Digifesto

Tag: oliver wendell holmes jr.

Existentialism in Design: Motivation

There has been a lot of recent work on the ethics of digital technology. This is a broad area of inquiry, but it includes such topics as:

  • The ethics of Internet research, including the Facebook emotional contagion study and the Encore anti-censorship study.
  • Fairness, accountability, and transparnecy in machine learning.
  • Algorithmic price-gauging.
  • Autonomous car trolley problems.
  • Ethical (Friendly?) AI research? This last one is maybe on the fringe…

If you’ve been reading this blog, you know I’m quite passionate about the intersection of philosophy and technology. I’m especially interested in how ethics can inform the design of digital technology, and how it can’t. My dissertation is exploring this problem in the privacy engineering literature.

I have a some dissatisfaction towards this field which I don’t expect to make it into my dissertation. One is that the privacy engineering literature and academic “ethics of digital technology” more broadly tends to be heavily informed by the law, in the sense of courts, legislatures, and states. This is motivated by the important consideration that technology, and especially technologists, should in a lot of cases be compliant with the law. As a practical matter, it certainly spares technologists the trouble of getting sued.

However, being compliant with the law is not precisely the same things as being ethical. There’s a long ethical tradition of civil disobedience (certain non-violent protest activities, for example) which is not strictly speaking legal though it has certainly had impact on what is considered legal later on. Meanwhile, the point has been made but maybe not often enough that legal language often looks like ethical language, but really shouldn’t be interpreted that way. This is a point made by Oliver Wendell Holmes Junior in his notable essay, “The Path of the Law”.

When the ethics of technology are not being framed in terms of legal requirements, they are often framed in terms of one of two prominent ethical frameworks. One framework is consequentialism: ethics is a matter of maximizing the beneficial consequences and minimizing the harmful consequences of ones actions. One variation of consequentialist ethics is utilitarianism, which attempts to solve ethical questions by reducing them to a calculus over “utility”, or benefit as it is experienced or accrued by individuals. A lot of economics takes this ethical stance. Another, less quantitative variation of consequentialist ethics is present in the research ethics principle that research should maximize benefits and minimize harms to participants.

The other major ethical framework used in discussions of ethics and technology is deontological ethics. These are ethics that are about rights, duties, and obligations. Justifying deontological ethics can be a little trickier than justifying consequentialist ethics. Frequently this is done by invoking social norms, as in the case of Nissenbaum’s contextual integrity theory. Another variation of a deontological theory of ethics is Habermas’s theory of transcendental pragmatics and legitimate norms developed through communicative action. In the ideal case, these norms become encoded into law, though it is rarely true that laws are ideal.

Consequentialist considerations probably make the world a better place in some aggregate sense. Deontological considerations probably maybe the world a fairer or at least more socially agreeable place, as in their modern formulations they tend to result from social truces or compromises. I’m quite glad that these frameworks are taken seriously by academic ethicists and by the law.

However, as I’ve said I find these discussions dissatisfying. This is because I find both consequentialist and deontological ethics to be missing something. They both rely on some foundational assumptions that I believe should be questioned in the spirit of true philosophical inquiry. A more thorough questioning of these assumptions, and tentative answers to them, can be found in existentialist philosophy. Existentialism, I would argue, has not had its due impact on contemporary discourse on ethics and technology, and especially on the questions surrounding ethical technical design. This is a situation I intend to one day remedy. Though Zach Weinersmith has already made a fantastic start:

“Self Driving Car Ethics”, by Weinersmith

SMBC: Autonomous vehicle ethics

What kinds of issues would be raised by existentialism in design? Let me try out a few examples of points made in contemporary ethics of technology discourse and a preliminary existentialist response to them.

