Figure 1. Ethical obligations related to preventing human deaths.

Of particular interest is Point B which represents a massive population loss of hundreds of millions of humans or more. The ethical standard is approximately 10-¹³.

It is interesting to compare the thresholds represented by the green line to various published risks. For example, it has been estimated that the risk of a major incident involving the tear in the space time fabric associated with experiments being run at Brookhaven National Laboratory ranges from 2x10^-6 (slow destruction) to 2x10^-8 (fast destruction) to 1x10^-5 to 2x10^-11 (Busza, Faffe, Sandweiss and Wilczek, 2000; Dar, Rujula and Heintz, 1999). Within this range of risks, any incident would need to cause the death of fewer than one person (10^-5) to one million people (10^-11) so as not to exceed ethical thresholds set out in Figure 1. In reasoning similar to the Fairness Criterion above and essentially consistent with Figure 1, Kent (2004) suggests that an acceptable risk of killing not only the current population but succeeding future populations due to an catastrophic high energy physics experiment should be less than 10^-22 per year.

According to NASA, the risk of earth being hit by any known and tracked near-earth object that could cause a major loss of life over the next 100 years seems to be very small. The only NEO that appears to be capable of causing a major loss of life is 2004 BX159, which is 1.2 km in diameter and which has a probability of 2.6 x 10-⁹ of hitting the earth during the next 100 years. If the number of possible deaths exceeds several thousand, then this event by itself could exceed the risk threshold set out in Figure 1 and would warrant efforts to reduce its probability of impact. It is also possible that the cumulative risks of asteroid impacts exceed ethical thresholds but these numbers have not been calculated.(see http://neo.jpl.nasa.gov/risk/).

It must be emphasized that Figure 1 contains a very disparate range of risks, from risks to individual humans to those that threaten the entire human race. To accommodate this range, it must be acknowledged that the scope of responsibility for meeting ethical risk thresholds changes as the risks move from the individual to the collective. It is proposed here that risks impacting tens to thousands of individuals be handled by communities (e.g., cities, counties); risks impacting up to millions be handled by regions (e.g., multiple provinces, multiple countries in a region), and that risks impacting larger populations be handled through global initiatives.

The red line in Figure 1 represents the author’s judgment of the current situation facing humanity. It is certain that there are people alive today that will die pre-maturely from environmental risks imposed upon them involuntarily from the same risks that threaten other species. At this point, it needs to be stated that war needs to be incorporated into the set of risks that threaten massive population loss and human extinction.

In fact, the red line suggests that it is certain that close to 100 million people will die prematurely over the next 1000 years. There is small hope that the pre-mature deaths will not exceed a billion. The only bit of optimism expressed by the red line is that it is very unlikely that humans will become extinct in the next 1000 years. This is more by luck than design because humans are adaptive, creative and flexible, leaving only a few singular and highly unlikely paths open to extinction (Tonn and MacGregor 2009). In totality, though, it can be argued that ethical thresholds are being exceeded at all points from A to D except the most extreme.

This judgment represented by the red line in Figure 1 is at odds with several other published estimates of the probability of human extinction. For example, Rees (2003) estimates the risk of human extinction to be at least one in two over the next one hundred years. Leslie (1996) puts the risk at 30% during this time period. Bostrom (2002) estimates that the risk is at least 25%. The Stern Review conducted for the United Kingdom Treasury assumes probability of human extinction during next century is 10% (United Kingdom Treasury 2006).

Finally, the yellow line represents an intermediate solution between the highly unacceptable red line and the acceptable green line. This solution removes catastrophic risks from almost all potential human futures through global cooperation and moves towards moving responsibility for remaining risks to regions and then eventually to communities.

III. OBLIGATIONS TO OTHER SPECIES

The purpose of this section is to generalize the framework presented in the previous section to massive species extinction. Figure 2 presents graphically an ethical framework to use when developing policies to prevent species extinction. The approaches to human risk and species risk are fundamentally the same, although the details differ substantially. Both approaches begin with a de facto standard, end with an extinction standard, and interpolate in between. One difference is that there is no explicit de facto standard related to an ethical standard to prevent the extinction of an individual species. Interpretations of the consequences of implementation of the Endangered Species Act in the United States are that the risk to a species needs to become quite substantial before a species is listed as endangered or threatened. In this regard it needs to be pointed out that animals and major plants species are addressed by the ESA. Insects and the various micro-organisms are not.

Figure 2. Ethical obligations related to preventing massive species extinction.

Therefore, in regard to plants and animals, it could be argued that the risk may rise to be as high as one chance in ten. Then plans are put in place to reduce the danger of extinction and species are delisted when the risk appears to be several orders of magnitude lower. Allowing that the second viewpoint is most defensible from an ethical framework, one could argue that the green line begins at Point A at a risk standard of about 10-4 for an individual species.

