The evolution of technological preference
How technology is advanced by the mechanism involved in Fisherian runaway
What is beauty? This question has fascinated Man since antiquity. Plato formulated beauty as a fundamental quantity that transcends the physical world; many theologists went on to argue that beauty is derived from God; and enlightenment thinkers began to posit that beauty is related to the harmonious arrangement of the parts. In all of these theories, beauty is conceptualized as static, metaphysical, and inexplicable by the means of scientific reasoning. This brief historical picture of the notion of beauty highlights the ingenuity of Ronald Fisher’s theory that the perception of beauty in sexual selection is a dynamical system whose existence can be justified by science and reason.
Charles Darwin’s 1859 work, On the Origin of Species represented an important paradigm shift, marking the inception of evolutionary biology. This landmark work sparked much discussion about the role of sexual reproduction in the evolution of a species. Interestingly enough, this paradigm shift led to an evolutionary explanation of why we and certain animals find certain aspects of the world beautiful.
At the start of the twentieth century, Fisher introduced a revolutionary framework for thinking about sexual reproduction. In his paper, “The Evolution of Sexual Preference“, he explains the evolution of sexual ornamentations such as fine plumage on certain bird species and the complicated “love dances” performed by the birds of paradise found in New Guinea. Fisher’s innovative idea was that the preferences held by the opposite sex for these sexual ornamentations are developed and enhanced through sexual reproduction in the same way that the ornaments themselves are passed to offspring. This coevolution of the ornamentation in one sex and the sexual preference for the ornamentation in the other sex creates a positive feedback system that eventually results in exacerbated sexual ornamentations. Often, these ornamentations do not improve the survivability of the species as a whole, but only the likelihood of the individual possessing the traits. In some cases, the survivability of the species is reduced due to worsened mobility or increased visibility to predators. These natural selective forces counteract the coevolution of the ornaments and the sexual preferences until an equilibrium is reached in which the disadvantage imposed by natural selective pressure is balanced by the advantage in sexual selection.
The positive feedback system described by Fisher is referred to as Fisherian runaway, and is used to explain many of the secondary sexual characteristics of several species, including humans1. On a systems level, the mechanism behind Fisherian runaway is one of the many examples of how a system may pursue a sub-goal (in this case, sexual reproduction) at the possible expense of an overarching goal (the survival of the species).
This article explains how the mechanism of Fisherian runaway is at play in the evolution of technology. To clarify, I am not referring to the role of technologies like makeup and luxury watches in the sexual ornamentation of humans, but rather how technology coevolves with human preference for the comfort provided by technology. By showing that technological progress is in part driven by the mechanism of Fisherian runaway, I show that efforts to advance technology have the potential to be detrimental for the survivability of humans.
Positive feedback mechanisms at play
The principle of Fisherian runaway applied to technological progress follows immediately from the premise that the inception and use of new technology remains both possible and desirable for at least some individuals.
To see that the premise is true, remark that by our working definition, technology is intentionally created; intentional creation implies the possibility and desire to create. Since technology is still being developed and used today, and has been for the last 10000 years, it will most likely continue in the future as long as there remains a human society.
Under the premise that there are some people who want to (for any reason) and can create a new technology, they will create it. If the use of that technology is possible and desirable for some, then the technological landscape evolves, and by the initial premise, there is a new desire for new technology. The positive feedback cycle continues.
Before we take a detailed look at the various drivers that might contribute to the inception and use of a technology, let us consider reasons why more technology might actually enhance the desire for more technology. There are at least five reasons for this.
The first is a phenomenon known as the Jevons paradox. It has been observed that technological improvements that increase the efficiency of a resource’s usage tend to lead to an overall increase in the consumption of that resource, rather than a decrease. This happens because the increased efficiency often reduces the cost of using the resource, which can increase demand and thus overall consumption. The perfect example is the concurrent improvements in the efficiency of combustion engines and increase in fossil fuel consumption.
Another reason is that new technologies open doors to further developments. Each new technology brings several additional possibilities for technological evolution, some of which will be compatible with the human-technology environment.