Ethical Charge Existentialist Response
A superintelligent artificial intelligence could, if improperly designed, result in the destruction or impairment of all human life. This catastrophic risk must be avoided. (Bostrom, 2014) We are all going to die anyway. There is no catastrophic risk; there is only catastrophic certainty. We cannot make an artificial intelligence that prevents this outcome. We must instead design artificial intelligence that makes life meaningful despite its finitude.
Internet experiments must not direct the browsers of unwitting people to test the URLs of politically sensitive websites. Doing this may lead to those people being harmed for being accidentally associated with the sensitive material. Researchers should not harm people with their experiments. (Narayanan and Zevenbergen, 2015) To be held responsible by a state’s criminal justice system for the actions taken by ones browser, controlled remotely from America, is absurd. This absurdity, which pervades all life, is the real problem, not the suffering potentially caused by the experiment (because suffering in some form is inevitable, whether it is from painful circumstance or from ennui.) What’s most important is the exposure of this absurdity and the potential liberation from false moralistic dogmas that limit human potential.
Use of Big Data to sort individual people, for example in the case of algorithms used to choose among applicants for a job, may result in discrimination against historically disadvantaged and vulnerable groups. Care must be taken to tailor machine learning algorithms to adjust for the political protection of certain classes of people. (Barocas and Selbst, 2016) The egalitarian tendency in ethics which demands that the greatest should invest themselves in the well-being of the weakest is a kind of herd morality, motivated mainly by ressentiment of the disadvantaged who blame the powerful for their frustrations. This form of ethics, which is based on base emotions like pity and envy, is life-negating because it denies the most essential impulse of life: to overcome resistance and to become great. Rather than restrict Big Data’s ability to identify and augment greatness, it should be encouraged. The weak must be supported out of a spirit of generosity from the powerful, not from a curtailment of power.

As a first cut at existentialism’s response to ethical concerns about technology, it may appear that existentialism is more permissive about the use and design of technology than consequentialism and deontology. It is possible that this conclusion will be robust to further investigation. There is a sense in which existentialism may be the most natural philosophical stance for the technologist because a major theme in existentialist thought is the freedom to choose ones values and the importance of overcoming the limitations on ones power and freedom. I’ve argued before that Simone de Beauvoir, who is perhaps the most clear-minded of the existentialists, has the greatest philosophy of science because it respects this purpose of scientific research. There is a vivacity to existentialism that does not sweat the small stuff and thinks big while at the same time acknowledging that suffering and death are inevitable facts of life.

On the other hand, existentialism is a morally demanding line of inquiry precisely because it does not use either easy metaethical heuristics (such as consequentialism or deontology) or the bald realities of the human condition as a stopgap. It demands that we tackle all the hard questions, sometimes acknowledging that they are answerable or answerable only in the negative, and muddle on despite the hardest truths. Its aim is to provide a truer, better morality than the alternatives.

Perhaps this is best illustrated by some questions implied by my earlier “existentialist responses” that address the currently nonexistent field of existentialism in design. These are questions I haven’t yet heard asked by scholars at the intersection of ethics and technology.

  • How could we design an artificial intelligence (or, to make it simpler, a recommendation system) that makes the most meaningful choices for its users?
  • What sort of Internet intervention would be most liberatory for the people affected by it?
  • What technology can best promote generosity from the world’s greatest people as a celebration of power and life?

These are different questions from any that you read about in the news or in the ethical scholarship. I believe they are nevertheless important ones, maybe more important than the ethical questions that are more typically asked. The theoretical frameworks employed by most ethicists make assumptions that obscure what everybody already knows about the distribution of power and its abuses, the inevitability of suffering and death, life’s absurdity and especially the absurdity if moralizing sentiment in the face of the cruelty of reality, and so on. At best, these ethical discussions inform the interpretation and creation of law, but law is not the same as morality and to confuse the two robs morality of what is perhaps most essential component, which is that is grounded meaningfully in the experience of the subject.

In future posts (and, ideally, eventually in a paper derived from those posts), I hope to flesh out more concretely what existentialism in design might look like.

References

Barocas, S., & Selbst, A. D. (2016). Big data’s disparate impact.

Bostrom, N. (2014). Superintelligence: Paths, dangers, strategies. OUP Oxford.

Narayanan, A., & Zevenbergen, B. (2015). No Encore for Encore? Ethical questions for web-based censorship measurement.

Weinersmith, Z. “Self Driving Car Ethics”. Saturday Morning Breakfast Cereal.

Advertisements

The FTC and pragmatism; Hoofnagle and Holmes

I’ve started working my way through Chris Hoofnagle’s Federal Trade Commission Privacy Law and Policy. Where I’m situated at the I School, there’s a lot of representation and discussion of the FTC in part because of Hoofnagle’s presence there. I find all this tremendously interesting but a bit difficult to get a grip on, as I have only peripheral experiences of actually existing governance. Instead I’m looking at things with a technical background and what can probably be described as overdeveloped political theory baggage.

