Civilization Shaped by Fundament Physical Postselection

I delayed writing on this for a long time because I was worried it may be an infohazard. I don’t think it is, and I think if this is true then it is valuable that people know about it, so I’m publishing it.

Imagine there was a physically real effect of postselection. That is, there is a simple bare probability distribution among possible initial conditions for the universe, and the universe we inhabit is not sampled from this bare distribution but rather from a subsample of all initial conditions that satisfy some criterion about the future, or a distribution that’s weighted by some function of the future. In other words, the history of the universe as a direct physical law is optimized to satisfy some criterion or maximize some function about its future.

One possible history to optimize this criterion is if intelligent life evolved that deliberately took action to optimize this criterion. This might be the optimal history if there were a long time-period that was weakly constrained by postselection where life can evolve which temporally precedes a highly constrained region of spacetime that is constrained to a conformation that is far more easily achieved by intelligent design than any abiotic mechanism. This intelligent species will become fully devoted and competent towards satisfying this optimization criterion, except to the extent that physics does not force the criterion to be maximally satisfied or the components that are more easily satisfied by luck than deliberate design.

The civilization of this species will develop in the most likely way possible that reaches this end state. For the most part it looks just like history that is not postselected. At some point the civilization commits to the value of satisfying the postselection criteria. This could start with a small number of people proposing that something like this is possible, then the idea gets refined as the underlying physics is better understood and the moral aspects thought out, and then it gets popular or influential to the point of complete fixation. The few people that first propose the idea are in the unusual situation where they are also the first to anticipate that it will become influential for a rather different reason than what ordinarily makes ideas influential. This weird state of affairs is what was making me worried about infohazards.

However, all things considered I have decided even in the unlikely case where something like this is true, it is preferable for me to share this idea in the open than to try to keep it private. My general impression is suspicious of “here’s an idea that shouldn’t be spread” so I need a strong countervailing reason against this one. In the likely event that nothing like this idea is true then there is no anomalous effect that could provide such a reason. If this idea is true then an individual choice to not spread it is unlikely to be effective. Finally, I don’t think a block-at-all-points-of-intervention strategy for such a future is what we should be doing; while the philosophical issues are subtle here trying to prevent the future discussed above seems more it would make us not exist than give anything good to us. I was thinking maybe there’s some variant of this theory where it would be harmful for us to know about it but while it is conceivable such a variant exists I don’t see a plausible worldview where such a variant is more likely than one where knowing about it benefits us. I still thought I was really too much on my reasoning for subtle philosophical issues that may affect many other peopler, so I consulted with a few other people before publishing, and nobody else objected. I then procrastinated far longer than reasonable to make a definitive decision, and now here we are.

I know of two plausible physical models for post-selection. Both are simple and natural modifications of our current understanding of the laws of physics. The first is the two-state formalism for quantum mechanics. In addition to an initial state of a quantum system a final state is specified as well. For any observable , considered in the Heisenburg picture, its actual expectation value is . This doesn’t actually work as stated because these “expectation values” will have phase terms in every history that contributes to them and won’t be real numbers so there has to be a more nuanced interpretation of them, but I think this problem is solvable. The second model is making the dynamics slightly non-unitary. In a path-integral formulation this can be done by adding a small imaginary component to the Lagrangian. This adds a term to the path integral which can be thought of as a weighing factor that favors some histories over others, specifically favoring histories where is minimized throughout all of space. For small this won’t perceptibly change the probabilities of small-scale events but will direct the large-scale trajectory of the universe to minimize .

Postselection provides a potential solution to the cosmological constant problem, in that a low cosmological constant allows long-lasting stable structures that are more conducive to evolution, and that might more than make up for the improbability of fine-tuning a low cosmological constant. This is similar to Steven Weinberg’s anthropic explanation for the cosmological constant, but is different in two ways:

1. Postselection may require intelligently designed devices to fill all of space, and which requires all of space to be in the future lifecone of places where life evolves. This is a stricter criterion than requiring that intelligent life exists at all and may give smaller estimates for the cosmological constant,
   closer to the observed value.
2. Postselection is a purely physical mechanism and doesn’t depend on the controversial philosophy behind the use of the anthropic principle. I particularly don’t buy the use of the anthropic principle in Weinberg’s anthropic explanation but I can imagine other people objecting to some philosophical stance behind postselection, so this one is more subjective.

I think this is the strongest reason to think physical postselection is actually true and not just a intriguing possibility; in fact, in my personal opinion this theory is the best explanation for the low cosmological constant that I am aware of.

It may be worth thinking thinking about applying a postselection theory to explain the other unnatural physical constant, the Higgs mass. Current measurements of its value place the universe just at the tipping point of metastability, where its still not known whether the universe can undergo vacuum decay. This seems “teleological” in a way potentially amenable to an explanation involving postselection with some future vacuum decay.her the universe can undergo vacuum decay. This seems “teleological” in a way potentially amenable to an explanation involving postselection with some future vacuum decay.