Hannah Arendt Center for Politics and Humanities
7Jun/130

In the Age of Big Data, Should We Live in Awe of Machines?

ArendtWeekendReading

In 1949, The New York Times asked Norbert Wiener, author of Cybernetics, to write an essay for the paper that expressed his ideas in simple form. For editorial and other reasons, Wiener’s essay never appeared and was lost. Recently, a draft of the never-published essay was found in the MIT archives. Written now 64 years ago, the essay remains deeply topical. The Times recently printed excerpts. Here is the first paragraph:

By this time the public is well aware that a new age of machines is upon us based on the computing machine, and not on the power machine. The tendency of these new machines is to replace human judgment on all levels but a fairly high one, rather than to replace human energy and power by machine energy and power. It is already clear that this new replacement will have a profound influence upon our lives, but it is not clear to the man of the street what this influence will be.

Wiener draws a core distinction between machines and computing machines, a distinction that is founded upon the ability of machines to mimic and replace not only human labor, but also human judgment. In the 1950s, when Wiener wrote, most Americans worried about automation replacing factory workers. What Wiener saw was a different danger: that intelligent machines could be created that would “replace human judgment on all levels but a fairly high one.”  

Today, of course, Wiener’s prophecy is finally coming true. The IBM supercomputer Watson is being trained to make diagnoses with such accuracy, speed, and efficiency that it will largely replace the need for doctors to be trained in diagnostics.

watson

Google is developing a self-driving car that will obviate the need for humans to judge how fast and near to others they will drive, just as GPS systems already render moot the human sense of direction. MOOCs are automating the process of education and grading so that fewer human decisions need to be made at every level. Facebook is automating the acquisition of friends, lawyers are employing computers to read and analyze documents, and on Wall Street computer trading is automating the buying and selling of stocks. Surveillance drones, of course, are being given increasing autonomy to sift through data and decide which persons to follow or investigate. Finally, in the scandal of the day, the National Security Agency is using computer algorithms to mine data about our phone calls looking for abnormalities and suspicious patterns that would suggest potential dangers. In all these cases, the turn to machines to supplement or even replace human judgment has a simple reason: Even if machines cannot think, they can be programmed to do traditionally human tasks in ways that are faster, more reliable, and less expensive than can be done by human beings. In ways big and small, human judgment is being replaced by computers and machines.

It is important to recognize that Wiener is not arguing that we will create artificial human beings. The claim is not that humans are simply fancy machines or that machines can become human. Rather, the point is that machines can be made to mimic human judgment with such precision and subtlety so that their judgments, while not human, are considered either equal to human judgment or even better. The result, Wiener writes, is that “Machines much more closely analogous to the human organism are well understood, and are now on the verge of being built. They will control entire industrial processes and will even make possible the factory substantially without employees.”

Wiener saw this new machine age as dangerous on at least two grounds. First, economically, the rise of machines carries the potential to upend basic structures of civilization. He writes:

These new machines have a great capacity for upsetting the present basis of industry, and of reducing the economic value of the routine factory employee to a point at which he is not worth hiring at any price. If we combine our machine-potentials of a factory with the valuation of human beings on which our present factory system is based, we are in for an industrial revolution of unmitigated cruelty.

The dangers Wiener sees from our increased reliance on computing machines are not limited to economic dislocation. The real threat that computing machines pose is that as we cede more and more power to machines in our daily lives, we will, he writes, gradually forfeit our freedom and independence:

[I]f we move in the direction of making machines which learn and whose behavior is modified by experience, we must face the fact that every degree of independence we give the machine is a degree of possible defiance of our wishes. The genie in the bottle will not willingly go back in the bottle, nor have we any reason to expect them to be well disposed to us.

In short, it is only a humanity which is capable of awe, which will also be capable of controlling the new potentials which we are opening for ourselves. We can be humble and live a good life with the aid of the machines, or we can be arrogant and die.

For Wiener, our eventual servitude to machines is both an acceptable result and a fait accompli, one we must learn to accept. If we insist on arrogantly maintaining our independence and freedom, we will die. I gather the point is not that machines will rise up and kill their creators, but rather that we ourselves will program our machines to eliminate, imprison, immobilize, or re-program those humans who refuse to comply with paternalistic and well-meaning directives of the machines systems we create in order to provide ourselves with security and plenty.