So a clearly written and knowledgeable account of the history and contemporary practice of the FTC is exactly what I need to read, I figure.

With the poor judgment of commenting on the book having just cracked it open, I can say that the book reads so far as, not surprisingly, a favorable account of the FTC and its role in privacy law. In broad strokes, I’d say Hoofnagle’s narrative is that while the FTC started out as a compromise between politicians with many different positions on trade regulation, and while its had at times “mediocre” leadership, now the FTC is run by selfless, competent experts with the appropriate balance of economic savvy and empathy for consumers.

I can’t say I have any reason to disagree. I’m not reading for either a critique or an endorsement of the agency. I’m reading with my own idiosyncratic interests in mind: algorithmic law and pragmatist legal theory, and the relationship between intellectual property and antitrust. I’m also learning (through reading) how involved the FTC has been in regulating advertising, which endears me to the adjacency because I find most advertising annoying.

Missing as I am any substantial knowledge of 20th century legal history, I’m intrigued by resonances between Hoofnagle’s account of the FTC and Oliver Wendell Holmes Jr.’s “The Path of the Law“, which I mentioned earlier. Apparently there’s some tension around the FTC as some critics would like to limit its powers by holding it more narrowly accountable to common law, as oppose to (if I’m getting this right) a more broadly scoped administrative law that, among other things, allows it to employ skilled economist and technologists. As somebody who has been intellectually very informed by American pragmatism, I’m pleased to notice that Holmes himself would have probably approved of the current state of the FTC:

At present, in very many cases, if we want to know why a rule of law has taken its particular shape, and more or less if we want to know why it exists at all, we go to tradition. We follow it into the Year Books, and perhaps beyond them to the customs of the Salian Franks, and somewhere in the past, in the German forests, in the needs of Norman kings, in the assumptions of a dominant class, in the absence of generalized ideas, we find out the practical motive for what now best is justified by the mere fact of its acceptance and that men are accustomed to it. The rational study of law is still to a large extent the study of history. History must be a part of the study, because without it we cannot know the precise scope of rules which it is our business to know. It is a part of the rational study, because it is the first step toward an enlightened scepticism, that is, towards a deliberate reconsideration of the worth of those rules. When you get the dragon out of his cave on to the plain and in the daylight, you can count his teeth and claws, and see just what is his strength. But to get him out is only the first step. The next is either to kill him, or to tame him and make him a useful animal. For the rational study of the law the blackletter man may be the man of the present, but the man of the future is the man of statistics and the master of economics. It is revolting to have no better reason for a rule of law than that so it was laid down in the time of Henry IV. It is still more revolting if the grounds upon which it was laid down have vanished long since, and the rule simply persists from blind imitation of the past. (Holmes, 1897)

These are strong words from a Supreme Court justice about the limitations of common law! It’s also a wholehearted endorsement of quantified science as the basis for legal rules. Perhaps what Holmes would have preferred is a world in which statistics and economics themselves became part of the logic of law. However, he goes to pains to point out how often legal judgment itself does not depend on logic so much as the unconscious biases of judges and juries, especially with respect to questions of “social advantage”:

I think that the judges themselves have failed adequately to recognize their duty of weighing considerations of social advantage. The duty is inevitable, and the result of the often proclaimed judicial aversion to deal with such considerations is simply to leave the very ground and foundation of judgments inarticulate, and often unconscious, as I have said. When socialism first began to be talked about, the comfortable classes of the community were a good deal frightened. I suspect that this fear has influenced judicial action both here and in England, yet it is certain that it is not a conscious factor in the decisions to which I refer. I think that something similar has led people who no longer hope to control the legislatures to look to the courts as expounders of the constitutions, and that in some courts new principles have been discovered outside the bodies of those instruments, which may be generalized into acceptance of the economic doctrines which prevailed about fifty years ago, and a wholesale prohibition of what a tribunal of lawyers does not think about right. I cannot but believe that if the training of lawyers led them habitually to consider more definitely and explicitly the social advantage on which the rule they lay down must be justified, they sometimes would hesitate where now they are confident, and see that really they were taking sides upon debatable and often burning questions.

