Monday, March 29, 2010

Jamming, Blockers, and the Google-China Conflict

The ongoing struggle between Google, which wants its customers and readers in mainland China to have full access to all its services, and the government of the People's Republic of China, which legally requires Google to impose censorship on some aspects of what it provides, is unprecedented in scale and number of people involved, perhaps, but not in its essentials. Something very similar went on for decades on a smaller scale in a different medium: shortwave radio. But important technological differences in the two media make it harder for Google to do what shortwave broadcasters often succeeded in doing: namely, evading the efforts of hostile governments to "protect" their citizens from harmful foreign influences.

Around 1965, I somehow came into the possession of an old tabletop tube radio with a shortwave band. Back then, shortwave broadcasting was one of the few electronic media which naturally and inexpensively (at least for the user) gave one instant access to information from around the globe. Most nations of any size maintained some sort of shortwave broadcasting presence, and in particular the Communist-bloc countries kept up a barrage of English-language propaganda broadcasts aimed at Europe and the U. S. Radio Moscow came in regularly, and once or twice I even heard Radio Peking, as it called itself then. But even more prominent and easy to find at any time of the day or night were some strange buzz-saw noises. No, it wasn't interference from a neighbor's buzz saw. It was jammers: high-power shortwave stations designed explicitly to block reception of "undesirable" broadcasts, usually Voice of America programs. I don't have statistics to back this up, but it's likely that the Soviet Union and China spent more money on jammers than they ever did on propaganda broadcasts directed at foreign countries. It was more important to them to keep their own citizens from hearing unapproved information from the free world than it was to tell the free world about life behind the Iron Curtain.

By the nature of the medium, jamming was a dodgy and unreliable method. Broadcasters could and did change frequencies unexpectedly, forcing the jammer to take expensive steps to follow suit, and the vagaries of shortwave propagation meant that depending on where you were listening from—especially if you were not near a large city—you might hear Voice of America loud and clear and the jamming signal might not even be audible. But it must have been effective enough for jammers to clutter up the airwaves with dozens of buzz-saw sounds at all hours of the day and night, for years. Then the Internet came along.

The Web is both much better and much worse than shortwave broadcasting with regard to allowing free exchange of information across the boundaries of a nation which wants to control its media. It's better, in that more information can be posted on a single Web page than you could read over the air in half an hour or more. And since pictures are worth some large number of words, the Internet is much richer than any shortwave broadcast could ever be in that regard. But unfortunately for free interchange, the Web works through fiber-optic cables and equipment which must physically reside inside the receiving country. And the government which controls access and operation of that equipment has ultimately an absolute say over what the equipment transmits.

So if the government of China got up some fine day (to personalize an impersonal organization) and decided that it could do without all Internet connections to the outside world altogether, it could pull the plug and everything on the Web outside China would promptly disappear, without recourse. Something like that was never possible in the days of jamming, since one-hundred-percent leakproof jamming was economically and technically impossible. So Google, in its frustrated attempts to keep delivering its search engine and email services in China, has not been able to make that much progress within the mainland itself.

This is one reason why on March 23 Google announced a new approach: they were going to quit self-censoring their services and still stay within the letter of the Chinese law by moving their search services provided in the mainland to Google.com.hk in Hong Kong, where different and more liberal laws evidently still prevail. It is too soon to tell how this move will work out, although a website maintained by Google to show the status of various services within mainland China indicates that the government is blocking access to YouTube, many websites, Google's blogger function (on which this blog appears, unfortunately), and began interfering on Mar. 28 (yesterday) with certain mobile services.

Although I haven't listened to shortwave broadcasting much lately, it's unusual to hear any jamming anymore, partly because shortwave radio has been overtaken by the Internet and partly because the world's biggest jammer, the USSR, no longer exists. Jamming was always a shady business anyway. To my knowledge, the U. S. never engaged in jamming of broadcasts intended for the general U. S. public (although jamming is a time-honored practice in wartime electronic countermeasures to disable enemy radars and communications). Any government which insists on blocking the free flow of information to its citizens thereby pronounces a kind of self-delivered judgment on itself and betrays a kind of fundamental insecurity—a fear that the light of day will show its citizens their true situation and make them want to change it. Google's shift to Hong Kong may make things better for a while, but the only permanent solution is for the government of 1.3 billion Chinese citizens to change its ways.

Sources: Google's March 23 announcement of its changes in service is posted at http://googlepublicpolicy.blogspot.com/2010/03/new-approach-to-china-update.html, and the website for monitoring services available in China is at http://www.google.com/prc/report.html#hl=en.

