Monday, April 28, 2008

Should Google Be the World's Librarian?

Book Search is a portal that Google, Inc. is developing to provide access to all the world's books in digital form. How many is that? If you count editions (not individual copies), a recent Associated Press article about the project says there are between 50 and 100 million books in the world. The largest research library that I deal with on a regular basis, at the University of Texas at Austin, has only eight million of these. So clearly, Google will have done a great thing if and when it finishes—although with new books coming out all the time, a project like that is never really finished.

At first glance, this sounds like a great step forward in the history of information, on a par with the invention of printing. There are many parallels between the two events. Before movable type made it possible to produce thousands of identical copies of a manuscript, hand-copied books were rare, expensive treasures that only the wealthy and powerful classes could afford, by and large. But once Europe had dozens of print shops churning out books and pamphlets by the hundreds, prices came down to the point that artisans, shopkeepers, and even some farmers and peasants could afford them. You can make arguments that the Renaissance, the Protestant Reformation, and the Industrial Revolution all depended vitally on the invention of printing.

However, there is one critical difference between the invention of printing and what Google is doing. Print shops, publishers, and the whole network of book production, distribution, and the libraries that developed to house them were under the control of a diverse array of entrepreneurs, private organizations, schools, and governments. On the other hand, Google is, well, Google—a single, monolithic, centrally controlled corporation. Is there any ethical problem with that? It depends.

One thing that may be in danger is what I would term the universal freedom of library access. At any university library worthy of the name, anywhere in the world, any person can simply walk in and look at the general collections, generally without charge. And if you can produce scholarly credentials, you will usually be allowed to examine even the rarest items in their collections, under proper security controls, of course. The only limitation (and this is a severe one, admittedly) is that you have to travel physically to the library in question. But once you're there, you're in.

We have already seen how many internet firms have submitted to the will of dictatorial nations in exchange for the privilege of operating there. In my Mar. 30, 2006 blog, I criticized Google, Yahoo, and Microsoft for kowtowing to the government of the Peoples' Republic of China by restricting users' access to certain sites that the government deemed objectionable. Surely the books and other published works of Chinese dissidents will not be welcome there in electronic form any more than the people themselves, many of whom have endured long prison terms or even death for the "crime" of expressing their opinions.

But that is only one example of how Google, or any entity which has exclusive legal rights to the propagation of large amounts of information in a single medium, could distort or restrict access to the written heritage of the human race.

Am I being paranoid in sensing the potential for some sinister goings-on? I do not presently attribute evil or malign motives to Google, but sometimes things that look good to start with have bad unintended consequences. All I'm saying is that letting a single firm be in control of the way most of the world will in the future access its own written heritage, is at the least an unprecedented step, and potentially a very dangerous one.

The management of Google may all be nice folks now. But what if China gets more prosperous and has so much money in its government-controlled stock investment option that one day it hauls off and buys Google? Sounds ridiculous now, but if you had said in 1965 that in forty years, General Motors would be a money-losing basket case and Japanese car makers would beat them in worldwide sales, you would have gotten peculiar glances then too. Then China would get to say who gets access to what—an eventuality that few people would enjoy or benefit from.

My point is that the concentration of information control in the hands of a few is something to be regarded with caution, to say the least. Same goes for news media, but here we're talking a lot more than just news media—the intellectual heritage of the entire human race is at stake.

Do I have any suggestions? Well, no, in this case I'm just trying to get the ball rolling on a discussion. Even if I owned stock in Google, I have no illusions that they would listen to my opinions about their project. But if we're going to go ahead with this thing, we should at least go into it with our eyes open—as long as we can still see on our own.

Sources: The Associated Press article by Natasha Robinson on Google's Book Search project and its efforts toward the preservation of historical books was published in numerous venues. I saw it in print in the Austin American-Statesman (p. D3 of the Apr. 28, 2008 edition), and a version is accessible online at http://abcnews.go.com/Technology/wireStory?id=4722073.

Monday, April 21, 2008

Human Biological Enhancement and the Ethics of Personhood

Some philosophers of the mind like to try a little thought experiment on their students. It goes something like this. Suppose some years from now, a person—an ordinary human being—gets some dreaded brain disease that gradually destroys his gray matter. But also suppose that medical technology has advanced to the point that as the brain's biological tissue dies, it can be replaced by silicon (or some equivalent futuristic material) that is functionally equivalent to the dying brain part. And so as time goes on, Mr. Brain Patient has more and more of his brain replaced by the future's equivalent of computer chips. At what point, the philosopher asks, does the patient cease to be a human and begin to be a computer?

