, but otherwise he’s straitjacketed.

And if our goal is to approach that perfection, then we’ll be straitjacketed too; only worse, because we’ll be identical and redundant: all responding to the same perfect information with the same perfect morality, and supplying the same optimal output. (Not to belabor this point, but it’s just one more nightmarish consequence of Plato’s insistence on philosophical tidiness, his need for everything to converge on a single point.)

Leaving aside religious questions, the idea that hard choices always have One Right Answer is a troubling one, because it implies that your “identity” is just your peculiar configuration of stupidity and moral brokenness — the unique way in which you fail to arrive at the optimal conclusion. And in the paradisiacal post-singularity future, when all computational problems become trivial, we’ll all be slaves to our perfectly defined utility functions.

But Chang rejects the idea that all hard choices are “solvable”.

In other words, there’s no way of breaking down the various results of each decision into “happiness points” and deciding which pile is bigger, because they involve measures of value that simply aren’t comparable. It’s not just computationally difficult to determine whether marrying one person or another is your optimal decision — it may well be impossible.

This doesn’t mean there are no sub-optimal choices you could take– there certainly are, both for moral decisions and pragmatic ones. But there may be multiple “best” choices, each of which maximizes a different set of values, none of which are comparable to each other. (For what it’s worth, this appears to be the Church’s official position, at least on the question of “soul mates”.)

Of course, utilitarianism (and by extension, rationalism) is built around the opposite assumption. If we can boil down all the possible consequences of a decision into quantifiable, comparable units, then there is an analytically “correct” choice — or at least, a choice that is the clear best fit to one’s preferences.

Intuitively, it seems wrong (or maybe just distasteful) for there to be exactly one optimal response to all the decisions that make us who we are — but it’s hard idea to disprove.

No matter how complex a problem is, it’s always possible that the answer could be “moar data” or a more sophisticated analysis. So far, every objection I’ve come up with could be answered with, “Well, that’s just one more thing to factor into your utility function”.

But when I go back to the thought experiment — imagining all the possible outcomes laid out before me — it seems like it really would be impossible to compare the different lives I might live. This seems especially obvious when I consider the granular details of spending my life with one person versus another, or raising one set of children versus another.

I simply wouldn’t trade my wife and daughter for any other family, no matter how ideally suited to my preferences. It’s a choice with an effectively infinite opportunity cost. But that doesn’t mean it was the only right answer. If I could see all those hypothetical forgone relationships with as much clarity — maybe a dozen different happy lifetimes of intimacies and old stories and shared struggle, with a dramatically different cast of characters — it seems like each of them would also become incomparable and irreplaceable to me.

There are certainly a lot of sub-optimal outcomes; but it’s also easy enough to imagine more than one happy outcome, and all of them different.  (I know Tolstoy said all happy families are alike, but he was also a 19th-c. Russian novelist and therefore a huge buzzkill.)

This is possible partly because preferences are dynamic, and (at least to a certain extent) within our control.

In other words, I don’t just have the freedom to live in New York City or Omaha. I also have at least some capacity to deliberately fall in love with my choice — to make it right for me (or more accurately, make myself right for it.)

Likewise, as long as you find the right sort of person (one of the many “right” people for you), you can have that incomparable, irreplaceable happiness — though you are making a decision about who you will become in order to achieve it. And I’ve never met a couple so perfect for each other that their love wasn’t at least partly a conscious choice.

To some extent, these preference changes happen whether you want them or not. If I had majored in journalism or computer science, I would have been constantly surrounded by other wannabe journalists/programmers, and I would have chosen mentors and role models in “my” industry, and my definition of success and what’s cool and what matters would soak in from a completely different social ecosystem. The effect would be even stronger if I chose to nurse and encourage it.

In short, I could work hard to become the sort of person for whom journalism/computer science is the right choice. That doesn’t mean it would work out, but the point is that your choices can change your preferences — so that an “optimal” career (or spouse, or community, or whatever) is one that is close enough for you to manually nudge your preferences into harmony with it. And people who master that skill enjoy a much larger scope of possibilities.

Needless to say, if we can liberate the idea of perfection from the idea of uniformity, that’s a pretty big deal.

If nothing else, it can make life a little less terrifying for folks like me who are constantly re-litigating their choices, always wondering if they did the exact right thing, or did enough. It makes happiness an ongoing choice, rather than a train you might miss; and it helps explain why greater freedom to make the “right” choice doesn’t seem to make people happier.

It also resolves a few theological paradoxes: for example, it might explain why God gave Adam what seems like conflicting commandments — he could either keep the commandment to stay away from the fruit, or he could keep the commandment to cleave unto his wife. It wasn’t a trap, or a test for him to pass or fail — it was a choice. (And it became the right choice after the fact.)

For that matter, it explains why God would be interested in free will and consciousness in the first place (since it’s more than just “the ability to get the wrong answer”). It explains why he would aspire to create “Saints; Gods; things like himself” — since those Saints and Gods will be more than just an army of redundant, identical, fully optimized utility zombies.

More importantly, it makes theosis, or heaven, or singularity, or whatever eschaton you’re anticipating, seem like something worth having. Instead of merging with the undifferentiated divine (extinguishing choice and consciousness and identity), or becoming wirehead gods on lotus thrones, perfection looks more like infinite diversity, in infinite combinations.

Perfection is not a straitjacket