Dear Legal Academics, Please Stop Misusing The Word ‘Algorithms’

Everyone is concerned about ‘algorithms.’ Especially legal academics; law review articles, conferences, symposia all bear testimony to this claim. Algorithms and transparency; the tyranny of algorithms; how algorithms can deprive you of your rights; and so on. Algorithmic decision making is problematic; so is algorithmic credit scoring; or algorithmic stock trading. You get the picture; something new and dangerous called the ‘algorithm’ has entered the world, and it is causing havoc. Legal academics are on the case (and they might even occasionally invite philosophers and computer scientists to pitch in with this relief effort.)

There is a problem with this picture. ‘Algorithms’ is the wrong word to describe the object of legal academics’ concern. An algorithm is “an unambiguous specification of how to solve a class of problems” or a step-by-step procedure which terminates with a solution to a given problem. These problems can be of many kinds: mathematical or logical ones are not the only ones, for a cake-baking recipe is also an algorithm, as are instructions for crossing a street. Algorithms can be deterministic or non-deterministic; they can be exact or approximate; and so on. But, and this is their especial feature, algorithms are abstract specifications; they lack concrete implementations.

Computer programs are one kind of implementation of algorithms; but not the only one. The algorithm for long division can be implemented by pencil and paper; it can also be automated on a hand-held calculator; and of course, you can write a program in C or Python or any other language of your choice and then run the program on a hardware platform of your choice. The algorithm to implement the TCP protocol can be programmed to run over an Ethernet network; in principle, it could also be implemented by carrier pigeon. Different implementation, different ‘program,’ different material substrate. For the same algorithm: there are good implementations and bad implementations (the algorithm might give you the right answer for any particular input but its flawed implementation incorporates some errors and does not); some implementations are incomplete; some are more efficient and effective than others. Human beings can implement algorithms; so can well-trained animals. Which brings us to computers and the programs they run.

The reason automation and the computers that deliver it to us are interesting and challenging–conceptually and materially–is because they implement algorithms in interestingly different ways via programs on machines. They are faster; much faster. The code that runs on computers can be obscured–because human-readable text programs are transformed into machine-readable binary code before execution–thus making study, analysis, and critique of the algorithm in question well nigh impossible. Especially when protected by a legal regime as proprietary information. They are relatively permanent; they can be easily copied. This kind of implementation of an algorithm is shared and distributed; its digital outputs can be stored indefinitely. These affordances are not present in other non-automated implementations of algorithms.

The use of ‘algorithm’ in the context of the debate over the legal regulation of automation is misleading. It is the ‘automation’ and ‘computerized implementation’ of an algorithm for credit scoring that is problematic; it is so because of specific features of its implementation. The credit scoring algorithm is, of course, proprietary; moreover, its programmed implementation is proprietary too, a trade secret. The credit scoring algorithm might be a complex mathematical algorithm readable by a few humans; its machine code is only readable by a machine. Had the same algorithm been implemented by hand, by human clerks sitting in an open office, carrying out their calculations by pencil and paper, we would not have the same concerns. (This process could also be made opaque but that would be harder to accomplish.) Conversely, a non-algorithmic, non-machinic–like, a human–process would be subject to the same normative constraints.

None of the concerns currently expressed about ‘the rule/tyranny of algorithms’ would be as salient were the algorithms not being automated on computing systems; our concerns about them would be significantly attenuated. It is not the step-by-step solution–the ‘algorithm’–to a credit scoring problem that is the problem; it is its obscurity, its speed, its placement on a platform supposed to be infallible, a jewel of a socially respected ‘high technology.’

Of course, the claim is often made that algorithmic processes are replacing non-algorithmic–‘intuitive, heuristic, human, inexact’–solutions and processes; that is true, but again, the concern over this replacement would not be the same, qualitatively or quantitatively, were these algorithmic processes not being computerized and automated. It is the ‘disappearance’ into the machine of the algorithm that is the genuine issue at hand here.

