extinctionethics

Moral philosophers have not written much about human extinction. This may be because they underestimate the potential benefits of human survival and/or the risks of human extinction. If we survive the next few centuries, humanity could allow Earth-originating life to survive a trillion years or more. If we do not survive, Earth-originating life will probably perish within a billion years. If prolonging the survival of Earth-originating life is morally important, then there may be nothing more important than reducing the near-term risks of human extinction.

Keywords: extinction, population ethics, intergenerational justice, catastrophic risk, existential risk, risk analysis, animal welfare, environmental ethics

It was only in the last century, with the invention of nuclear weapons, that the probability of human extinction could be appreciably affected by human action. Ever since, human extinction has generally been considered a terrible possibility. It’s surprising, then, that a search of JSTOR and the Philosopher’s Index suggests contemporary philosophers have written little about the ethics of human extinction. In fact, they seem to have written more about the extinction of other animals. Maybe this is because they consider human extinction impossible or inevitable; or maybe human extinction seems inconsequential compared to other moral issues.

In this paper I argue that the possibility of human extinction deserves more attention. While extinction events may be very improbable, their consequences are grave. Human extinction would not only condemn to non-existence all future human generations, it would also cut short the existence of all animal life, as natural events will eventually make Earth uninhabitable.

The value of future lives

Leslie (1996) suggests philosophers’ nonchalance toward human extinction is due in large part to disagreements in population ethics. Some people suppose it does not matter if the number of lives lived in the future is small -- at its limit, zero.[2]

In contrast, I assume here that moral value is a function of both the quality and number of lives in a history.[3] This view is consistent with most people’s intuition about extinction (that it’s bad) and with moral theories under which life is considered a benefit to those who have it, or under which life is a necessary condition for producing things of value (Broome, 2004; Hare, 1993; Holtug 2001, Ng, 1989; Parfit 1984; Sikora, 1978).

For instance, some moral theories value things like experiences, satisfied preferences, achievements, friendships, or virtuous acts, which take place only in lives. On this view, an early death is bad (at least in part) because it cuts short the number of these valuable things. Similarly, on this view, an early extinction is bad (at least in part) because it cuts short the number of these valuable things. I think this view is plausible and think our best reasons for believing an early death is bad are our best reasons for believing an early extinction is bad. But such a view is controversial and I will not settle the controversy here.

I start from the premise that we ought to increase moral value by increasing both the quality and number of lives throughout history. I also take it, following Singer (2002), this maxim applies to all sentient beings capable of positive subjective feelings.

Life’s prospects

The human population is now 6 billion (6 x 10⁹). There are perhaps another trillion (10¹²) sentient animals on Earth, maybe a few orders more, depending on where sentience begins and ends in the animal kingdom (Gaston, Blackburn, and Goldewijk, 2003; Gaston and Evans, 2004).

Animal life has existed on Earth for around 500 million years. Barring a dramatic intervention, all animal life on Earth will die in the next several billion years. Earth is located in a field of thousands of asteroids and comets. 65 million years ago, an asteroid 10 kilometers in size hit the Yucatan , creating clouds of dust and smoke that blocked sunlight for months, probably causing the extinction of 90% of animals, including dinosaurs. A 100 km impact, capable of extinguishing all animal life on Earth, is probable within a billion years (Morrison et al., 2002).

If an asteroid does not extinguish all animal life, the Sun will. In one billion years, the Sun will begin its Red Giant stage, increasing in size and temperature. Within six billion years, the Sun will have evaporated all of Earth’s water, and terrestrial temperatures will reach 1000 degrees -- much too hot for amino acid-based life to persist. If, somehow, life were to survive these changes, it will die in 7 billion years when the Sun forms a planetary nebula that irradiates Earth (Sackmann, Boothroyd, Kraemer, 1993; Ward and Brownlee, 2002).

Earth is a dangerous place and animal life here has dim prospects. If there are 10¹² sentient animals on Earth, only 10²¹ life-years remain. The only hope for terrestrial sentience surviving well beyond this limit is that some force will deflect large asteroids before they collide with Earth, giving sentients another billion or more years of life (Gritzner and Kahle, 2004); and/or terrestrial sentients will colonize other solar systems, giving sentients up to another 100 trillion years of life until all stars begin to stop shining (Adams and Laughlin, 1997). Life might survive even longer if it exploits non-stellar energy sources. But it is hard to imagine how life could survive beyond the decay of nuclear matter expected in 10³² to 10⁴¹ years (Adams and Laughlin, 1997). This may be the upper limit on the future of sentience.[4]

Deflecting asteroids and colonizing space could delay the extinction of Earth-originating sentience from 10⁹ to 10⁴¹ years. Assuming an average population of one trillion sentients is maintained (which is a conservative assumption under colonization[5]), these interventions would create between 10²¹ and 10⁵³ life-years.