Subjectively, this makes the value of one human life more valuable than that of an entire species. This may or may not be defensible ethically. Another philosophical question one could raise here is whether all species ought to be treated equally.

At the end of the green line in Figure 2, at Point D, is the acceptable risk of the extinction of all life on earth. This standard, 10-²³, is three orders of magnitude stricter than is the standard for human extinction. One reason for this is that humans would surely become extinct somewhat before all life itself became extinct. Thus, in order to reduce the threat of human extinction, one such path to extinction, massive extinction of other species, must be kept at least several orders of magnitude lower. Additional philosophical considerations could push this standard even lower.

The green line in Figure 2 is interpolated between these two points, as in Figure 1. In this case, the drop off to the acceptable risk of all species extinction is a bit less dramatic than is the dropoff in Figure 1. The acceptable risk of massive species extinction, defined in the graph as existing near Point B is approximately 10-¹⁶. As in Figure 1, unacceptability increases as the points move up and to the right above the green line. Also, similarly, the worst point is Point E, which represents close to certain extinction of all life within the planning horizon.

Figure 2 also contains a red line which representsthe present with respect to species extinction. It is certain that tens of thousands of species will become extinct during the next 1000 years. There are few signs that the situation will turn around any time soon, leaving the probability high that hundreds of thousands of species will also become extinct. As with the story about human survival, there is only a very remote chance that all life on earth will become extinct, when one considers that life exists in the deepest depths of the oceans, in the cauldrons of geysers, in soils, and in pools of water deep underground. Again, this is by luck, not policy. So, in summary, it can be argued that we are not close to meeting our ethical responsibilities to other species.

IV. CONCLUSIONS

This paper provides a framework to view ethical obligations to other species generally and a way to understand ethical thresholds related to preventing massive species extinction. However, this framework should only be considered as an initial effort in this area.

Much additional thought is needed to ground the ethical thresholds related to both involuntary, pre-mature human mortality from environmentally-based risks and massive species extinction. As mentioned above, philosophical obligations to future generations were used to developed thresholds related to human extinction but reasonable people could disagree with the approaches and assumptions required to operationalize the approaches. The theoretical grounding for the ethical standards related to species extinction is even less secure. Much work is needed to sharpen and defend the green line in Figure 2.

Despite these qualifications, the outline of the framework seems sound. There should be some acceptable risk associated with the pre-mature death of a human due to an involuntary environmental risk as well as associated with the extinction of an specie. To continue, there should be some ultimate acceptable risk standards associated with human extinction and extinction of all life on earth, respectively. Whether the interpolation between these two points follows the essentially linear functions found in Figures 1 and 2, until one reaches population extinctions is sound is another topic for further study, but the assertion that the extinction risk standard must be much stricter than the individual/species standard seems to be reasonable. It can be argued that we do not currently have the ‘technology’ to estimate the probabilities over a 1000 year time span required to instantiate Figures 1 and 2. It could be that a set of approaches are needed to estimate probabilities appropriate to the scale of the risk. For example, known approaches to risk analysis could still be used to determine risks to individuals and small numbers of people that would be under the responsibility of communities to deal with. In some instances, historical data might be available to estimate risks to larger numbers of people (e.g., hurricanes, earthquakes) within a regional purview. Similar approaches can be taken at these scales with respect to species extinction.

Beyond these two areas, probabilities would need to be estimated using subjective means. One could do this holistically using Bayesian methods or constructed using evidential approaches (see Tversky and Shafer 1985). Another suggested approach to estimating probabilities of risks to massive populations up to extinction would be to model as many paths into the future as possible. As the time horizon extends, the number of paths would grow exponentially but there are powerful brute force search algorithms (see Deep Blue, IBM’s system that used brute force computing technology to defeat the reigning world chess champion) that could tirelessly work through vast numbers of future paths into the future. If this could be done, then the probability of human extinction, for example, would simply be the number of paths that lead to extinction over all possible paths into the future. Then this probability could be compared to the ethical standard, as presented in Figure 1. One cannot underestimate the modeling challenge associated with modeling the future, when human behavior, the invention and adoption of new technologies, and a plethora of other complicating factors are considered over the 1000-year period. This could be one of those 1000-year challenges, maybe too daunting to undertake if one requires results next year but surely achievable to an extent needed for policy purposes over a longer time frame. One could argue that this challenge could be a logical next challenge for an organization such as the Intergovernmental Panel on Climate Change.

Despite the lack of technology to estimate probabilities, it can be strongly argued that we are not meeting our ethical obligations to either our future generations or to other species. We know that individual humans and individual species will die pre-maturely or perish, respectively, without any probabilistic analysis. Most of us feel that the probabilities of massive loss of human life and extinction of species are unacceptably high. We also know that the likelihood of human extinction and complete loss of life is very low but this is not a situation we can take any credit for. The technologies are not needed, really, to tell us where we are at, but to help us shape policies to move the red lines in Figures 1 and 2 down towards and someday coincident with the green lines.