Saturday, November 30, 2019

Person-affecting views may be dominated by possibilities of large future populations of necessary people

On symmetric person-affecting views (PAV), in choosing between actions only necessary people, those who will exist regardless of our immediate choice, count while those whose existence is contingent on our choices are ignored. Chaotic 'butterfly effects' mean that almost any terrestrial action will shortly scramble the genetic makeup and identities of future terrestrial births. Thus philosopher Michael Plant invokes symmetric PAV to say that '[o]n this, roughly speaking, the well-being of future entities, human or non-human, does not count" and to argue in favor of a focus on short-run impacts on long-lived existing creatures (humans and dogs, but not chickens or ants) rather than long-run consequences of altruistic interventions, since vast future populations would be composed of contingent rather than necessary people. However, the empirical underpinnings of this move are questionable: there are reasonable possibilities on which astronomically large populations of necessary people will exist, and credence in such possibilities will mean that they account for the bulk of expected impact on necessary people. I raise three such possibilities: technological life extension, contact with distant aliens, and fission cases.

Life extension increases longtermist stakes for PAV

Common arguments for long-termism invoke the possibility of vast future populations and welfare, among other reasons:

  •  Large future potential populations
    • Orders of magnitude energy reaching Earth from the Sun than used by our civilization
    • ~billionfold greater energy emitted by the Sun than reaches Earth
    • Hundreds of billions of stars in the Milky Way
    • Billions of galaxies reachable at the speed of light, hundreds of millions at lower speeds
    • Potentially much greater energy efficiency in supporting populations or welfare
  • Long durations
    • Billions of years of starlight on average, hundreds of billions or more for the longer-lived stores
    • Longer periods potentially using other methods (such as feeding black holes) to reap more energy
  • Potentially high density of welfare
    • Advanced technology permitting more efficient creation of desired mental states and other properties
    • Potentially very high densities of minds
For PAV, an important additional possibility is that of technological life extension, so that necessary beings can be extended to take up more of the above factors. If advanced technology enables life extension, then existing beings may be impacted for very long durations, or with extreme welfare (e.g. by using increased computation to accelerate the speed of thought, living more subjective life-years per calendar year). Section IV of Bostrom's paper Astronomical Waste discusses implications for longtermism and PAV:

Suppose instead that we adopt a “person-affecting” version of utilitarianism, according to which our obligations are primarily towards currently existing persons and to those persons that will come to exist.[9] On such a person-affecting view, human extinction would be bad only because it makes past or ongoing lives worse, not because it constitutes a loss of potential worthwhile lives. What ought someone who embraces this doctrine do?...Suppose one thinks that the probability is negligible that any existing person will survive long enough to get to use a significant portion of the accessible astronomical resources, which, as described in opening section of this paper, are gradually going to waste. Then one’s reason for minimizing existential risk is that sudden extinction would off cut an average of, say, 40 years from each of the current (six billion or so) human lives.[10]... 
Arguably, however, we ought to assign a non-negligible probability to some current people surviving long enough to reap the benefits of a cosmic diaspora. A so-called technological “singularity” might occur in our natural lifetime[11], or there could be a breakthrough in life-extension...Even if you are skeptical about their prognostications, you should consider the poor track record of technological forecasting. In view of the well-established unreliability of many such forecasts, it would seem unwarranted to be so confident in one’s prediction that the requisite breakthroughs will not occur in our time as to give the hypothesis that they will a probability of less than, say, 1%. 
The expected utility of a 1% chance of realizing an astronomically large good could still be astronomical. But just how good would it be for (some substantial subset of) currently living people to get access to astronomical amounts of resources?...Bill Gates' level of well-being does not seem to dramatically exceed that of many a person of much more modest means. On the other hand, advanced technologies of the sorts that would most likely be deployed by the time we could colonize the local supercluster may well provide new ways of converting resources into well-being. In particular, material resources could be used to greatly expand our mental capacities and to indefinitely prolong our subjective lifespan...A long shot it may be, but for an expected utility maximizer, the benefit of living for perhaps billions of subjective years with greatly expanded capacities under fantastically favorable conditions could more than make up for the remote prospects of success. 
Now, if these assumptions are made, what follows about how a person-affecting utilitarian should act? Clearly, avoiding existential calamities is important, not just because it would truncate the natural lifespan of six billion or so people, but also – and given the assumptions this is an even weightier consideration – because it would extinguish the chance that current people have of reaping the enormous benefits of eventual colonization. However, by contrast to the total utilitarian, the person-affecting utilitarian would have to balance this goal with another equally important desideratum, namely that of maximizing the chances of current people surviving to benefit from the colonization.
For a total utilitarian symmetric PAV view like some of those discussed by Plant in Doing Good Badly, these considerations do appear to blow up the importance of the longterm by orders of magnitude given a significant chance of technological revolution and radical life extension within the lifetime of many existing terrestrial beings, particularly if we consider the possibility of uplifting long-lived nonhuman animals (or things like cryogenically preserving frogs, insects, and other more easily preserved creatures) and the possibility of a near-term technological revolution.

Distant extraterrestrial civilizations

On standard accounts of physics, there is no way for the influence of our actions on Earth to reach distant parts of the universe faster than the speed of light. Since our galaxy is on the order of 100,000 light-years across, and the most distant galaxies that can apparently be reached (due to the accelerating expansion of the universe which means sufficiently distant galaxies recede faster than we can reach them) are in the billions of light-years away. Thus it would appear that any causal impact on civilizations in distant parts of the galaxy and beyond would lie thousands or millions of years into the future, i.e. they would most definitely be long term consequences of our actions.