At one time, you could laugh off the whole thing by saying nobody has ever done such a thing and it's unlikely that they ever will. But no longer. Writing in Technology and Culture, historian Michael D. Bess points out that numerous blind and otherwise disabled people have received brain implants that allow them to see or communicate in ways that are utterly impossible for the rest of us mortals. Having a bunch of wires attached to your brain is not the same thing as replacing your cerebellum with a mainframe, but the border has been crossed. What happens from now on is more a matter of degree than of kind.

Bess foresees not just advances in brain science, but in genetic engineering and pharmacology as well, all leading to what he calls "human biological enhancement." Currently, the goal of most such projects is to use technology to restore the abilities of disabled people to something close to normal: curing genetic diseases, allowing the blind to see, allowing people with strokes or myasthenia gravis who end up "locked in" (unable to move or talk) to communicate via brain waves, and so on. But what is to prevent a person who sees through a computer from attaching an infrared camera to their input so they can see in the dark? Or what if we find a drug that restores Alzheimer's patients to normal brain function, and also gives normal people an IQ of 200? What is to keep us from taking human nature as merely raw material, a rough design to be improved on with increasingly advanced engineering? And what do we call these improved beings? People? Cyborgs? Or something in between?

Bess, for his part, sees no practical way to avoid these changes. The science will keep progressing, and as the natural desire on the part of people to take advantage of enhancements pulls the technology into the marketplace, we will face the issue of how to treat folks who have version numbers after their names (Bess titled his essay "Icarus 2.0"). He imagines that the only way to stop or regulate human biological enhancement would be to pass a worldwide set of laws together with a huge enforcement mechanism to chase down any miscreants trying to do enhancments under the table, so to speak. He sees the very public failure of the attempt to regulate performance-enhancing drugs in sports as a sign that this road is doomed to futility.

What we ought to do instead, he says, is get used to it. Start now to develop an "ethics of personhood" that in his words constitutes "an expanded conception of human dignity, a more generous understanding of the word 'us'." If one day you go to your job and find that the new hire you have to work with moves on wheels, sees through cameras, and accesses the Internet just by thinking, Bess is concerned that somehow you will be tempted to view that being as something other than human. We need to start now to work on that problem so that it doesn't lead to disastrous social consequences.

Well, I'm doing my little bit by drawing your attention to this matter. I'm already working with a colleague who gets around on wheels—he has osteomyelitis and spends most of his day in an electric wheelchair. Perhaps if these changes come along slowly enough, we can get used to them.

But for some reason, in searching history for an encounter between two very different orders of being who both happened to be human, the story of the early Spanish explorations of the New World comes to mind. With their armor, ships, and guns, the Spaniards must have looked to the native Americans like R2D2 looks to us. And sure enough, a whole lot of social disruption and suffering came about as a result of that encounter. But most of the misery and suffering was experienced by the native Americans, not the "enhanced" Spaniards.

Bess seems to be worried that un-enhanced humans will discriminate against the enhanced types, because they'll look odd or peculiar. But the case of Spanish exploitation of the New World suggests that the problems will mostly be experienced by those who, for whatever reason, don't benefit from technologically enhanced abilities. Especially if enhancement is expensive (it will always be at first), you could easily end up with an elite class of enhanced humans who would regard political and social power as their right.

Aldous Huxley's 1932 dystopia Brave New World divided the genetically engineered population of the future into alphas, betas, and gammas, as I recall. The alphas were the natural-born leaders with enhanced intelligence, and the gammas were bred (or manufactured, really) for menial jobs such as elevator operators (Huxley's crystal ball didn't include much in the way of automation). Huxley avoided the problem of having the gammas rise up in revolt when he made their genetic makeup include a natural-born enjoyment of menial tasks.

I don't know about you, but I wouldn't want to live in such a world. Bess is to be congratulated for raising a concern that we ought to start thinking about now. But I believe he's looking in the wrong places for problems. The enhanced types will do just fine—the people we need to start thinking about defending are the poor, the discriminated against, and the unborn, now and perhaps even more in the future.

Sources: Bess's essay "Icarus 2.0: A Historian's Perspective on Human Biological Enhancement" appears in the January 2008 issue of Technology and Culture (vol. 49, no. 1, pp. 114-126).