Monday, March 22, 2010

U. S. Health Care Bill: Killing the Messenger

Last night I watched a tiny video image on my laptop that showed the floor of the U. S. House of Representatives. Superimposed on the image was a set of numbers tallying the votes for and against the famous (or infamous) health-care bill that has obsessed Washington for the last year or so. When the "yea" vote got to 216, I turned it off, knowing that the die had been cast. Whatever happens after this, the federal government has now taken to itself massive new powers over the one-sixth of the U. S. economy that is health care.

Part of the reason that health care costs so much more than it did decades ago, both proportionally and absolutely, is that owing largely to engineered products and services, doctors can do a lot more about disease more effectively than they used to. And medical technology is one of the shrinking number of areas in which U. S. manufacturing still leads the world. Our peculiar cobbled-together third-party way of paying for health care, which largely insulates the consumer against the real cost of medical technology, may be one reason that the U. S. is a favorable environment for developing such technology. Another may be the fruitful interaction between our research universities and medical companies, both of which benefit from federal and state research support as well as private research funding. Whatever the reasons, the healthy and innovative state of the medical technology industry is a bright spot in the otherwise murky health-care scene.

But that is not the way it is viewed in the philosophy embodied in the new health-care bill that just passed Congress. In this view, medical technology companies are unfairly profiting from the misery of the patients they serve, and should be penalized with a 2.3-percent excise tax on their products. Although the final form of this tax is not clear as of this writing, an earlier press account of it in December estimated it would cost the medical-device industry some $2 billion annually, rising to $3 billion in 2018.

It is not clear whether it was ever an ancient practice to kill the runner who brought bad news to the king. Even old-time despots probably understood the sheer wastefulness of such a move, let alone the difficulties it would cause in recruiting new messengers. But the vengeful, short-sighted, political-grandstanding attitude that leads to taxes such as the one on medical device companies is about as counterproductive as you can get, and shows a basic misunderstanding of the nature of business and private enterprise.

I am personally tone-deaf when it comes to entrepreneurship, but even I understand enough about it to see that medical-device companies, and health-insurance companies too for that matter, typically do the best they can within the regulatory and economic environment they work in. If MRI machines cost X hundred thousand dollars, and the MRI-machine market is doing well with several companies competing for good business, as I understand to be the case, then those who have the money to buy them (namely hospitals and clinics) must think the price is about right compared to the money they can make by using them. The rules and restrictions that people chafe about concerning health insurance arise largely from the way we have historically tied such insurance to employers, who "own" the policies rather than the individuals insured. Such third-party situations always lead to higher prices, and in my view this is the single most significant reason why health-care costs have risen faster than otherwise. If individuals owned their own health-care plans and decided in advance how much they wanted to spend, maybe there wouldn't be so many "Cadillac" plans out there and medical technology might be cheaper. But it would still be innovative, because the market would be free enough to reward innovation and competition.

The health-care bill that the president is so eager to sign moves things in the exact wrong direction: toward more bureucratic control, less freedom to do business in a way that makes economic sense, and more third-party interference between the consumer of health care and the provider. Yes, some more people will have insurance than before, and that is a worthwhile goal and almost the only good thing about the legislation. But there were far less intrusive, and probably more effective, ways of increasing the number of people who could afford health insurance. Only now, we may never know whether they might have worked. Instead, we have embarked on a huge, complex social experiment that will probably take decades to implement, and whose consequences we will all have to live with regardless of the outcome.

Besides the specific policy problems, the year-long debate over health care has further frayed the already worn fabric of our democracy. Congress has passed legislation that was clearly unpopular, and led to party-line divisions that may be very difficult to heal. Plus which, it may have contributed to an unhealthy cynicism about government in general. Last Friday I took the trouble to visit my congressman's office in downtown Austin to hand-deliver a note about the health-care vote. I had never done this before. It was rather depressing to see how the federal building is surrounded by a high black steel fence and guarded by cops who make you pass through a metal detector and run your personal items through an X-ray machine, just like at the airport. I understand the reasons for these precautions, but all the same, it seems like if the government had not expanded to be such a big target, we wouldn't have to go to so much trouble to protect it.

I will try to leave this topic alone in the future, since it is largely a done deal now and following it has not done my personal state of mind any good. But if anything comes up about it in the future that credibly concerns engineering ethics, I reserve the privilege to say something about it.

Sources: A description of last-minute changes to the medical-device tax was carried by the Huffington Post on Mar. 21 at http://www.huffingtonpost.com/2010/03/21/health-care-bill-insuranc_n_507506.html. The figures on total cost to the industry were from a Wall Street Journal article on Dec. 21, 2009 at http://online.wsj.com/article/SB10001424052748704398304574598130440164954.html.