A third reason is that technology has the ability to more efficiently connect humans on larger scales. This allows for large-scale organizations to collaborate and create new technological developments with less difficulty.
In addition, technology improves our efficiency at conducting scientific research. With the improved knowledge of what is possible, it becomes easier to develop new technology. It is for this reason that many large tech companies dedicate so many resources to an R&D department. Research begets development, which begets improved research capabilities, ad infinitum.
Finally, cultural conditioning contributes to more rapid development of technology. If people learn that rapid technological progress is the norm, they are more accepting of new technological advancements.
Remark that the two necessary and sufficient criteria for a new technology to emerge, namely its possibility and desirability for some individuals, are extremely general, and allow for innovations that can be wildly misaligned with the goals of humans as a species, and those of you and me. The more technology advances, the wider the horizon of possibility, so the possibility aspect is an ever weakening constraint. Due to the dynamic nature of the possibilities created by technology, it is difficult to examine the mechanism of Fisherian runaway on this dimension. The more accessible dimension to study is that of desirability. The broad principles that motivate us to develop new technologies remain relatively stable, and studying them can illuminate how Fisherian runaway manifests itself in our relationship with technology.
What motivates technological development?
Although there are often several motivations for the development of any given technology, it helps to outline what these drivers are so that we may measure to what extent they remain present in driving future technologies, even in a rapidly changing technological landscape. Moreover, we may measure to what extent the driving goal is aligned with the goal of humanity, for the goal of a technology has no reason to be more humanitarian than those of its creators and users.
In the following, I list seven general categories of driving goals for innovation. Each category applies to both the inception and use of a technology. I will describe how each contributes to positive feedback systems that drive technological development, and provide elementary comments on the alignment of the driving goal with the goals of humanity as a whole.
Efficiency: The desire to help oneself or others to realize their goals by facilitating the flow of resources. Examples are abundant, and most publicly advertised mission statements of large technology companies fall into this category, even if there are underlying alternative motivations.
I will come back to this point later in this article, but efficiency is often the idealized goal of technological progress. For many people, it is what technology represents in a large way. Whether or not increased efficiency through technological improvements is good or bad for humanity is a topic of debate, and both sides will certainly be discussed in future articles.
Once the technology is introduced and normalized into society, the benchmark for efficiency is raised, and human satisfaction restabilizes at an equilibrium. This phenomenon is analogous to the hedonic treadmill, in which human satisfaction is observed to stabilize back to a baseline happiness after life changing events. The same can be said for our satisfaction with improvements in efficiency, and so this allows for the push for more efficient technology to continue.
Capital: The desire to accumulate resources. The development of modern technology is backed by the belief that the development will be profitable. The market rewards those who create new technologies that are profitable, and profitability loosely translates to increased efficiency in consumers fulfilling their subgoals. The more we trust in technology motivated by capital gain, the more we trust in a system driven by consumption, and by extension, the more we believe that the average person — where the average is weighted by capital wealth — pursuing their consumption goals leads to our own well-being. As technology evolves, so do the goals of the average consumer. Should we trust that they will remain oriented towards fostering well-being, or might they drift in the absence of a restoring force?
The subgoals of the average consumer may be in direct opposition to the individual’s well-being. Technology companies will nevertheless profit off of this, and may even be responsible for this. Take the tobacco industry, for example. Remark that this industry is almost purely motivated by capital. Cigarettes create enormous inefficiencies while offering almost no efficiency2.
Technological advancement motivated by capital gives rise to the Fisherian runaway mechanism. History has shown that the rich tend to get richer, especially through technological advancements, which bolsters the premise that there is a positive feedback system at work.
Status / recognition / esteem: The desire for appreciation. This is similar to capital if you see status and recognition as resources that can be accumulated. With that being said, the actual function of the technology is quite irrelevant, only that these technologies are designed in ways that achieve something unlikely or unimagined (not necessarily virtuous). The desire for recognition is so strongly reinforced by the stories we tell about billionaires, scientists, athletes, and adventurers. Even diabolical wrong-doers have earned a spot in history books, a prospect that most likely contributed to motivating their evils.