Wiener counsels that instead of self-important resistance, we must learn to be in awe of our machines. Our machines will improve our lives. They will ensure better medical care, safer streets, more efficient production, better education, more reliable childcare and more human warfare. Machines offer the promise of a cybernetic civilization in which an entire human and natural world is regulated and driven towards a common good with super-human intelligence and calculative power. In the face of such utopian possibility, we must accept our new status as the lucky beneficiaries of the regulatory systems we have created and humble ourselves as beings meant to live well rather than to live free.

tech

Recent revelations about the U.S. government’s using powerful computers to mine and analyze enormous amounts of data collected via subpoenas from U.S. telecom companies is simply one example of the kind of tradeoff Wiener suggests we will and we should make. If I understand the conclusions of Glenn Greenwald’s typically excellent investigative reporting, the NSA uses computer algorithms to scan the totality of phone calls and internet traffic in and out of the United States. The NSA needs all of this data—all of our private data—in order to understand the normal patterns of telephony and web traffic and thus to notice, as well, those exceptional patterns of calling, chatting, and surfing. The civil libertarian challenges of such a program are clear: the construction of a database of normal behavior allows the government to attend to those whose activities are outside the norm. Those outliers can be terrorists or pedophiles; they may be Branch Davidians or members of Occupy Wall Street; they may be Heideggerians or Arendtians. Whomever they are, once those who exist and act in patterns outside the norm are identified, it is up to the government whether to act on that information and what to do with it. We are put in the position of having to trust our government to use that information wisely, with pitifully little oversight. Yet the temptation will always be there for the government to make use of private information once they have it.

In the face of the rise of machines and the present NSA action, we have, Wiener writes, a choice. We can arrogantly thump our chests and insist that our privacy be protected from snooping machines and governmental bureaucracies, or we can sit back and stare in awe of the power of these machines to keep us safe from terrorists and criminals at such a slight cost to our happiness and quality of life. We already allow the healthcare bureaucracy to know the most intimate details of our lives and the banking system to penetrate into the most minute details of our finances and the advertising system to know the most embarrassing details of our surfing and purchasing histories; why, Wiener pushes us to ask, should we shy away from allowing the security apparatus from making use of our communication?

If there is a convincing answer to this hypothetical question and if we are to decide to resist the humbling loss of human freedom and human dignity that Wiener welcomes, we need to articulate the dangers Wiener recognizes and then rationalizes in a much more provocative and profound way. Towards that end, there are few books more worth reading than Hannah Arendt’s The Human Condition. Wiener is not mentioned in Hannah Arendt’s 1958 book; and yet, her concern and her theme, if not her response, are very much in line with the threat that cybernetic scientific and computational thinking pose for the future of human beings.

In her prologue to The Human Condition, Arendt writes that two threatening events define the modern age. The first was the launch of Sputnik. The threat of Sputnik had nothing to do with the cold war or the Russian lead in the race for space. Rather, Sputnik signifies for Arendt the fact that we humans are finally capable of realizing the age-old dream of altering the basic conditions of human life, above all that we are earth-bound creatures subject to fate. What Sputnik meant is that we were then in the 1950s, for the first time, in a position to humanly control and transform our human condition and that we are doing so, thoughtlessly, without politically and thoughtfully considering what that would mean. I have written much about this elsewhere and given a TEDx talk about it here.

The second “equally decisive” and “no less threatening event” is “the advent of automation.”  In the 1950s, automation of factories threatened to “liberate mankind from its oldest and most natural burden, the burden of laboring and the bondage to necessity.” Laboring, Arendt writes, has for thousands of years been one essential part of what it means to be a human being. Along with work and action, labor comprises those activities engaged in by all humans. To be human has meant to labor and support oneself; to be human has for thousands of years meant that we produce things—houses, tables, stories, and artworks—that provide a common humanly built world in which we live together; and to be human has meant to have the ability to act and speak in such a way as to surprise others so that your action will be seen and talked about and reacted to with a force that will alter the course and direction of the human world. Together these activities comprise the dignity of man, our freedom to build, influence, and change our given world—within limits.

But all three of these activities of what Arendt calls the vita activa, are now threatened, if not with extinction, then at least with increasing rarity and public irrelevance. As automation replaces human laborers, the human condition of laboring for our necessary preservation is diminished, and we come to rely more and more on the altruism of a state enriched by the productivity of machine labor. Laboring, part of what it has meant to be human for thousands of years, threatens to become ever less necessary and to occupy an ever smaller demand on our existence. As the things we make, the houses we live in, and the art we produce become ever more consumable, fleeting, and temporary, the common world in which we live comes to seem ever more fluid; we move houses and abandon friends with the greater ease than previous ages would dispose of a pair of pants. Our collective focus turns toward our present material needs rather than towards the building of common spiritual and ethical worlds. Finally, as human action is seen as the statistically predictable and understandable outcome of human behavior rather than the surprising and free action of human beings, our human dignity is sacrificed to our rational control and steering of life to secure safety and plenty. The threat to labor, work, and action that Arendt engages emerges from the rise of science—what she calls earth and world alienation—and the insistence that all things, including human beings, are comprehensible and predictable by scientific laws.