Monday, March 15, 2010

Epistemology for Dummies, or, Where Do Engineers Come From?

My wife edits the About.com breast cancer website, and every so often she feels obliged to take down scurrilous rumors by running them to their source and shining the light of day on them. The one she's working on right now says that bras cause breast cancer. Turns out that an anthropologist and his wife published a book in 1995 saying just that, but apparently basing it on nothing more than some hazy suspicions about restraining lymph and breast cancer rates in industrial societies (in which women tend to wear bras more). Judging by the fact that this couple now lives on sixty-some acres in Hawaii, a lot of people must have bought the book.

This is an example of the type of thinking that cannot help you build a bridge, or design a computer, or develop a cure for cancer that will work. It begins from a feeling or a desire, even, rather than from scientifically gleaned data. If bras cause breast cancer, then there's something you can do to avoid it: don't wear a bra! Wouldn't it be nice if it were that simple? And evidently, at least some readers of the book seem to think it is.

Epistemology is the branch of philosophy dealing with how we know things. There are many ways of knowing, and I am not here to tell you that the scientific way—which means gathering carefully marshalled data, doing highly controlled experiments, and constructing theories based on reliable premises, all of this being a lot of tedious, difficult work—is the only way we can know things. But for certain questions, such as "do bras cause breast cancer?" the scientific way has proved itself to be better than the path taken by the aforementioned authors. I'm not sure what they did has a name, but let's call it the wish-fulfillment way. I'm also not sure that what they get out of it is knowledge—I'd prefer to call it a story. But it's a story that a lot of people want to hear.

If I were the kind of person who indulged myself in telling similar stories, there's a story I'd like to tell, and one that feels right to me. After doing or teaching engineering for some thirty-odd years, it seems to me that the kind of person who is good at engineering comes from a certain kind of background. This background includes, typically, a solid family structure (if possible, being raised by the person's biological father and mother), a culture in which few material things are taken for granted and actions are valued above words, and an atmosphere in which complex mental achievements are honored and praised. Up to the 1960s, these descriptions applied to most of the U. S., especially outside large coastal cities. But as average incomes rose and the culture changed significantly, this type of background became relatively rare.

Outside the States, however, especially in places where opportunities for higher education were strictly limited and highly competitive, these conditions still prevailed, or at least they did until recently. We are currently interviewing candidates for faculty openings in my engineering school. I haven't done a scientific study of the question, but I estimate that no more than 10% of the applicants were born in the U. S. A. All the rest are immigrants. I have no problem with this, because it's great for the immigrants and shows once again that America is the land of opportunity. But all the same, it's a curious phenomenon, and in my more pessimistic moments I want to explain it with the conclusion of this little story.

What if the blessings of modern technology and engineering lead inevitably after a generation or so to a population of individuals who are incapable of doing engineering? Instead, you get the kind of folks who will buy and believe a fuzzy-headed book about bras causing breast cancer. These kinds of people can still sell things. They can get retail jobs, or factory jobs (if there are any factories left), or they can even become good artists or politicians (especially politicians). But if on some fundamental level, they don't believe in the external, objective, stubborn, don't-care-what-you-think-but-that's-the-way-it-is-anyway reality that engineers have to deal with, they will not even consider engineering as a career, no matter how much money the National Science Foundation or anybody else spends on science, technology, engineering, and math education enhancement. And anyway, they wouldn't be any good at it if they tried.

In this dismal fantasy, the only reason we haven't run out of engineers yet is that we can still find people who are the products of cultures which haven't yet enjoyed the fruits of technology long enough to quit producing young men and women who can be engineers. But given enough time, and the spread of high-tech prosperity into enough places, we may run out.

There, that's my story that I don't exactly wish were true, but have often suspected of being true. Is it true? If I weren't an engineer, I might go out and write a fuzzy-headed wish-fulfillment-type book claiming that it is, and aimed at getting people all worked up about how bad it is and what we should do about it. But being an engineer, I know that it would take probably millions of dollars' worth of statistical, sociological, and longitudinal-study research even to come close to guessing about the truth of this idea. I have neither the support nor the skills to contemplate such research, so my story will have to just remain what it is—a story. The only way I can imagine finding out it's true without doing a lot of research is to let time pass and see what happens. By that time I will probably not be around to find out. But if it does happen, remember—you read it here first.