At present on Earth, only a human civilization would be remotely capable of carrying out such projects. If humanity survives the next few centuries, it’s likely we will develop technologies needed for at least one of these projects. We may already possess the technologies needed to deflect asteroids (Gritzner and Kahle, 2004; Urias et al., 1996). And in the next few centuries, we’re likely to develop technologies that allow colonization. We will be strongly motivated by self-interest to colonize space, as asteroids and planets have valuable resources to mine, and as our survival ultimately requires relocating to another solar system (Kargel, 1994; Lewis, 1996).

Extinction risks

Being capable of preserving sentient life for another 10⁴¹ years makes human survival important. There may be nothing more important. If the human species is extinguished, all known sentience and certainly all Earth-originating sentience will be extinguished within a few billion years. We ought then pay more attention to what Bostrom (2002) has called “existential risks” -- risks “where an adverse outcome would either annihilate Earth-originating intelligent life or permanently and drastically curtail its potential.”

Such risks include: an asteroid or comet strikes Earth, creating enough debris to shut down photosynthesis for months; a supervolcano erupts, creating enough debris to shut down photosynthesis; a nearby supernova unleashes deadly radiation that reaches Earth; greenhouse gasses cause a radical change in climate; a nuclear holocaust creates enough debris to cause a “nuclear winter,” shutting down photosynthesis; a genetically engineered microbe is unleashed, by accident or design, killing most or all of humanity; or a high-energy physics experiment goes awry, creating a “true” vacuum or strangelets, destroying the Earth (Bostrom 2002; Bostrom and Cirkovic 2006; Leslie 1996, Posner 2004, Rees 2003).

To me, most of these risks seem very unlikely. But dishearteningly, in their catalogs of these risks, Britain ’s Astronomer Royal, Sir Martin Rees (2003), gives humanity 50-50 odds of surviving the next few centuries, and philosophers John Leslie (1996) and Nick Bostrom (2002) put our chances at 70% and 75%, respectively.

Estimating the probabilities of unprecedented events is subjective, so we should treat these numbers skeptically. Still, even if the probabilities are orders lower, because the stakes are high, it could be justified to invest in extinction countermeasures. Matheny (2007) found that, even with traditional social discounting, investing in asteroid detection and mitigation is justified under standard cost-effectiveness analysis.

Ought humanity be saved?

Even accepting that future lives have value and that extinction risks can be cost-effectively reduced, there could still be reasons not to worry about human extinction. For instance, human lives might have negative moral value, in which case human extinction could be a good thing. This might have been Bertrand Russell’s sentiment when he wrote, “Although it is a gloomy view to suppose that life will die out, sometimes when I contemplate the things that people do with their lives I think it is almost a consolation.”[6]

In the 20th century, more people, in absolute numbers, died of war, famine, and pestilence than ever before. But in the same century, more people did not die of war, famine, and pestilence than ever before. So even if we're especially pessimistic about average human welfare during the last century compared to others, it would be hard to argue that total welfare decreased. As long as average welfare was greater than zero – that is, the average life was preferable to suicide – then the century was a success for humanity. We will be capable of even greater moral nightmares in this century than in the last, but we will also be capable of securing greater welfare for a larger fraction of humanity. I suspect in this century, the average life will again be worth living, assuming we survive the century to judge.

We should be more pessimistic when we review how nonhuman animals have fared in the last century. At present around 50 billion animals are raised and killed each year to feed humanity. (Many million animals are used for clothing, product testing, research, and entertainment, but their numbers are insignificant by comparison.) Since World War 2, with the invention of "factory farming," farm animals’ welfare has significantly deteriorated, as they now live in conditions that frustrate their most basic instincts (Singer, 2002, chapter 3).

At the same time, we’re probably the only animal on Earth that routinely demonstrates compassion for other species. Such compassion is nearly universal in developed countries but we usually know too little, too late, for deeply ingrained habits, such as diets, to change. If improvements in other public morals were possible without any significant biological change in human nature, then the same should be true for our treatment of nonhuman animals, though it will take some time.

Even without any change in public morals, it seems unlikely we will continue to use animals for very long – at least, nowhere near 50 billion per year. Our most brutal use of animals results not from sadism but from old appetites now satisfied with inefficient technologies that have not fundamentally changed in 10,000 years. Ours is the first century where newer technologies -- plant or in vitro meats, or meat from brainless animals -- could satisfy human appetites for meat more efficiently and safely (Edelman et al, 2005). As these technologies mature and become cheaper, they will likely replace conventional meat. If the use of sentient animals survives much beyond this century, we should be very surprised.