Monday, April 14, 2008

Thoughts on the Passing of a Zip Drive

In my household we try not to let too much old technology pile up, so after my wife bought a new laptop the other day, we began saying good-bye to her old Mac tower. It gave good service from about 2002 to a couple of years ago, and one of its features we're going to miss is its Zip drive. Zip disks were a removable magnetic-disk storage medium that were popular from the mid-nineties until flash drives came along. The first Zip disks held 100 MB, which was later boosted to 250 MB, but with 1-gig flash drives so cheap now I can't imagine there's much of a market for Zip drives now. Thing is, we have about 40 or so Zip disks that have stuff on them going all the way back to 1988, when my wife first learned to do graphics on a computer. Some of it has been backed up here and there, but if I had to tell you where, I'd be in trouble. So I spent yesterday afternoon transferring a good many of those old Zip disks to a backup drive, and it got me to thinking about the permanent impermanence of digital storage.

Every two to five years or so, a new generation of storage media come along. If the new generation didn't rise up and commit parricide on the previous generation, it wouldn't be so bad. But the hallmark of modern technology is "creative destruction," so for a new storage medium to be successful, it has to drive the previous medium out of existence. True, you can usually find antique drives, media, and even computers that use them if you look hard enough, but having to hunt around and assemble your own computer museum just to read some old files is hardly practical for most people. So the only alternative if you don't want your old data to go away as definitely as if you wrote it on paper and threw the paper on a bonfire, is to transfer it to the next medium. Which is fine for another two to five years, and then. . . .

And that gets me to wondering, what am I saving all this stuff for anyway? The inventor and futurist Ray Kurzweil wrote about this in one of the most human-sounding passages of a book about how we're all eventually going to live as software on hardware that will take over the universe (you think I'm kidding, go read The Singularity Is Near). His father Fredric was a musician and music teacher who fled Germany in the 1930s for the U. S. When he died at 58, the son inherited a large volume of paper documents, recordings, and other memorabilia. After starting a project to digitize all this stuff, Ray reached a conclusion which is as simple as it is startling. It was this: "Information lasts only so long as someone cares about it."

Like many of Kurzweil's philosophical epigrams, it contains elements of truth. I'm sure lots of information, in the form of paper, hard drives, old floppy disks, and so on, is eradicated every day simply because nobody needs or wants it any more, and the space or money it takes up is needed for something else. But just because somebody cares about information doesn't mean it will necessarily endure. Along with caring, the people interested in the data need the resources it takes to preserve it—whether that means space, funding for periodic migrations to new media, or archeological work.

In a way there's nothing new about this. People have been making choices about what information to save and what to toss ever since the invention of writing. Writing and paper are different in degree from Zip disks and flash drives, but not in kind. They are all technologies for the storage of a non-material entity—namely, information—using material media. You can make a good argument that the invention of writing made civilization possible, in that laws, history, customs, religious traditions, and most of what makes a culture could then be preserved independently of particular people with both good memories and the ability to pass their memories on to other people who could do the same. And I'm not one of these people who sit up at night worrying that historians of the future will have nothing to go on after the global catastrophe that wipes out all computer memories everywhere—although if that did happen, we'd all have a lot to worry about, not just the historians.

If we knew for certain whether anybody in the future would care about this or that data file, things would be easier. But you never know. Certain kinds of information, such as emails in the Executive Branch of the U. S. government, are just assumed to have historical importance, which is why the Bush administration got in some trouble a few months ago after admitting that they appear to have "lost" some emails covering several years, and had to recover them from backup tapes.

But for most ordinary, non-historical personages like myself, the candidates for people who will care about your information include yourself in the future, your relatives and children, and maybe a few friends and associates. It's actually a pretty short list. And unless you're a professional historian or plan to become the subject of one, if you don't think your list of carers-in-the-future would be interested in your tax return for 1982, you can just go ahead and throw it away.

Sources: Ray Kurzweil's The Singularity Is Near (Viking, 2005) carries the story of his attempts to archive his father's legacy on pp. 326-330. Zip is a registered trademark of Iomega Corporation, which still sells Zip drives, so maybe I won't worry about backing up those remaining disks just yet.

Monday, April 07, 2008

Whistleblowing on Southwest Airlines: Cracks of Doom or Paperwork Errors?

The lot of a whistleblower is not an easy one. And I'm not talking about football referees. In engineering ethics parlance, a whistleblower is someone who goes public with information about a safety issue, after trying without success to deal with the problem through normal organizational channels. Whistleblowers can toot either before or after something terrible happens, but the consequences for them are usually the same: isolation, criticism, and often the loss of a job or even a career. Their only compensation is the knowledge that, in most cases at least, they did the right thing.

