Existential Risk and Existential Uncertainty
6 March 2013
Frank Knight on risk and uncertainty
Early Chicago school economist Frank Knight was known for his work on risk, and especially for the distinction between risk and uncertainty, which is still taught in economics and business courses. Like Schumpeter, Knight was interested in the function of the entrepreneur in the modern commercial economy, and he employed his distinction between risk and uncertainty in order to illuminate the function of the entrepreneur.
Although it is easy to conflate risk and uncertainty, and to speak as though facing a risk were the same thing as facing uncertain or unknown circumstances, Knight doesn’t see it like this at all. A risk can be quantified and calculated, and because risks can be quantified and calculated, they can be controlled. This is the function of insurance: to quantify and price risk. If you have correctly factored risk into your calculation, it is no longer an uncertainty. You might not know the exact date or magnitude of losses, but you know statistically that there will be a certain number of losses of a certain magnitude. It is the job of actuaries to calculate this, and one buys insurance to control the risk to which one is exposed.
The ordinary business of life, and of business, according to Knight, involves risk management, but the unique function of the entrepreneur is to accept uncertainty that cannot be quantified, priced, or insured. The entrepreneur makes his profit not in spite of uncertainty, but because of uncertainty. No insurance can be bought for uncertainty, so that in taking on an uncertain situation the entrepreneur enters into a realm in which it is recognized that there are factors beyond control. If he is not destroyed financially by these uncontrollable factors, he may profit from them, and this profit is likely to exceed the profit made in ordinary business operations exposed to risk but not to uncertainty.
Here is how Knight formulated his distinction between risk and uncertainty:
To preserve the distinction which has been drawn in the last chapter between the measurable uncertainty and an unmeasurable one we may use the term “risk” to designate the former and the term “uncertainty” for the latter. The word “risk” is ordinarily used in a loose way to refer to any sort of uncertainty viewed from the standpoint of the unfavorable contingency and the term “uncertainty” similarly with reference to the favorable outcome; we speak of the “risk” of a loss, the “uncertainty” of a gain. But if our reasoning so far is at an correct, there is a fatal ambiguity in these terms which must be gotten rid of and the use of the term “risk” in connection with the measurable uncertainties or probabilities of insurance gives some justification for specializing the terms as just indicated. We can also employ the terms “objective” and “subjective” probability to designate the risk and uncertainty respectively, as these expressions are already in general use with a signification akin to that proposed.
Frank Knight, Risk, Uncertainty, and Profit, CHAPTER VIII, STRUCTURES AND METHODS FOR MEETING UNCERTAINTY
Knight went on to add…
The practical difference between the two categories, risk and uncertainty, is that in the former the distribution of the outcome in a group of instances is known (either through calculation a priori or from statistics of past experience), while in the case of uncertainty this is not true, the reason being in general that it is impossible to form a group of instances, because the situation dealt with is in a high degree unique.
Frank Knight, Risk, Uncertainty, and Profit, CHAPTER VIII, STRUCTURES AND METHODS FOR MEETING UNCERTAINTY
The growth of knowledge and experience can transform uncertainty into risk if it contextualizes a formerly unique situation in such a way as to demonstrate that it is not unique but belongs to a group of instances. Of the tremendous gains made in the space sciences during the last forty years, during our selective space age stagnation, it could be said that the function of this considerable gain in knowledge has been to transform uncertainty into risk. But this goes only so far.
Even if the boundary between risk and uncertainty can be pushed outward by the growth of knowledge, the same growth of civilization that attends the growth of knowledge and technology means that the boundaries of civilization itself will also be pushed further out, with the result being that we are likely to always encounter further uncertainties even as old uncertainties are transformed by knowledge into risk.
The evolution of the existential risk concept
In many recent posts I have been discussing the idea of existential risk. These posts include, but are not limited to, Moral Imperatives Posed by Existential Risk, Research Questions on Existential Risk, and Six Theses on Existential Risk. The idea of existential risk is due to Nick Bostrom. (I first heard about this at the first 100YSS symposium in Orlando in 2011, when I was talking to Christian Weidemann.)