What I find interesting about this essay is that it somehow endorses both the use of economics and statistics in advancing legal thinking and also endorses what has become critical legal theory, with its specific consciousness of the role of social power relations in law. So often in contemporary academic discourse, especially when it comes to discussion of regulation technology businesses, these approaches to law are considered opposed. Perhaps it’s appropriate to call a more politically centered position, if there were one today, a pragmatist position.

Perhaps quixotically, I’m very interested in the limits of these arguments and their foundation in legal scholarship because I’m wondering to what extent computational logic can become a first class legal logic. Holmes’s essay is very concerned with the limitations of legal logic:

The fallacy to which I refer is the notion that the only force at work in the development of the law is logic. In the broadest sense, indeed, that notion would be true. The postulate on which we think about the universe is that there is a fixed quantitative relation between every phenomenon and its antecedents and consequents. If there is such a thing as a phenomenon without these fixed quantitative relations, it is a miracle. It is outside the law of cause and effect, and as such transcends our power of thought, or at least is something to or from which we cannot reason. The condition of our thinking about the universe is that it is capable of being thought about rationally, or, in other words, that every part of it is effect and cause in the same sense in which those parts are with which we are most familiar. So in the broadest sense it is true that the law is a logical development, like everything else. The danger of which I speak is not the admission that the principles governing other phenomena also govern the law, but the notion that a given system, ours, for instance, can be worked out like mathematics from some general axioms of conduct. This is the natural error of the schools, but it is not confined to them. I once heard a very eminent judge say that he never let a decision go until he was absolutely sure that it was right. So judicial dissent often is blamed, as if it meant simply that one side or the other were not doing their sums right, and if they would take more trouble, agreement inevitably would come.

This mode of thinking is entirely natural. The training of lawyers is a training in logic. The processes of analogy, discrimination, and deduction are those in which they are most at home. The language of judicial decision is mainly the language of logic. And the logical method and form flatter that longing for certainty and for repose which is in every human mind. But certainty generally is illusion, and repose is not the destiny of man. Behind the logical form lies a judgment as to the relative worth and importance of competing legislative grounds, often an inarticulate and unconscious judgment, it is true, and yet the very root and nerve of the whole proceeding. You can give any conclusion a logical form. You always can imply a condition in a contract. But why do you imply it? It is because of some belief as to the practice of the community or of a class, or because of some opinion as to policy, or, in short, because of some attitude of yours upon a matter not capable of exact quantitative measurement, and therefore not capable of founding exact logical conclusions. Such matters really are battle grounds where the means do not exist for the determinations that shall be good for all time, and where the decision can do no more than embody the preference of a given body in a given time and place. We do not realize how large a part of our law is open to reconsideration upon a slight change in the habit of the public mind. No concrete proposition is self evident, no matter how ready we may be to accept it, not even Mr. Herbert Spencer’s “Every man has a right to do what he wills, provided he interferes not with a like right on the part of his neighbors.”

For Holmes, nature can be understood through a mathematized physics and is in this sense logical. But the law itself is not logical in the narrow sense of providing certainty about concrete propositions and the legal interpretation of events.

I wonder whether the development of more flexible probabilistic logics, such as those that inform contemporary machine learning techniques, would have for Holmes adequately bridged the gap between the logic of nature and the ambiguity of law. These probabilistic logics are designed to allow for precise quantification of uncertainty and ambiguity.

This is not a purely academic question. I’m thinking concretely about applications to regulation. Some of this has already been implemented. I’m thinking about Datta, Tschantz, and Datta’s “Automated Experiments on Ad Privacy Settings: A Tale of Opacity, Choice, and Discrimination” (pdf). I know several other discrimination auditing tools have been developed by computer science researchers. What is the legal status of these tools? Could they or should they be implemented as a scalable or real-time autonomous system?

I was talking to an engineer friend the other day and he was telling me that internally to Google, there’s a team responsible for building the automated system that tests all of its other automated systems to make sure that it is adherence to its own internal privacy standards. This was a comforting thing to hear and not a surprise, as I get the sense from conversations I’ve had with Googler’s that they are in general a very ethically conscientious company. What’s distressing to me is that Google may have more powerful techniques available for self-monitoring than the government has for regulation. This is because (I think…again my knowledge of these matters is actually quite limited) at Google they know when a well-engineered computing system is going to perform better than a team of clerks, and so developing this sort of system is considered worthy of investment. It will be internally trusted as much as any other internal expertise. Whereas in the court system, institutional inertia and dependency on discursive law mean that at best this sort of system can be brought in as an expensive and not entirely trusted external source.