This thought is a cure for misanthropy. As long as most humans in the future don't use sentient animals, the vast number of good lives we can create would outweigh any sins humanity has committed or is likely to commit. Even if it takes a century for animal farming to be replaced by vegetarianism (or in vitro meats or brainless farm animals), the century of factory farming would represent around 10¹² miserable life-years. That is one-billionth of the 10²¹ animal life-years humanity could save by protecting Earth from asteroids for a billion years.

The century of industrialized animal use would thus be the equivalent of a terrible pain that lasts one second in an otherwise happy 100-year life. To accept human extinction now would be like committing suicide to end an unpleasant itch. If human life is extinguished, all known animal life will be extinguished when the Sun enters its Red Giant phase, if not earlier. Despite its current mistreatment of other animals, humanity is the animal kingdom’s best long-term hope for survival.

Wishful thinking

There could be other reasons to not worry about human extinction. For instance, it might be hoped that if humanity is extinguished, another animal will evolve and develop the technologies needed to preserve terrestrial life.

This possibility should not give us much comfort. Some events, such as asteroid impacts, threaten all animal species, not just our own -- there may be no animal that survives to evolve. Moreover, the evolution of a civilization like ours may be a rare event. Another species that evolves to fill our intelligence niche could do as much environmental damage as humanity, and could have a worse ethical system. E.O. Wilson (1978, p. 104), describing species more violent than humans, remarked, “I suspect that if hamadryas baboons had nuclear weapons, they would destroy the world in a week. And alongside ants, which conduct assassinations, skirmishes, and pitched battles as routine business, men are all but tranquilized pacifists.” Alternatively, another intelligent species could have a better ethical system but too few ambitions to develop technologies essential to delaying extinction.

Given the (probably) improbable position we find ourselves in, as a species with both the technological potential and the motivation to delay extinction, it would be imprudent to trust another species will evolve and possess the same. Moreover, it would take many millions of years for such a species to evolve. During that time, Earth will be exposed to astronomical risks, and entropy will have consumed resources that could otherwise have supported sentient life (Bostrom, 2003).

The same can be said of sentient extraterrestrials, who, it might be hoped, exist or will exist. Even if we did not care about preserving terrestrial life, the absence of apparent extraterrestrial signals or colonies should worry us. The conditions necessary for supporting sentient life may be highly restrictive, making such life a rare phenomenon in the galaxy (Mayr, 1995; Ward and Brownlee, 2000). It is possible that humanity is at the leading edge of technology in our region of space. If so, we should not assume sentient life elsewhere will develop a civilization capable of space colonization. Extraterrestrials may be overwhelmed by their own evolutionary hurdles or extinction risks.

If we are not at the leading edge, and there are extraterrestrials more advanced than we are, then they are probably either not benevolent or not nearby. Otherwise, it is hard to fathom why they haven’t intervened (overtly or covertly) in Earth’s affairs to prevent the mass of suffering on this planet. The presence of preventable misery here militates against the existence of an advanced, benevolent, and nearby extraterrestrial, as it militates against the existence of a powerful and benevolent god.

If a benevolent, advanced extraterrestrial civilization exists, then it is probably distant. We would waste some fraction of remaining starlight if we waited for them to reach this region of space. We ought to prevent human extinction, then, if only to decrease the opportunity costs caused by entropy (Bostrom, 2003).

Conclusion

I believe that if we destroy mankind, as we now can, this outcome will be much worse than most people think. Compare three outcomes:

2 would be worse than 1, and 3 would be worse than 2. Which is the greater of these two differences? Most people believe that the greater difference is between 1 and 2. I believe that the difference between 2 and 3 is very much greater. . . . The Earth will remain habitable for at least another billion years. Civilization began only a few thousand years ago. If we do not destroy mankind, these thousand years may be only a tiny fraction of the whole of civilized human history. The difference between 2 and 3 may thus be the difference between this tiny fraction and all of the rest of this history. If we compare this possible history to a day, what has occurred so far is only a fraction of a second.

This paper supports Parfit’s conclusion. Human extinction would likely condemn all sentience of terrestrial origin to extinction. We take extraordinary measures to protect some endangered species from extinction. It would be reasonable to take extraordinary measures to protect humanity from the same. If we survive the next few centuries, we will probably survive long enough to colonize space and disperse, ensuring the survival of sentient life for perhaps trillions of years. The next few centuries could be the most critical in our past or future.

The moral weight of human extinction does not mean we can ignore other moral problems. There is no conflict between helping to delay human extinction and, for instance, boycotting animal farms or consuming fewer natural resources. We can do both. But when instances of conflict arise, as they do in cases of public funding, we ought to prioritize projects that reduce extinction risks. Our primary goal in the next few centuries should be to survive long enough to colonize space.

References

Adams FC, Laughlin G, 1997, A dying universe: the long-term fate and evolution of astrophysical objects, Reviews of Modern Physics 69(2): 337-370.