Consider a scientist who discovers a detrimental system that has the potential to hurt humanity in a devastating way. Should he publish his result? If he does not, then surely it will be only a matter of time before someone else discovers the same result, and steals the recognition. It is even likely that the scientist convinces himself that he is doing the world a favour by warning about the looming threat. For these reasons, and the obvious reason that there is recognition that accompanies well-cited publications, I think the scientist would publish the result. He might later admit publicly that he regrets his work, as has been the case in several instances in the history of science. The bottom line is: all surprising results get published.
There are several examples where CEOs of big tech companies and other prominent figures with high-profile projects and public persona that reflect a desire for recognition. Steve Jobs once said, “We’re here to put a dent in the universe. Otherwise why else even be here?". I believe that status and recognition were strong motivators for Elon Musk to buy Twitter and subsequentially rebrand it X3. After all, X is used by over 300 million people, and in controlling X, he has their attention.
Humans have, and will always have an insatiable thirst for fame, recognition, and status. This motivator for technological development pushes technology in directions that may be aligned with, orthogonal to, or against human development.
Curiosity: A desire to learn about what is possible. This is the idealized goal of science, even though working in the scientific domain has shown me that even science is largely motivated by finances and prestige. Speaking of motivations of curiosity, I have met artificial intelligence (AI) researchers that are publicly pessimistic about the future of AI, and at the same time, excited on an intellectual level to see where it goes.
We have seen before that scientific discovery is difficult to quarantine, and that nearly all that is found is eventually mixed into society, regardless of its benefit or harm. This is especially important to keep in mind when considering René Descartes’ advice,
Nature, in her indifference, makes nothing imperfect, nothing superfluous, and nothing in vain; but she is not kind or indulgent to her creatures, for she follows her own laws without caring whether they conform to our desires or not.
Furthermore, discovery usually incites further discovery. Albert Einstein writes,
As our circle of knowledge expands, so does the circumference of darkness surrounding it.
Weapons and defence: This category of motivations is similar to efficiency, in that the new technology helps to achieve the goal of successful combat. This leads to arms races, where the positive feedback system is symmetric — each actor is incentivised to have more advanced technology than their competitors. It is clear how the evolving equilibrium may be pushed away from one that promotes mutual well-being.
Future potential gain / strategic position: It may be unknown why a technology might be useful at the time it is being created, but rather the creator knows that the possession of the technology will likely be a strategic advantage once it is better understood. Artificial general intelligence is an example.
Technology that is motivated by future gains can give rise to the same positive feedback systems as those motivated by efficiency and capital, only with higher variance due to the uncertainty inherent in the prediction. Moreover, once the technology spreads to other individuals, the competitive advantage is lost.
Humanitarianism: A desire to improve overall well-being through technological development. Technologies for humanitarian action are created with a global vision of the evolution of humans and technology. They are based on values such as protecting the environment, improving education and healthcare, promoting equality, and fostering cooperation between people and communities.
The degree to which these values actually translate to positive change through humanitarian efforts and technological advancements may vary, especially when there are few resources devoted in this direction. Overall, I believe that with intelligent planning and sufficient resources, technology motivated by virtue can drastically improve well-being.
Driving goals, technology’s goals, and well-being
It was mentioned at the beginning of this article that Fisherian runaway in the evolution of species can decrease the survivability of the species. With the arguments presented in this article thus far, it remains unclear whether the positive feedback systems involving technological development are beneficial or detrimental to the long-term survivability of our developed society.
Technology is a collection of systems, each with a goal. By the fact that technologies are active participants in positive feedback systems involving our desires, we can study how the system goals of a technology relate to both its driving goals and the overarching goals of humanity. The effect of technology on the well-being of humans can be understood in terms of the alignment of these goals.
Let us first remark that the goal of a technology will not align exactly with the driving goals responsible for its inception and use. For example, an engine has the goal to combust fossil fuels in such a way that the excess chemical energy is transferred into kinetic energy. The production of CO2, water, and other by-products of combustion are just as much of the goal of the engine as the production of kinetic energy. This subgoal of the engine, the production of CO2, is neither desired by the user of the engine, nor by humanity. The user of the engine might not care about carbon pollution, but they are probably not a fan of the engine’s subgoal of consuming their fossil fuel.