Nick Bostrom defined existential risk as follows:
Existential risk – One where an adverse outcome would either annihilate Earth-originating intelligent life or permanently and drastically curtail its potential.
An existential risk is one where humankind as a whole is imperiled. Existential disasters have major adverse consequences for the course of human civilization for all time to come.
“Existential Risks: Analyzing Human Extinction Scenarios and Related Hazards,” Nick Bostrom, Professor, Faculty of Philosophy, Oxford University, Published in the Journal of Evolution and Technology, Vol. 9, No. 1 (2002)
In his papers on existential risk and the book on Global Catastrophic Risks, Bostrom steadily expanded and refined the parameters of disasters that have (or would have) major adverse consequences for human beings and their civilization.
The table from an early existential risk paper above divides qualitative risks into six categories. the table below from the book Global Catastrophic Risks includes twelve qualitative risk categories and implies another eight; the table further below from a more recent paper includes fifteen qualitative risk categories and implies another nine. From a philosophical point of view, these further distinctions represent in advance in clarity, contextualizing both existential risks and global catastrophic risks in a matrix of related horrors.
The specific possible events that Bostrom describes range from the imperceptible loss of one hair to human extinction. Recently in Moral Imperatives Posed by Existential Risk I tried to point out how further distinctions can be made within the variety of human extinction scenarios, and that some distinct outcomes might be morally preferable over other outcomes. For example, even if human beings were to become extinct, we might want some of our legacy to remain to potentially be discovered by alien species visiting our solar system. Given the presence of space probes throughout our solar system, it seems highly likely that these would survive any human extinction scenario, so that we have left some kind of mark on the cosmos — a cosmic equivalent of “Kilroy was here.”
Further distinction can be made, however, and the distinction that I would like to urge today is that of distinguishing existential risks from existential uncertainties.
The need to explicitly formulate existential uncertainty
Once the distinction is made between existential risks and existential uncertainties, we recognize that existential risks can be quantified and calculated. Ultimately, existential risks can also be insured. The industrial and financial infrastructure is not now in place to do this, although we can clearly see how to do this. And this much is obvious, because much of the discussion of existential risk focuses on potential mitigation efforts. Existential risk mitigation is insurance against extinction.
We can clearly understand that we can guard against the existential risks posed by massive asteroid impacts by a system of observation of objects in space likely to cross the path of the Earth, and building spacecraft that could deflect or otherwise render harmless such threatening asteroids. It was once thought that the appearance of comets or “new stars” (novae) in the sky heralded the death of kings of the end of empires. No longer. This is the perfect example of a former uncertainty that has been transformed into a risk by the growth of knowledge (or, at very least, is in the process of being transformed from an uncertainty into a risk).
We can also clearly see that we could back up the Earth’s biosphere about a truly catastrophic global disaster by transplanting Earth-originating life elsewhere. Far in the future we can even understand the risk of the sun swelling into a red giant and consuming the Earth in its fires — unless by that time we have moved the Earth to an orbit where it remains safe, or perhaps even transported it to another star. All of these are existential risks where “risk” is used sensu stricto.
There are a great many existential risks and global catastrophic risks that have been proposed. When it comes to geological events — like massive vulcanization — or cosmological events — the death of our sun — the sciences of geology and cosmology are likely to mature to the point where these risks are quantifiable, and if industrial-technological civilization continues its path of exponential development, we should also someday have the technology to adequately “insure” against these existential risks.
The vagaries of history and civilization
When it comes to scenarios that involve events and processes not of the variety that contemporary natural science can formulate, we are clearly pushing the envelope of existential risks and verging on existential uncertainties. Such scenarios would include those predicated upon the development of human history and civilization. For example, scenarios of wars of an order of magnitude that far exceed the magnitude of the global wars of the twentieth century are on the outer edges of risk and, as they become more speculative in their formulation, verge onto uncertainty. Similarly, scenarios that involve the intervention of alien species in human history and human civilization — alien invasion, alien enslavement, alien visitation, etc. — verge onto being existential uncertainties.