Arendt’s response to these collective threats to the human condition is that we must “think what we are doing.” She writes at the end of her prologue:

What I propose in the following is a reconsideration of the human condition from the vantage point of our newest experiences and our most recent fears. This, obviously, is a matter of thought, and thoughtlessness—the heedless recklessness or hopeless confusion or complacent repetition of “truths” which have become trivial and empty—seems to me among the outstanding characteristics of our time. What I propose, therefore, is very simple: it is nothing more than to think what we are doing.

Years before Arendt traveled to Jerusalem and witnessed what she saw as the thoughtlessness of Adolf Eichmann, she saw the impending thoughtlessness of our age as the great danger of our time. Only by thinking what we are doing—and in thinking also resisting the behaviorism and materialism of our calculating time—can we humans hope to resist the impulse to be in awe of our machines and, instead, retain our reverence for human being that is foundation of our humanity. Thinking—that dark, irrational, and deeply human activity—is the one meaningful response Arendt finds to both the thoughtlessness of scientific behaviorism and the thoughtlessness of the bureaucratic administration of mass murder.

think

There will be great examples of chest thumping about the loss of privacy and the violation of constitutional liberties over the next few days. This is as it should be. There will also be sober warnings about the need to secure ourselves from terrorists and enemies. This is also necessary. What is needed beyond both these predictable postures, however, is serious thinking about the tradeoffs between our need for reliable and affordable security along with honest discussion of what we today mean by human freedom. To begin such a discussion, it is well worth revisiting Norbert Wiener’s essay. It is your weekend read.

If you are interested in pursuing Arendt’s own response to crisis of humanism, you can find a series of essays and public lectures on that theme here.

-RB

14Dec/120

The Dorm Wars

The Dorm Wars have not yet caused the numerous bankruptcies amongst minor and maybe even some more established colleges that seem inevitable. What they have done is change the nature of college education. Whether at Harvard or Ramapo, students want luxury dorms with private bathrooms and glitzy campus centers. And since students—fueled with cheap student debt are the all-powerful consumers—campus administrators have followed the money. Unfortunately, they also too often followed their students into debt. As the NY Times reports today,

A decade-long spending binge to build academic buildings, dormitories and recreational facilities — some of them inordinately lavish to attract students — has left colleges and universities saddled with large amounts of debt. Oftentimes, students are stuck picking up the bill.

I recently visited my alma mater for a reunion and was housed in the building where decades ago I labored long into the night as an editor for my beloved Prism magazine. It is the dorm in which I once put my hand through a glass door in the midst of a late-night editing and layout session.  I barely recognized the Pratt Dormitory, which resembled more a Tablet style hotel than a college dormitory.

Such lavish quarters are now seen as necessary to attract the best students—something that is sad if it is true. And this perception, true or false, has unleashed the dorm wars. Some colleges, like the one I attended, don’t need to borrow to build. But many others think that they do.

“If Ramapo College was going to respond to what students wanted, which was larger, more comprehensive programs and residential housing, then we were going to have to go out and borrow,” Peter P. Mercer, President of the public liberal arts college in New Jersey told the Times

How wrong is that. Borrowing can of course be justified. But if you want to build something, there are other options. You can, for example, go out and raise the money. That requires work, convincing people, many of whom have no personal connection to your college, that what you are doing is important and worthy of support. Excessive borrowing is, too often, the resort of those unwilling to take the longer and yet more responsible path of building an institution that people are willing to invest in and support.

More importantly, the enormous borrowing of colleges reported by the Times is evidence of an educational system that has simply lost its way.  The fastest growing costs at colleges across the country are for administrators and for capital projects. Much of the borrowing is financing new luxury buildings and a bloated services staff. The priorities are wrong and real focus on teaching and learning seems to have been largely ignored. As students and parents confront extraordinary costs that go increasingly to pay interest on debt and support lavish undergraduate living, many are increasingly rebelling.

And for the first time in generations, students have other options. The rise of Internet learning is going to disrupt college education in this country as the Internet has transformed nearly every other area of life. And it will do so at the very moment when the finances of colleges and universities around the country are shakier than they have been in generations. The shake out will be painful.