Barry B, 1977, Rawls on average and total utility: a comment, Philosophical Studies 31: 317-325.

Bostrom N, 2002, Existential Risks: Analyzing Human Extinction Scenarios and Related Hazards, Journal of Evolution and Technology, Vol. 9. http://www.nickbostrom.com/existential/risks.html

Bostrom N and Cirkovic M (eds.), 2006, Global Catastrophic Risk, Oxford University Press, Oxford.

Edelman, P.E., D.C. McFarland, V.A. Mironov, J.G. Matheny, “In vitro cultured meat production,” Tissue Engineering 11 (5/6), 2005.

Freese K, Kinney W, 2003, The ultimate fate of life in an accelerating universe, Physics Letters B 558: 1-8.

Freitas RA, Merkle RC, Kinematic Self-Replicating Machines, Landes Bioscience, Georgetown, TX , 2004, ch. 3.

Gritzner C, Kahle R, 2004, Mitigation technologies and their requirements, in Belton MJS et al (eds.) Mitigation of Hazardous Impacts Due to Asteroids and Comets, Cambridge UK: Cambridge University Press.

Hare RM. 1993. “Possible People,” Essays on Bioethics. Oxford : Clarendon Press, 67-83.

Highfield R, 2001, Colonies in space may be only hope, says Hawking, Daily Telegraph (UK), 16 October 2001.

Holtug N, 2001, On the value of coming into existence, Journal of Ethics 5: 361-384.

Kargel J, 1994, Metalliferous Asteroids as Potential Sources of Precious Metals, Journal of Geophysical Research 99 (E10): 21129-21141.

Leslie J, 1996, The End of the World: The Science and Ethics of Human Extinction. London : Routledge.

Matheny JG, 2006, Reducing the risk of human extinction, Risk Analysis, forthcoming.

Morrison D, A Harris, G Sommer, C Chapman, A, 2002, Dealing with the Impact Hazard, in Asteroids III, (W. F. Bottke Jr. et al., eds.), Univ. of Arizona, Tucson.

Ng, Yew-Kwang. 1989. What Should We Do About Future Generations? Economics and Philosophy 5: 235-251.

Posner R. 2004. Catastrophe: Risk and Response. New York: Oxford University Press.

Sackmann IJ, Boothroyd AI, Kraemer KE, 1993, Our Sun III: Present and Future, Astrophysical Journal 418, 457-468.

Sikora RI . 1978. “Is it wrong to prevent the existence of future generations?” in Obligations to Future Generations, RI Sikora and Brian Berry, ed. (Philadelphia: Temple University Press, 1978).

Schopenhauer A, World as Will and Representation v. II, New York: Dover, 1958 [1819], p. 605.

Ward PD, Brownlee D, 2002, The Life and Death of Planet Earth, New York : Henry Holt.

[1] The author is grateful to John Broome, John Leslie, Richard Posner, Nick Bostrom, Peter Singer, Jonathan Wiener, David Bishai, Matthew Ball, Anjula Razdan, Robin Hanson, Les Knight, and Tyler Cowen for comments on an earlier draft.

[2] See, for instance, “Sui genocide”, The Economist, 349 (8099): 130 (Dec 19, 1998) and Jan Narveson’s “Utilitarianism and New Generations,” Mind 76 (1967): 62-72.

[3] This view needn’t imply Parfit’s (1984) “Repugnant Conclusion,” since the allowable tradeoffs of quality for number could be minimal – even zero. Here I remain agnostic about whether we ought to bring into existence those people Hare (1993) calls “merely possible,” or whether we only ought to ensure the existence of people who will certainly exist if humanity is not extinguished. As Parfit (1984) has pointed out in his Non-Identity Problem, the distinction between these two classes of people is unclear, as policies preventing extinction would affect the timings of some births and the identities of some future persons. But the number of all “merely possible” persons is much larger than the number of persons who will certainly exist if humanity delays its extinction, and here I limit discussion to the latter.

[4] However, some cosmologies might allow life to exist indefinitely (Freese and Kinney, 2003).

[5] At a population doubling time of 1000 years, a group of 1000 colonists could create 10¹² colonists in 34,000 years. This number could be much larger. There are around 10¹³ stars in the Virgo Supercluster and, assuming each star could provide enough energy to sustain 10¹⁰ sentients living on nearby planets, asteroids, or artificial satellites, our Supercluster could support 10²³ sentients. More ambitious colonization strategies could lead to even larger populations (Bostrom, 2003).

[6] Schopenhauer (1958 [1819]) was even more pessimistic: “existence is certainly to be regarded as an error or mistake." Similarly, “negative utilitarians” would be glad to see all life extinguished (Smart, 1958).

Ought we worry about human extinction? [1]