The anthropogenic existential risks that are of primary concern to Nick Bostrom, Martin Rees, and others — risks from artificial intelligence, machine consciousness, unintended consequences of advanced technologies, and the “gray goo” problem potentially posed by nanotechnology — are similarly problematic as risks, and many must be accounted as uncertainties. In regard to the anthropogenic dimension of many existential uncertainties I am reminded of a passage from Carl Sagan’s Cosmos:
“Biology is more like history than it is like physics. You have to know the past to understand the present. And you have to know it in exquisite detail. There is as yet no predictive theory of biology, just as there is not yet a predictive theory of history. The reasons are the same: both subjects are still too complicated for us. But we can know ourselves better by understanding other cases. The study of a single instance of extraterrestrial life, no matter how humble, will deprovincialize biology. For the first time, the biologists will know what other kinds of life are possible. When we say the search for life elsewhere is important, we are not guaranteeing that it will be easy to find – only that it is very much worth seeking.
Carl Sagan, Cosmos, CHAPTER II, One Voice in the Cosmic Fugue
This strikes me as one of the most powerful and important passages in Cosmos. When Sagan writes that, “[t]here is as yet no predictive theory of biology, just as there is not yet a predictive theory of history,” while leaving open the possibility of a future predictive science of biology and history — he wrote as yet — he squarely recognized that neither biology nor human history (much of which derives more or less directly from biology) can be predicted or quantified or measured in a scientific way. If we had a science of history, such as Marx thought we had discovered, then the potential disasters of human history could be quantified, and we could insure against them.
Well, we can insure against some eventualities of history, though certainly not against all. This is a point that Machiavelli makes:
It is not unknown to me how many men have had, and still have, the opinion that the affairs of the world are in such wise governed by fortune and by God that men with their wisdom cannot direct them and that no one can even help them; and because of this they would have us believe that it is not necessary to labour much in affairs, but to let chance govern them. This opinion has been more credited in our times because of the great changes in affairs which have been seen, and may still be seen, every day, beyond all human conjecture. Sometimes pondering over this, I am in some degree inclined to their opinion. Nevertheless, not to extinguish our free will, I hold it to be true that Fortune is the arbiter of one-half of our actions, but that she still leaves us to direct the other half, or perhaps a little less.
I compare her to one of those raging rivers, which when in flood overflows the plains, sweeping away trees and buildings, bearing away the soil from place to place; everything flies before it, all yield to its violence, without being able in any way to withstand it; and yet, though its nature be such, it does not follow therefore that men, when the weather becomes fair, shall not make provision, both with defences and barriers, in such a manner that, rising again, the waters may pass away by canal, and their force be neither so unrestrained nor so dangerous. So it happens with fortune, who shows her power where valour has not prepared to resist her, and thither she turns her forces where she knows that barriers and defences have not been raised to constrain her.
Nicolo Machiavelli, The Prince, CHAPTER XXV, “What Fortune Can Effect In Human Affairs, And How To Withstand Her”
What remains beyond the predictable storms of floods of history are the true uncertainties, the unknown unknowns, and these pose a danger we cannot predict, quantify, or insure. They are not, then, risks in the strict sense. They are existential uncertainties.
It could be argued that our inability to take specific, concrete, effective measures to mitigate the obvious uncertainties of life has resulted in religious responses to uncertainty that systematically avoid falsifiability and thereby secure the immunity of hopes to exterior circumstances. Whether or not this has been true in the past, merely the recognition of existential uncertainty is the first step toward rationally assessing them.
Existential risk suggests a clear course of mitigating action; existential uncertainty cannot, on the contrary, be the object of planning and preparation. The most that one can do to address existential uncertainty is to keep oneself open and flexible, ready to roll with the punches, and responsive to any challenge that might arise, meeting it at the height of one’s powers; any attempt to prepare specific measures will be fruitless, and quite possibly counter-productive because of the wasted effort.
. . . . .
. . . . .
. . . . .
. . . . .