A term that may be used to describe how aligned a technology’s goal is with the goal of a reference system is efficiency. Remark that by this definition, efficiency is relative to the reference system. A cancerous cell may efficiently succeed in reproducing itself, but the resulting tumour is inefficient for the animal that the cell belongs to. A perfectly efficient machine, by this definition, gets the job done with no unintended manipulation of the environment, and as fast as desired. Of course, all reasonable examples illustrate that perfect efficiency is impossible.
Inefficiencies tend to be misaligned with the goals of living systems. All living systems rely on homeostasis, and thus local entropy management. Inefficiencies are by definition unintended manipulations of the environment, which tend to increase entropy.
We may term the misalignment of a technology’s goal with the driving goal as technical inefficiency. The goal of a technology will be misaligned with the driving goal to some extent. Technological advancement reduces technical inefficiencies at least temporarily, until the driving goals evolve in the new technological environment.
All of the driving goals responsible for technological development outlined in the previous section do not totally align perfectly with the goals of humanity. While humanitarian goals are probably very close, I would argue that they are still only a model of what is truly good for humanity. Thus, even technology that is perfectly efficient in satisfying a driving goal may still be largely inefficient for humanity. We may think of the misalignment between the goals of humanity and the driving goals as scaling inefficiency4. Game theory and empirical evidence has proven that, left to their own devices, individuals acting for themselves will not always produce a system that optimally fosters the well-being of the group.
The systems that are responsible for creating scaling inefficiencies are often referred to as social traps. Examples are abundant in the sphere of climate preservation, where the tragedy of the commons takes centre stage. Other more subtle examples include Braess’s paradox, where it has been observed (with theoretical justification) that adding roads or lanes to a road network can have the overall effect of slowing down traffic.
To summarize, no technology is perfectly efficient in satisfying the goals of humanity. A simplified model of the inefficiency of technology in delivering the goals of humanity is described by the equation,
Total inefficiency = technical inefficiency + scaling inefficiency.
Technological development can improve technical inefficiencies, but scaling inefficiencies can only be improved by human systems implementing legislation, cooperation, education, etc.
The intricate relationship between technical inefficiency and scaling inefficiency is hidden in the above, simplified model. Technological advancements are often accompanied by a trend towards a more globalized society, where scaling inefficiencies become more relevant. Thus, we cannot treat the two types of inefficiencies as independent.
Conclusion
We have seen that technological developments have a tendency to instigate further technological developments. This is a phenomenon that continues independently of efforts to make the world a better place. It appears that even if society has sufficient reason to believe that rapid technological development is suboptimal, the forces that drive technological development are sufficiently strong to resist our efforts to stop them.
The question is: will we one day reach this conclusion? Some evidence suggests that it may be the case that new technologies are creating more harm than good, and that the classical view that progress can be measured by our technical abilities needs to be rethought. This article from The Roots of Progress outlines some compelling arguments for why this might be the case.
Rethinking Tech is dedicated to curating a wholistic view of technology and its interaction with humans. By adhering to a top-down approach, the conclusions that we reach about the benefits and harms of technology may be understood through a well-established philosophical framework. I hope that this framework empowers the ethical use and development of technology, and that efficiency in obtaining humanitarian goals may be improved. As for the other goals that are ever-so abundant in todays society, I hope that their power in driving technological development may be limited, and I believe that this may be achieved through intelligent discourse about what we as a society really want.
A man’s Adam’s apple is an example of Fisherian runaway in humans.
There’s the argument that cigarettes spark conversations between smokers when they ask each other for lighters.
He might as well have named it “Elon”, since the name “X” has a well-documented association with Elon Musk. He founded SpaceX, and has a child named “X Æ A-12“.
Scaling inefficiencies can be further broken down into inefficiencies between an actor’s subgoal and their well-being (like smoking tobacco) and inefficiencies between an individual’s desires and the well-being of society, i.e., social traps.