Finding purpose in this world is not a simple problem. People talk about looking to have impact, creating value, making themselves useful, or working on important things. However they phrase it, there is meaning in looking to make things better, in trying to create progress.

But what is progress? There are many answers: it’s folks feeling happier, living longer, things being more efficient, people being more equal, future generations not having the same problems we have, … Progress means so many things, it’s tempting to just say “I know it when I see it”. But it’s not enough to just consider types of progress in isolation; there are often trade-offs between them. Do we build that bridge or give people better healthcare? Should I spend my time on fundamental research or helping people in need? Which kinds of progress should we prioritize as we build general Artificial Intelligence? It would help to have a single definition of progress that folks can agree on.

Before we get into the definition of progress, let’s think about what’s good for people. For a long time, we’ve optimized for the obvious things like access to food, shelter, and basic healthcare. We need to continue doing that, but beyond these, it’s not obvious anymore. There isn’t a clear list of unequivocally good things we want to achieve. There isn’t a single ‘way to be’ that we can push on folks. That’s why we should focus on what people want, that is, on human will. Not just in material things but whatever it is that people want to experience. The fundamental value we should be optimizing for is the exercise and fulfillment of human will. Every other good thing comes from people’s will being fulfilled. Even though there is some logic around it, this is ultimately an axiom, a principle that you either believe in or you don’t. I like to call it the sanctity of will. 

The definition of progress follows: Progress is improvements in the ability for human will to be fulfilled, totaled over all humans, present and future, and all time. That means that progress has happened when it has become more likely that humans get to experience things they want to experience, or the things they would want to if they weren’t already experiencing them. That is the definition of progress based on the sanctity of will.

In order to estimate how much progress our actions lead to, we need to think about the total sum of will-fulfillment-improvements across all people, present and future. We are considering all people’s will-fulfillment to be weighted equally for progress. Does that mean that everyone should have access to the same resources so that no resources are wasted? If the world was going to last only a week, that would be the right way to optimize. But not if we care about will-fulfillment in the near and far future. Folks who help others’ will become fulfilled, like folks who provide services for others, make things more efficient, or make new inventions to help current and future generations, need to be rewarded to the extent that makes others get more of what they do (and no more). From this point of view, giving folks different abilities to exercise their will is a necessary consequence of optimizing for total will-fulfillment into the future, not because it’s some rule of nature. It’s not “greed is good”, it’s “this is your reward for helping people exercise their will”.

If we’re counting total will-fulfillment, does this logic mean that if 90% of people want to treat the remaining 10% unfairly, that’s ok for progress? Definitely not. Usually, the 10% will want to not be treated unfairly a lot more than the 90%’s will, and they will often be willing to give up a lot more to prove it. There are a lot of examples in history of this, for better or worse.

How do we tell how much people really want something? Asking is not enough since people might just say they want everything, a lot. The true way to learn how much people want things and direct resources accordingly is for folks to signal that they are willing to risk/give the only thing they have that matters: part of their ability to exercise their will. This is the true magic of the marketplace. When it works, it’s not just a tool for self-organization of production, it’s also a way to find out what people really want and direct resources accordingly. Democracy is also a great example, you only have one vote, by giving that vote, you’re spending that ability to exercise your will. We need to double down on mechanisms that give us true signal on what people really want.

This way of thinking is already there for many. When people talk about freedom, about making themselves useful to others. They are thinking about the sanctity of will. For others, who talk about impact at scale, potential for value, or the will to do good, it is perhaps unconscious, or implicit. Focusing explicitly on the exercise of human will simplifies thinking about progress, while encompassing the other things people think of as good.

Thinking about the sanctity of will also helps think about the effects of Artificial Intelligence (AI). Will general AI be good for people? It will be good if it makes it easier for folks to exercise their will. It doesn’t matter how many jobs general AI will take, as long as the net probability that a human fulfills their will is improved. And, from this point of view, AI represents the biggest opportunity ever. The thing that’s unusual with general AI is that there is a substantial probability that human decision making and effort become increasingly decoupled from technological or market improvements, at which point the winners might say: “Oops, I happen to own the economy, which runs itself at this point, so I don’t need you people.”. We will need mechanisms to ensure that we are still optimizing for total human will-fulfillment; not perpetuating some old legacy of how we got there.

Next time you are thinking about progress, trust others, present and future, to know what is good for them, and center your thinking on empowering them to fulfill their will, and work backwards.