As our causal influence expands at the speed of light the surface area of the sphere of causal influence will grow, as will the volume within a few light-decades or centuries of the surface, to the point that after billions of years one might expect an average of hundreds of billions of stars within a century of contact. If radiation emitted by our everyday activities is sufficient to scramble chaotic events at such distances, then the surface of the sphere would mark the end of the production of new necessary (with respect to us now) beings, and begin a 'clock' for impact on the stocks of necessary beings contacted by the sphere.

If at least some non-negligible minority of alien civilizations eventually use technology to greatly extend lifespans at scale, then vast populations could survive for centuries, or far longer. This would provide ample time for Earth-derived civilization to technologically advance and send out communication signals or a fast colonization wave. With shorter lifespans, humanity would need to advance faster to send out active impacts, but such advance might nonetheless occur.

On this story, influencing the development of human civilization in such a way as to benefit the necessary aliens it will eventually influence has astronomical instrumental importance, but a potentially enormous Earth-derived civilization of contingent descendants is a means to an end from the current perspective, as are contingent aliens subjected to non-identity effects after causal contact is made. The total stakes would also be lower than on a non-PAV view, since it would seem in expectation most impacts will be on contingent rather than necessary beings.

There are a variety of empirical uncertainties here, from the possibility of new physics with faster-than-light causal contact (as in the Bohmian interpretation of quantum mechanics), to the existence of alien life, and the statistics for use of life extension technology by other civilizations. Nonetheless, the combination of standard physics and the distribution of views in exobiology makes me think this possibility should be considered by a PAV-consequentialist.

Creating vast populations of necessary people through fission

An alternative with a relatively greater proportion of philosophical uncertainty to empirical uncertainty is fission. From the SEP on personal identity:

A more serious worry for psychological-continuity views is that you could be psychologically continuous with two past or future people at once. If your cerebrum—the upper part of the brain largely responsible for mental features—were transplanted, the recipient would be psychologically continuous with you by anyone’s lights (even if there would also be important psychological differences). Any psychological-continuity view will imply that she would be you. If we destroyed one of your cerebral hemispheres, the resulting being would also be psychologically continuous with you. (Hemispherectomy—even the removal of the left hemisphere, which controls speech—is considered a drastic but acceptable treatment for otherwise-inoperable brain tumors: see Rigterink 1980.) What if we did both at once, destroying one hemisphere and transplanting the other? Then too, the one who got the transplanted hemisphere would be psychologically continuous with you, and would be you according to the psychological-continuity view.
But now suppose that both hemispheres are transplanted, each into a different empty head. (We needn’t pretend that the hemispheres are exactly alike.) The two recipients—call them Lefty and Righty—will each be psychologically continuous with you. The psychological-continuity view as we have stated it implies that any future being who is psychologically continuous with you must be you. It follows that you are Lefty and also that you are Righty. But that cannot be: if you and Lefty are one and you and Righty are one, Lefty and Righty cannot be two. And yet they are: there are indisputably two people after the operation. One thing cannot be numerically identical with two things that are distinct from each other. Suppose Lefty is hungry at a time when Righty isn’t. If you are Lefty, you are hungry at that time. If you are Righty, you aren’t. If you are Lefty and Righty, you are both hungry and not hungry at once: a straight contradiction.

More plausible technological implementations of fission might be preceded by gradual technological augmentation and replacement of parts of the brain, or use destructive methods of scanning to create a psychological continuant. See Chalmers for a recent discussion of such 'mind uploading.'

If a PAV is cashed out to consider both minds produced by fission as necessary people in light of their psychological/physical continuity with the previously existing person, then it would permit the indefinite expansion of the necessary population, and all the usual arguments for longtermism could go through with PAV, provided minds from the existing population turn out to be widely used as seeds for that process.

This would correspond with multiple-occupancy views:

Psychological-continuity theorists have proposed two different solutions to this problem. One, sometimes called the “multiple-occupancy view”, says that if there is fission in your future, then there are two of you, so to speak, even now. What we think of as you is really two people, who are now exactly similar and located in the same place, doing the same things and thinking the same thoughts. The surgeons merely separate them (Lewis 1976, Noonan 2003: 139–42; Perry 1972 offers a more complex variant)...
The multiple-occupancy view is usually combined with the general metaphysical claim that people and other persisting things are composed of temporal parts (often called “four-dimensionalism”; see Heller 1990: ch. 1, Hudson 2001, Sider 2001a, Olson 2007: ch. 5). For each person, there is such a thing as her first half: an entity just like the person only briefer, like the first half of a meeting. On this account, the multiple-occupancy view is that Lefty and Righty coincide before the operation by sharing their pre-operative temporal parts or “stages”, and diverge later by having different temporal parts located afterwards. They are like two roads that coincide for a stretch and then fork, sharing some of their spatial parts but not others. At the places where the roads overlap, they are just like one road. Likewise, the idea goes, at the times before the operation when Lefty and Righty share their temporal parts, they are just like one person. Even they themselves can’t tell that they are two. Whether we really are composed of temporal parts, however, is disputed. (Its consequences are explored further in section 8.)

However, on the alternative view that fission is death, PAV conclusions would be little affected.
The other solution to the fission problem abandons the intuitive claim that psychological continuity by itself suffices for us to persist. It says, rather, that a past or future being is you only if she is then psychologically continuous with you and no other being is. (There is no circularity in this. We need not know the answer to the persistence question in order to know how many people there are at any one time; that comes under the population question.) This means that neither Lefty nor Righty is you. They both come into existence when your cerebrum is divided. If both your cerebral hemispheres are transplanted, you cease to exist—though you would survive if only one were transplanted and the other destroyed. Fission is death. (Shoemaker 1984: 85, Parfit 1984: 207; 2012: 6f., Unger 1990: 265, Garrett 1998: ch. 4).

No comments: