Introduction

In A Duty to Cognitively Enhance Animals, Yasha Rohwer argues that we humans have a duty to promote biodiversity. In some cases, it may be preferable to do so by increasing the cognitive abilities of a species at risk of extinction to prevent that species from going extinct. It might be preferable (albeit not perfect nor a strict improvement with no costs) to eliminate the “bio” part altogether and emulate animal (and human) brains. While some might disagree, I think ems would count as biodiversity in some meaningful way.

Our Duty to Biodiversity

Rohwer argues that we humans have a duty to preserve biodiversity. While there are many definitions of biodiversity, Rohwer uses a conception of biodiversity from DeLong 1996 that includes organisms and their relationships¹ (e.g. the symbiotic relationship between myself and the bacteria in my gut) and functions within the context of a greater ecosystem with a focus on long-run biodiversity, so invasive species are bad because they will reduce biodiversity in the long-run by eliminating likely more than one local species. Rohwer and others claim that biodiversity has inherent value rather than only instrumental value² i.e. we should try to preserve biodiversity because it valuable in and of itself. Inherent value is tricky and I don’t really buy it; I think there’s always some end (other than “to behave morally”), and acting as if biodiversity has inherent value will probably be a more useful way of thinking once we’ve solved other problems like, say, aging. Luckily, biodiversity likely has many different reasons for being instrumentally valuable to humans e.g., the aesthetic pleasure it brings us, increased potential for scientific advances, etc. Insofar as we have obligations to do morally good things at all, we have obligations to promote biodiversity because we are often complicit in its reduction. Rohwer’s arguments rest on obligations to species (as a sort of proxy measure for biodiversity) which means sometimes harming individuals can be an acceptable tradeoff. This could mean killing an invasive species to preserve the other species on an island or capturing and holding captive the last members of a species in a zoo. We should, according to Rohwer, avoid making more animals suffer than we need to, so if we can preserve biodiversity without killing, then we should avoid killing. Generally, we should seek actions that maximize biodiversity while minimizing the suffering of animals. These are reasonable conclusions to draw with an end of biodiversity.

Why Make Animals Smarter?

Cognitive enhancement is not the only way to prevent a species from going extinct, and it seems like a rather roundabout way of doing so, but in the case of Australian rufous hare-wallabies, it would be rather effective. In Rohwer’s words “Lack of sophisticated cognition is indeed a hallmark of Australian marsupials.” They are unable to recognize the introduced red fox as a threat, and therefore are easy prey for it. Frankly, it’s also a much more interesting solution to biodiversity. Rohwer does address alternatives from habitat restoration and breeding programs to zoos and sanctuaries, but there could be issues even there.³

It’s Cool and Ethical!

Cognitive enhancement is an interesting solution because the focus shifts from harming the invasive species or restoring the habitat (likely to the short-run cost of the humans, potentially to their long-run cost) to helping the endangered species adapt to its new environment, becoming better fit to the new world in which it finds itself. Ethically, it’s a relatively better way to go about promoting biodiversity with likely only short-term costs and long-run benefits if done properly. There are lots of ways Rohwer mentions that we can enhance animals, from training to drugs or even genetic modification Cognitively enhancing animals probably has at least a few positive externalities.

Positive Externalities

If we’re really successful, we might get more species to “human-level” cognition which could mean not just more numerous and varied interactions between that species and other animals (including humans) but also interactions among members of that species. While this might essentially just increase the number of humans, we can do better. If we can increase the cognitive abilities of other animals, we might then be able to increase cognitive abilities of humans and animals with human-level cognitive abilities (hereon: AHCs [pronounced “ACK”]). I don’t think this is very likely without large changes in social attitudes and technology, but if we can do it, we might experience quite a bit of change: as de Menard points out, “smart people produce benefits for everyone else in the form of greater innovation, faster growth, and better governance.” If there’s a relationship between cognition and Dunbar’s number, we might increase Dunbar’s number for AHCs (and humans), leading to more, stronger, and perhaps more varied relationships with each other. In turn, this would mean better coordination between AHCs, which is a community-oritented lense through which we may achieve the results de Menard mentions. This increase in number and richness of relationships also adds to biodiversity (our original goal)!

Objections

Most objections Rohwer anticipates succumb to the naturalistic fallacy or are uninteresting, but I repeat myself.⁴

Not Just Smarter

What if we could modify animals such that their survival no longer depended on successfully hunting, scavenging, or foraging for food (or if they’re AHCs, such that they are rich enough to purchase food); such that they could have more direct control over their bodies, willingly and cheaply modifying how it appears to others; such that rather than succumb to an unexpected death, they could ease into a slow retirement?

Emulations

The excellent Robin Hanson wrote The Age of Em, forecasting what the Earth would be like when emulated human brains are possible and cheap enough to use widely.⁵ He calls these emulations “ems” for short. If human brains could be emulated, then we could plausibly emulate animal and AHC brains. Emulated life is preferable in some ways to physical, biological life.

In the virtual world, “older ems suffer little or no health issues or other physical decline.” Hanson estimates “if the rate of subjective aging for ems is only 10% of that of humans today, that would plausibly move the age of peak work productivity from roughly 50 to 500 subjective years,” suggesting an increased lifespan of at least as many subjective years.⁶ Because of the ease with which ems can be copied, ems would live at subsistence or “Malthusian” levels. While this might be distasteful to modern readers, these wage levels are the case for many humans today, most humans from the past, and nearly all animals that have existed. While controversial in the late 20th and early 21st centuries, ems might wirehead themselves into greatly enjoying the lives they live, e.g., ems might be able to modify their memories to retain only pleasant memories of recent vacations if they are useful for advancing their goals while working. The em world will be much more competitive and therefore productive than our world. There can also be many more ems than physical beings. Em cities might have populations of up to trillions of ems.

In a pure utilitarian sense, ems not only experience more utility⁷ than modern humans per unit time, but might also experience it over a greater time period and there might be thousands of times more ems than there could be humans on Earth.⁸

If we cared about the cultivation of virtue, we might prefer ems even more highly as Hanson writes:

What if we look instead at the virtues that have been admired in most eras and cultures, such as intelligence, insight, benevolence, loyalty, determination, etc.? Here the em world can look very good. Ems are strongly selected for their impressive productivity, which tends to correlate with most of these virtues. In this sense, the em world is packed full of people who are more virtuous than most people so far have ever met in a lifetime.

Biodiversity

If we accept that these ems count as life just as much as biochemical with which we’re so familiar,⁹ and if we can add ems without much cost to biological life, we might have an obligation to do so. We can create new kinds of beings who rather enjoy their lives and feel relatively little harm. Ems would have rich interactions with each other that would probably become only richer faster as time goes on since they would probably run at faster-than-human speeds. Ems could also have rich interactions with humans and the rest of the environment.

However, the increased number and productivity of ems may cost us some biodiversity. Whether ems contribute to biodiversity depends on the definition we use. If we use DeLong’s definition, ems might not count since they have no genes and most might not interact directly with the physical world, but the definition requires little tweaking to fit ems. Rather than genes we can consider and em’s ancestry. Perhaps we can consider species if we are able to emulate AHC and other animal brains but that might be a merely nominal difference. In place of biomass, em mass could be measured by the mass of the hardware used to run ems and em physical bodies. The point is for most biological concepts relating to biodiversity, we can probably find an equivalent that maps well to ems.

Indeed counting ems for biodiversity is important because they might take over the world!

While charities may pay to keep a few small nature preserves, the resources available for such efforts are likely be tiny compared with the economic pressures to use available resources to support the growing economy.

There may yet be other bad news when it comes to ems and diversity.

Em Diversity

With the ability to cheaply make copies and an extremely competitive economy¹⁰, we might not need as many truly unique minds as we do today.

A thousand diverse able scanned humans seems sufficient to induce competition in most labor markets, as the best few ems can dominate each labor market. Thus it seems likely that most ems are copies of fewer than a thousand or so, and perhaps only dozens, of the original humans. These few highly copied em “clans” of copies might be known by a single name, as are celebrities today such as Madonna or Beyoncé.

[…] most ems may come from a dozen to a thousand one-name clans, each of which has millions to billions of members. In contrast, there may be millions of two-name clans, each with only thousands to millions of members. And most of the roughly 10 billion ordinary humans may each give rise to a three-name clan with perhaps only dozens of members.

With only hundreds of main clans, em labor markets become concentrated like today’s product markets, with a few clans supplying most workers in any given skill area.

Relative to today’s number of humans, there might be a similar number of “unique” minds, that is em descendants of a particular human; however, as Hanson says, the vast majority of ems might be descended from just dozens to thousands of humans.

In the em world, another important metric is em speed. Faster ems are more expensive to run. In the scenario of voting, Hanson mentions that speed-weighted voting with votes proportional to em brain speeds could be useful to disincentivize election-time speed-ups or copying. For the same reasons one might use biomass rather than solely counting organisms to try to measure biodiversity today, one might use speed-weighted em counting to measure em biodiversity.

Conclusion

If we have a duty to promote biodiversity, and if we can (moderately) expand our definition of biodiversity to include artificial life, we might have a duty to promote emulated minds since they would add so much more utility and could add to the biodiversity of our world. If we don’t value ems much for biodiversity, then we’re out of luck.

Notes