Tesla’s ‘Irma Update’ Shows The Dangers Of Proprietary Software

By now, you know the story. Tesla magically (remotely) updated the software of its cars during Hurricane Irma:

Tesla remotely sent a free software update to some drivers across Florida over the weekend, extending the battery capacity of cars and giving extra range to those fleeing Hurricane Irma.

According to reports, the update temporarily unlocked the full-battery potential for 75-kilowatt-hour Model S sedans and Model X SUVs, adding around 30 to 40 miles to their range.

“Cars with a 75-kilowatt-hour battery pack were previously software limited to 210 miles of driving range per single charge and will now get 249 miles, the full range capacity of the battery,” the company wrote on a blog.

As is evident from this description, the software regulating battery life is ‘autonomous’ of the user; the user cannot change it, or tweak it in any way to reflect changing user needs or driving conditions (like, say, the need to drive to a distant point in order to escape a potentially life-threatening change in the weather.) In short, the software that runs on Tesla’s cars is not ‘free‘–not in the sense that you have to pay money for it, but in the sense that you cannot do what you, as the user of the software, might or might not want to do with it: like share it, copy it, modify it. If the user needs ‘help’ he or she must wait for the benevolent corporation to come to its aid.

We, as software users, are used to this state of affairs. Most of the software we use is indeed not ‘free’ in this sense: the source code is kept a trade secret and cannot be inspected to figure out how it does what it does, the binary executables are copyrighted and cannot be copied, lastly, the software’s algorithms are patented. You cannot read the code, you cannot change it to better reflect your needs, and you cannot make copies of something you ‘own’ to give it to others who might need it. As software users eventually come to realize, you don’t ‘own’ proprietary software in the traditional sense of the term, you license it for a limited period of time, subject to many constraints, some reasonable, others not.

In an interview with 3AM magazine, while talking about my book Decoding Liberation: The Promise of Free and Open Source Software I had made note of some of the political implications of the way software is regulated by law. The following exchange sums up the issues at play:

3:AM: One aspect of the book that was particularly interesting to me was your vision of a world full of code, a cyborg world where ‘distinctions between human and machine evanesce’ and where ‘personal and social freedoms in this domain are precisely the freedoms granted or restricted by software.’ Can you say something about what you argued for there?

SC: I think what we were trying to get at was that it seemed the world was increasingly driven by software, which underwrote a great deal of the technology that extends us and makes our cyborg selves possible. In the past, our cyborg selves were constructed by things like eyeglasses, pencils, abacuses and the like—today, by smartphones, wearable computers, tablets and other devices like them. These are all driven by software. So our extended mind, our extended self, is very likely to be largely a computational device. Who controls that software? Who writes it? Who can modify it? Look at us today, tethered to our machines, unable to function without them, using software written by someone else. How free can we be if we don’t have some very basic control over this technology? If the people who write the software are the ones who have exclusive control over it, then I think we are giving up some measure of freedom in this cyborg society. Remember that we can enforce all sorts of social control over people by writing it into the machines that they use for all sorts of things. Perhaps our machines of tomorrow will come with porn filters embedded in the code that we cannot remove; perhaps with code in the browsers that mark off portions of the Net as forbidden territory, perhaps our reading devices will not let us read certain books, perhaps our smartphones will not let us call certain numbers, perhaps prosthetic devices will not function in ‘no-go zones’, perhaps the self-driving cars of tomorrow will not let us drive faster than a certain speed; the control possibilities are endless. The more technologized we become and the more control we hand over to those who can change the innards of the machines, the less free we are. What are we to do? Just comply? This all sounds very sci-fi, but then, so would most of contemporary computing to folks fifty years ago. We need to be in charge of the machines that we use, that are our extensions.

We, in short, should be able to hack ourselves.

Tesla’s users were not free during Irma; they were at the mercy of the company, which in this case, came to their aid. Other users, of other technologies, might not be so fortunate; they might not be the masters of their destiny.