What I’d like to figure out is to what extent agency law in particular is flexible enough to be extended to algorithmic law.

algorithmic law and pragmatist legal theory: Oliver Wendell Holmes Jr. “The Path of the Law”

Several months ago I was taken by the idea that in the future (and maybe depending on how you think about it, already in the present) laws should be written as computer algorithms. While the idea that “code is law” and that technology regulates is by no means original, what I thought perhaps provocative is the positive case for the (re-)implementation of the fundamental laws of the city or state in software code.

The argument went roughly like this:

  • Effective law must control a complex society
  • Effective control requires social and political prediciton.
  • Unassisted humans are not good at social and political prediction. For this conclusion I drew heavily on Philip Tetlock’s work in Expert Political Judgment.
  • Therefore laws, in order to keep pace with the complexity of society, should be implemented as technical systems capable of bringing data and machine learning to bear on social control.

Science fiction is full of both dystopias and utopias in which society is literally controlled by a giant, intelligent machine. Avoiding either extreme, I just want to make the modest point that there may be scalability problems with law and regulation based on discourse in natural language. To some extent the failure of the state to provide sophisticated, personalized regulation in society has created myriad opportunities for businesses to fill these roles. Now there’s anxiety about the relationship between these businesses and the state as they compete for social regulation. To the extent that businesses are less legitimate rulers of society than the state, it seems a practical, technical necessity that the state adopt the same efficient technologies for regulation that businesses have. To do otherwise is to become obsolete.

There are lots of reasons to object to this position. I’m interested in hearing yours and hope you will comment on this and future blog posts or otherwise contact me with your considered thoughts on the matter. To me the strongest objection is that the whole point of the law is that it is based on precedent, and so any claim about the future trajectory of the law has to be based on past thinking about the law. Since I am not a lawyer and I know precious little about the law, you shouldn’t listen to my argument because I don’t know what I’m talking about. Q.E.D.

My counterargument to this is that there’s lots of academics who opine about things they don’t have particular expertise in. One way to get away with this is by deferring to somebody else who has credibility in field of interest. This is just one of several reasons why I’ve been reading “The Path of the Law“, a classic essay about pragmatist legal theory written by Supreme Court Justice Oliver Wendell Holmes Jr. in 1897.

One of the key points of this essay is that it is a mistake to consider the study of law the study of morality per se. Rather, the study of law is the attempt to predict the decisions that courts will make in the future, based on the decisions courts will make in the past. What courts actually decide is based in part of legal precedent but also on the unconsciously inclinations of judges and juries. In ambiguous cases, different legal framings of the same facts will be in competition, and the judgment will give weight to one interpretation or another. Perhaps the judge will attempt to reconcile these differences into a single, logically consistent code.

I’d like to take up the arguments of this essay again in later blog posts, but for now I want to focus on the concept of legal study as prediction. I think this demands focus because while Holmes, like most American pragmatists, had a thorough and nuanced understanding of what prediction is, our mathematical understanding of prediction has come a long way since 1897. Indeed, it is a direct consequence of these formalizations and implementations of predictive systems that we today see so much tacit social regulation performed by algorithms. We know now that effective prediction depends on access to data and the computational power to process it according to well-known algorithms. These algorithms can optimize themselves to such a degree that their specific operations are seemingly beyond the comprehension of the people affected by them. Some lawyers have argued that this complexity should not be allowed to exist.

What I am pointing to is a fundamental tension between the requirement that practitioners of the law be able to predict legal outcomes, and the fact that the logic of the most powerful predictive engines today is written in software code not words. This is because of physical properties of computation and prediction that are not likely to ever change. And since a powerful predictive engine can just as easily use its power to be strategically unpredictable, this presents an existential challenge to the law. It may simply be impossible for lawyers acting as human lawyers have for hundreds of years to effectively predict and therefor regulate powerful computational systems.

One could argue that this means that such powerful computational systems should simply be outlawed. Indeed this is the thrust of certain lawyers. But if we believe that these systems are not going to go away, perhaps because they won’t allow us to regulate them out of existence, then our only viable alternative to suffering under their lawless control is to develop a competing system of computational legalism with the legitimacy of the state.