Tolstoy, Berlin and Artificial Intelligence
I wonder what Isaiah Berlin would have had to say about artificial intelligence. This occurred to me as I sat listening to presentations on big data and the Fourth Amendment at a recent conference in New York City sponsored by the National Association of Criminal Defense Lawyers. It turns out we’re all blips on someone’s screen, and there doesn’t seem much we can do about it.
Lawyers in general do not read enough. Oh, we read plenty of case law, treatises, pleadings – all the flotsam and jetsam that make the law a semi-scholarly profession. But what we don’t read enough of are essayists, historians, poets, philosophers. The result is a sort of tunnel vision. And, I might add, turgid and scholastic Supreme Court opinions that promote neither respect for, not understanding of, the law by non-lawyers.
Lawyers ought to be required to spend a certain amount of time each week reading outside the law. What drives litigation, and the need for lawyers, is the raw material of life. Lawyers try to harness this energy and direct it into peaceful and largely civilizing channels. We’d benefit from more time swimming in the currents that drive clients to our doors.
So let me recommend Isaiah Berlin. We can start with one of his most famous, and accessible, essays: “The Hedgehog and the Fox.”
Berlin was an Oxford don and intellectual historian who wrote scores of essays and books in the second half of the twentieth century. He was what some call a political philosopher, a person who understood politics as Aristotle understood politics: the pursuit of knowledge about what constitutes a good life, both individually, and collectively.
The “Hedgehog and the Fox” was written in the 1950s. Berlin famously distinguished between hedgehogs, folks dedicated to the pursuit of one great, unifying truth, or vision about the good life, and the fox, folks content to master many things, understanding that the whole shall forever be beyond their grasp. Berlin writes of the monism and the pluralism.
Pluralism is, of course, a staple of liberalism. We believe that so long as folks are free to pursue their own individual ends, truth will, in the end, prevail. Liberty serves as the means to the good life. Monists can threaten the liberal project by imposing their vision of the good, thus the danger of totalitarianism.
But does this mean that there is no larger truth in the world? Pluralism was never intended by its founders to be an end in itself. John Stuart Mill’s classic work, On Liberty, supported freedom of speech as a means of setting the marketplace of ideas free from the dead hand of censors. Why? So that we could better learn the truth about ourselves and our world.
All pluralists, it turns out, are agnostic monists. There are many roads to the one place we all seek: the truth about the good life.
Berlin’s essay is really a meditation on Leo Tolstoy’s War and Peace. Totstoy was a virtuoso fox, dissecting the French invasion of Russia under Napoleon in minute imaginative detail. We might call his critical perspective granular. He was a master of portraying the lived, and idiosyncratic, lives of individual participants. He was a fox confidently describing the manifold lives caught up in the conflict.
He scorned hedgehogs, the theoreticians who described events in terms of grand political, military or economic theory. The war tumbled on almost as if inevitable; men of ideas tried to harness the fury. But these failed hedgehogs could never comprehend the minute, and infinite, causes of things. Understanding experience is simply too vast an undertaking. It is best, Tolstoy thought, to be moderate in what we expected. He drew inspiration from the felt sense of decency of common men and women.
What has this to do with artificial intelligence?
Tolstoy thought reason insufficient to capture the full sense of reality. There were too many causes, too many events – too many variables for the mind to master. But he did not deny that there was a reality. And that’s the point.
Google, Amazon, Facebook – the algorithmic masters assembling seemingly infinite data points can and do capture a far more robust sense of all the variables that constitute reality than can any individual mind. Police departments and governments now engage in predictive policing, aggregating data about our past behavior to predict what we are likely to do in the future. Is it possible that as computing power increases our world will be down with precision? What truth will that reveal? Can we bear it?
And what then of our cherished sense of liberty, of autonomy, of dignity?
I’ve been watching the opiod crisis with an increasing sense of foreboding. Do happy and productive people choose to dull the pain of living with painkillers? What if in a world increasingly understood, and mastered, by machine learning, the machines reach a point of superintelligence – what commentators call singularity? Do the machines then decide we humans are inefficient and unnecessary? Is the opiod crisis a thinning of the herd?
Dark thoughts, I know. I’d love to know what Isaiah Berlin would have made our madcap world. Few write as well and as piercingly as did he.
Tolstoy, Berlin wrote, was a frustrated monist. He believed there was a reality, and that wisdom meant accepting its limits without seeking to impose false theories on the raw data of experience. That seems somehow quaint in the age of AI. What if the machines see more clearly than do we? What if every cause can be plumbed, and freedom is shown, after all, to be an illusion?
It’s been years since I spent serious time with Isaiah Berlin. It’s time for a return to this wise man’s writings. Next up, another classic of his: “Two Concepts of Liberty.”