And I don't mean that they must concern themselves with death in the sense of ending death, or removing its sting through mental backups, or delaying it to the later ages of the universe; or in the sense of working to decrease the probability of extinction risks and other forms of megadeath; or even in the sense of saving as many lives as possible, as efficiently as possible. All of that is legitimate and interesting. But I mean something far more down to earth.

First, let me specify more precisely who I am talking about. I mean people who are trying to maximize the general welfare; who are trying to achieve the greatest good for the greatest number; who are trying to do the best thing possible with their lives. When someone like that makes decisions, they are implicitly choosing among possible futures in a very radical way. They may be making judgments about whether a future with millions or billions of extra lives is better than some alternative. Whether anyone is ever in a position to make that much of a difference is another matter; but we can think of it like voting. You are at least making a statement about which sort of future you think you prefer, and then you do what you can, and that either makes a difference or it doesn't.

It seems to me that the discussions about the value of life among utilitarians are rather superficial. The typical notion is that we should maximize net pleasure and minimize net pain. Already that poses the question of whether a life of dull persistent happiness is better or worse than a life of extreme highs and lows. A more sophisticated notion is that we should just aspire to maximize "utility", where perhaps we don't even know what utility is yet. Certainly the CEV philosophy is that we don't yet know what utility really is for human beings. It would be interesting to see people who took that agnosticism to heart, people whose life-strategy amounted to (1) discovering true utility as soon as possible (2) living according to interim heuristics whose uncertainty is recognized, but which are adopted out of the necessity of having some sort of personal decision procedure.

So what I'm going to say pertains to (2). You may, if you wish, hold to the idea that the nature of true utility, like true friendliness, won't be known until the true workings of the human mind are known. What follows is something you should think on in order to refine your interim heuristics.

The first thing is that to create a life is to create a death. A life ends. And while the end of a life may not be its most important moment, it reminds us that a life is a whole. Any accurate estimation of the utility of a life is going to be a judgment of that whole.

So a utilitarian ought to contemplate the deaths of the world, and the lives that reach their ends in those deaths. Because the possible futures, that you wish to choose between, are distinguished by the number and nature of the whole lives that they contain. And all these dozens of people, all around the world of the present, ceasing to exist in every minute that passes, are examples of completed lives. Those lives weren't necessarily complete, in the sense of all personal desires and projects having come to their conclusion; but they came to their physical completion.

To choose one future over another is to prefer one set of completed lives to another set. It would be a godlike decision to truly be solely responsible for such a choice. In the real world, people hardly choose their own futures, let alone the future of the world; choice is a lifelong engagement with an evolving and partially known situation, not a once-off choice between several completely known scenarios; and even when a single person does end up being massively influential, they generally don't know what sort of future they're bringing about. The actual limitations on the knowledge and power of any individual may make the whole quest of the "ambitious utilitarian" seem quixotic. But a new principle, a new heuristic, can propagate far beyond one individual, so thinking big can have big consequences.

The main principle that I derive, from contemplating the completed lives of the world, is cautionary antinatalism. The badness of what can happen in a life, and the disappointing character of what usually happens, are what do it for me. I am all for the transhumanist quest and the struggle for a friendly singularity, and I support the desire of people who are already alive to make the most of that life. But I would recommend against the creation of life, at least until the current historical drama has played itself out - until the singularity, if I must use that word. We are in the process of gaining new powers and learning new things, there are obvious unknowns in front of us that we are on the way to figuring out, so at least hold off until they have been figured out and we have a better idea of what reality is about, and what we can really hope for, from existence.

However, the object of this post is not to argue for my special flavor of antinatalism. It is to encourage realistic consideration of what lives and futures are like. In particular, I would encourage more "story thinking", which has been criticized in favor of "systems thinking". Every actual life is a "story", in the sense of being a sequence of events that happens to someone. If you were judging the merit of a whole possible world on the basis of the whole lives that it contained, then you would be making a decision about whether those stories ought to actually occur. The biographical life-story is the building block of such possible worlds.

So an ambitious utilitarian, who aspires to have a set of criteria for deciding among whole possible worlds, really needs to understand possible lives. They need to know what sort of lives are likely under various circumstances; they need to know the nature of the different possible lives - what it's like to be that person; they need to know what sort of bad is going to accompany the sort of good that they decide to champion. They need to have some estimation of the value of a whole life, up to and including its death.

As usual, we are talking about a depth of knowledge that may in practice be impossible to attain. But before we go calling something impossible, and settling for a lesser ambition, let's at least try to grasp what the greater ambition truly entails. To truly choose a whole world would be to make the decision of a god, about the lives and deaths that will occur in that world. The future of our world, for some time to come, will repeat the sorts of lives and deaths that have already occurred in it. So if, in your world-planning, you don't just count on completely abolishing the present world and/or replacing it with a new one that works in a completely different way, you owe it to your cause to form a judgement about the totality of what has already happened here on Earth, and you need to figure out what you approve of, what you disapprove of, whether you can have the good without the bad, and how much badness is too much.