Any decent gambler knows that being in a hand has some value, even if you’re not likely to win. Any decent venture capitalist knows that some ventures are worth investing in, even if they’re not likely to succeed. Why? Because if you invest in 100 startups, and one of them turns out to be Facebook, then it was probably worth losing money on the other 99. Likewise, being in a hand of poker has real financial value. In a simplified case, the expected payout is the possible payout multiplied by the probability of winning. If that number is higher than the price of being in the hand, then (all else being equal) it is worth being in the hand.

These are positive examples of expected value. Negative examples also abound: You want car insurance to mitigate the negative expected value of getting into an accident, even if an accident is unlikely; you don’t want to play a round of Russian roulette even though your odds of shooting yourself are only 1 in 6; and so on.

Now let’s do morality. If it is immoral to shoot someone, it is also immoral to pointlessly risk shooting someone. If shooting someone with a revolver is immoral (all else being equal), then putting one bullet into a revolver, spinning the cylinder, aiming at someone, and pulling the trigger is also immoral. It is one-sixth as immoral (all else being equal) as shooting them, because it has a one in six chance of having the result that is the source of the presumed immorality. This is why we think it is immoral to drive drunk. It doesn’t usually result in injury or property damage, but it typically increases the risk of injury and property damage, and we think it is usually not worth the increased risk.

Now let’s do nihilism. Humans might think nihilism is probably true, but since human beings are fallible and can’t know the nature of morality (or lack thereof) with 100 percent certainty, we can’t know for sure. There may be such a thing as objective morality for future philosophers to discover, even if no human has yet understood it. Even if this is highly unlikely, we can’t say it’s impossible.

Just like being in a hand of poker has some financial value because you have at least a slight chance of winning, operating in the universe as a fallible entity has some moral value because you have at least a slight chance of operating morally. Any discrete hypothesis that potentially vindicable should receive nonzero credence. Of course, the moral value could be positive or negative. To extend the poker analogy: we don’t know if we’re playing for money or debts here. If we behave badly, then we will produce negative moral value, not positive.

In order to eliminate any moral truth from the calculus of our lives, we would have to attain total certainty that there is no moral truth. As long as humans remain morally fallible, that will not happen.

But if that is all we know about morality, what does “operate morally” even mean? How do we avoid the problem that bedevils Pascal’s Wager: namely, that every possibility has some opposite possibility canceling it out? I have begun attempting to answer that question here, here, here, and here, among other places.

Knowledge Maximalism

Nihilism Is Incompatible with Fallibilism