Charlambe "Bobby" Boutris is finding out right now what life as a whistleblower is like. In 1998, the Federal Aviation Administration (FAA) hired him, and an important part of his job was to make sure that airlines complied with what are called Airworthiness Directives (ADs for short). These are rules that the FAA makes to ensure the safety of aircraft, and detail such things as regular fuselage inspections, especially for older planes.

You'd think nothing much could go wrong with the fuselage compared to moving parts like the engine and so on, but think again. If you've ever been on a jet aircraft and looked through a window with a view over the wing, you have probably noticed that the wingtip wiggles up and down several inches during air turbulence. That is perfectly normal, and designed into the way the plane works. If the wing was built solidly enough not to wiggle at all, it would make the plane so heavy that it couldn't get off the ground.

But if you've ever bent a paper clip back and forth until it breaks, you know about a thing called metal fatigue. And not only the wing, but all stress-bearing parts of the fuselage experience tiny movements that over time, can cause metal fatigue and cracks. Most of the time these cracks are small and don't spread. But in 1988, they were responsible for one of the most spectacular airline accidents in aviation history.

Passengers in the first-class section of an Aloha Airlines flight over Maui were astonished to see the roof of the plane pop off and rip away in the violent decompression, taking a flight attendant with it. The pilot, not even fully aware of what happened, quickly adapted to the altered flying characteristics of his plane and safely landed at a nearby airport. The attendant was the only fatality, but clearly, airlines did not want to take the chance of this kind of thing happening again. Investigation showed that the plane, which was one of the oldest in Aloha's fleet, had developed fatigue cracks that had spread to cause the whole top section of the fuselage to fly off.

For this and other very good reasons, the FAA requires air carriers to inspect their fleets for fatigue cracks on a regular basis. Now, these cracks are a statistical thing, like mortality rates. It's hard to predict whether a given plane will develop a crack at a given place by a given time, but the inspections are timed so that on average, any cracks can be caught and repaired well before they become dangerous. But the system works only if you keep to the schedule.

Well, it appears that Southwest Airlines didn't keep to the inspection schedule. In testimony before Congress on April 4, Inspector Boutris told the story of how he found numerous cases in which inspection records were either too mixed up to tell whether the inspections had been done, or showed definitely that planes had gone as long as 30 months past the time when ADs specified they had to be pulled out of service to be inspected. It's illegal to fly a plane in revenue service if it's behind in certain kinds of inspections.

What made matters worse was that when Boutris asked permission from his FAA supervisor to issue a letter of investigation to Southwest in 2007, the supervisor told him to tone it down to a letter of concern, which does not carry the same impact. Eventually, in late March of 2007, Southwest did finish up the late inspections, but only after some airplanes had gone months or years without them. The FAA has announced its intention to fine Southwest ten million dollars for flying the uninspected planes, at least one of which was found to have fatigue cracks after inspections were finally performed.

On a scale of "who cares?" to "stick it to 'em," you can identify two extremes of how one can view this story. If you take the side of Southwest Airlines, you can point out that besides being one of the most profitable airlines in the business, they have never had a catastrophic accident in which more than one person was killed. And that incident, when a ground crew member was pulled into an engine, was due to pilot error, not mechanical failure. True, they didn't follow all the rules, but no harm was done—none of their planes popped their tops like the Aloha Airlines flight did.

On the other extreme, you can say that you keep good safety records like that by following the rules, even if it means grounding a large fraction of your fleet to make overdue inspections. The attitude of Boutris' supervisor appears to be one of "don't rock the boat," which might indicate that he was more concerned with how Southwest Airlines would fare than he was worried about the safety of the flying public, despite the fact that he worked for the government. That indicates systemic organizational problems both within the FAA and Southwest Airlines.

Back in high school, I attended Explorer Scout meetings that were held in the basement of a telephone exchange building. On the wall of the break room was a brass plaque, as I recall, and its words went something like this: "No service is so urgent or no business need is so critical that we fail to perform our work safely." Back then, Ma Bell had a guaranteed monopolistic income, and could afford to make safety priority number one. But I thought it was a great motto at the time, no matter what the business was or how it was doing financially. And I still do. I hope Southwest Airlines agrees with me, not just in words, but in actions as well.

Sources: A video of Mr. Boutris' opening statement before a Congressional committee investigating this matter can be viewed at http://salon.glenrose.net/?view=plink&id=6899. A CNN article on the Southwest Airlines actions and the FAA's response is at http://www.cnn.com/2008/US/03/06/southwest.planes/. The Wikipedia article on Aloha Airlines has a brief description of the 1988 accident.