Sources: Pretty soon you should be able to read about bras and breast cancer at http://breastcancer.about.com./

Saturday, March 06, 2010

Toyota: Not So Fast, Guys

The automotive industry has furnished the field of engineering ethics with more than one "paradigm" case that generations of budding engineers have studied. In the way Toyota tried to deal with the now-infamous problem of unintended acceleration in many of its models, they have given us an example of how not to deal with a safety problem. Technical issues are only one aspect of the way an initially small-scale issue snowballed into a major financial disaster that shut down sales and may have permanently blotted Toyota's reputation. While the whole story has yet to emerge, it looks now as though a circle-the-wagons mentality was at least as much to blame as poor engineering.

We have the Los Angeles Times to thank for most of the initial journalistic spadework that pressured Toyota into reluctant action. Back in August of last year, an off-duty policeman and three of his family suddenly felt their new Lexus jet out of control at speeds of up to 100 mph before it hit another car, flew down an embankment, and caught fire, killing all four occupants. Investigation revealed that if a certain rubber floor mat is installed, a projection in the molding can catch the bottom tip of the rigid accelerator pedal so that the throttle is stuck wide-open. In late September, Toyota issued a floor-mat recall on over four million vehicles.

The August incident turned out to be the tip of an unintended-acceleration iceberg that Toyota has been struggling to minimize for over a decade, as further investigations published by the L. A. Times in October revealed. In October, Toyota sent out a letter to many of its car owners that the National Highway Traffic Safety Administration criticized four days later as "inaccurate" and "misleading." Toyota was clearly trying to minimize a problem that now appears to be more serious than simply a floor mat that interferes with accelerator-pedal motion. Some of the incidents dug up by the Times appear to involve the "drive-by-wire" electronic accelerator system which is, of course, operated by computer programs. Toyota admits that there is no software feature that disables the accelerator pedal when the brake pedal is pressed.

In all of the United States, Toyota has exactly one machine that can read a car's onboard data recorder to assist in diagnosing accidents and problems after the fact. This appears to be an attempt on the part of the automaker to control the production of potentially damaging information, since dealers and other interested parties cannot access the data. After a crash the day after Christmas in Texas that killed four people appeared to be due to unintended acceleration, things spun out of even Toyota's control. On January 26, they took the extraordinary step of halting sales of about half their models until the mechanical accelerator fix can be applied. It turns out this is at the request of the federal government, in a move that was probably motivated by words to the effect of, "if you don't do it by yourselves, we'll make you do it." As of this writing, the question of whether Toyota's electronic throttle system is to blame as well as the mechanical pedal problems is still under investigation. But there is plenty of smoke around to justify the conclusion that there's a fire in that department as well.

The phrase "damage control" can mean several things. If you're talking about something as straightforward as fighting a fire, it means putting the fire out as fast as you can to minimize physical damage. But the phrase has taken on a darker meaning in recent years. It has come to mean a strategy that an organization deploys in its public (and government) relations in order to convince outsiders that whatever really happened, things are not that bad after all. Although the story is still unfolding, there is enough evidence already to conclude that Toyota had taken the second meaning all too much to heart. In its misleading minimizing of the seriousness of recalls, in its failure to provide more than one data-reading device for the entire U. S., and in its Johnny-come-lately reactions to serious, long-term problems, it showed a signal lack of judgment and concern for the safety of the driving public. We do not know yet what Toyota knew about the problem, or problems, and when they knew it. But it is already clear that whatever they knew, they didn't want anyone else to know. It took a courageous news outlet (a print one, at that) to keep hammering on the issues and asking embarrassing questions, as well as a government agency that is arguably understaffed and underequipped to deal with the technical complexity of today's automotive industry. I wonder how long it would take a gang of clever Caltech undergrads to crack Toyota's automotive data-recorder code. If somebody hasn't done it already, they ought to as a public service.

The worst thing that can happen to a corporation (short of going out of business altogether) is to lose money, and Toyota has already been taken to the woodshed in that regard. But losing money is one thing, and losing one's life to a preventable technical defect is another order of thing altogether. Some estimates say that perhaps fifty or so people have died as a result of unintended acceleration in Toyota cars. That is a small fraction of the total of annual auto-related deaths in this country. But the engineers who could have prevented these deaths failed to do so. And that's what engineering ethics is about.

Sources: Motortrend Magazine online has a good chronology of the Toyota recall and related issues up to late January at http://www.motortrend.com/features/auto_news/2010/112_1001_toyota_recall_crisis/index.html. I also referred to a recent Los Angeles Times article on Congressional hearings about the matter at http://www.latimes.com/sns-ap-toyota-recall,0,7766384.story. For the record, I drive a Honda.