What needs to be thought here is what is it that allowed debt to become so infectious within and amongst our educational institutions. With $1 trillion in student debt and $200 billion in institutional debt, education more and more resembles the housing and financial sectors of our economy.

Education is supposed to be a conservative enterprise, a bastion of learning and teaching the accumulated history and knowledge of the past. Somewhere along the line, education changed from being an experience of teaching and forming young individuals and citizens and became something very different. Higher education is now a progressive launching pad for careers. It is job security for tenured professors. It is the center of research and the producer of valuable sports franchises. Lost in the mix, I fear, is original mission itself.  Just as banks and financial institutions abandoned their old job of lending and saving money and sought to become investment banks, so too have colleges changed from being educational institutions to being consumer brands selling luxury and success instead of the life of the mind. Some can do both. But many more will go the way of Pan Am and Hostess.

-RB

25Apr/122

Pension Crisis Primer

I have been trying to understand our pension mess for years now, and I will tell you that the numbers and acronyms are at times baffling. But help is here.

Two of the nation's Federal Reserve Banks (the Cleveland and the Atlanta Federal Reserve Banks) have joined together to form a "Financial Monitoring Team to study pension funds and municipal finance with an eye toward implications for the wider economy and financial system." In other words, these two banks are seeking to shine a light on the dark and difficult to understand corners of municipal finance. Chiefly, the banks are interested in learning about Municipal Pensions.

The Cleveland Fed publishes a newsletter, "Forefront", and the latest issue contains a number of incredibly helpful articles about the state of municipal pensions.  The main article is: Public Finances: Shining Light on a Dark Corner. This is a clear and helpful article, with a glossary and helpful sidebars. It also comes with a video primer on the pension crisis that is sober, clear, and helpful.

One question the article addresses is just how big the pension shortfall actually is. According to government numbers, the shortfall is $800 billion. The government estimates are based on assumptions of an 8% rate of return, which inflates the assumptions about the present value of pensions. In all likelihood, the return will be somewhere between the 8% historic average and the painfully low return offered by persistently low interest rates.  Thus, many private economists estimate the shortfall at around $4 Trillion. Here is what the Cleveland and Atlanta Feds say:

Some economists, however, have come up with a $4 trillion shortfall. They have pointed out that for most state and local plans, promised pension benefits are protected by constitutional, statutory, or common law guarantees. (See related article, “Navigating the Legal Landscape for Public Pension Reform.”) By definition, this ought to make them riskless obligations to the pensioners. Thus, the appropriate valuation methodology should discount promised benefits using the risk-free interest rate, usually calculated as the yield on long-term U.S. Treasuries. This method, argued cogently by Jeffrey Brown and David Wilcox in “Discounting State and Local Pension Liabilities” (2009), has the virtue of being supported by both economic and legal principles. It also produces substantially higher estimates of the present value of pension liabilities. Given the currently low yields on Treasury bonds, this approach implies a present value of accrued obligations as high as $6.7 trillion, leaving an unfunded liability of $4 trillion.

In other words, the actual size of the pension shortfall is probably somewhere between $800 Billion (the size of the 2009 stimulus package) and $4 Trillion. The likely shortfall, as the Fed says in its video on the site, is in the $3-$4 Trillion range.

So what does this mean? The Cleveland and Atlanta Feds offer a few conclusions:

1. At this point, it seems unlikely that any major pension fund will run out of cash in the next few years, barring a general worsening of economic and financial condition.

2. But we are not out of the woods yet. Many funds will require significant reforms to reduce underfunding levels, with painful new contributions from employers and employees.

3. Another concern is that some states’ legal protections may be too strong to give reforms enough time and flexibility to put plans on sustainable paths. In that case, states would ultimately be on the hook for covering pension benefits out of general revenues. This scenario, by creating crisis conditions in those states, could stress economic conditions more generally.

The real problem is the combination of #2 and #3. For if state laws make it too difficult to cut or reduce pensions, the only option is "painful new contributions from employers or employees." It may be that we cut the guaranteed pensions of pensioners, making them less well off in retirement. That would hurt the workers. Or, if legal protections prevent that option, we the taxpayers will have to dig deep to pay their pensions, probably as we at the same time cut other essential services. And that will not be pretty. Either way, the state and local government crisis is shaping up to be one of the most important challenges of our generation.

For more on the pension crisis, you can revisit our other posts on the subject here.

—RB