I'm going to go out on a limb here. But I'm sure that the limb's pretty solid.
On May 15, the California Supreme Court struck down a ban on same-sex marriages that the state has had in place for some time. I'm not going to talk about the issue of judicial activism, or the question of whether California's citizens will assert their rights to reverse this action by approving a referendum amending their state constitution next fall. Instead, I am going to argue that allowing same-sex marriage will endanger the future of the engineering profession in this country.
Seems like a stretch, doesn't it? Here is my line of reasoning.
First, let me show that allowing same-sex marriages damages the institution of marriage. Some people simply do not see how conventional marriages between a man and a woman are in any way affected if we also let men marry men or women marry women. For these readers, let me make an analogy.
We have a nice solid base of well-functioning, highly capable engineering colleges in the U. S. Most of them are accredited through a rigorous process of inspections, visits, and continuous improvement. Suppose we passed a law that said all employers must recognize engineering degrees from any institution calling itself a college of engineering, whether it was accredited or not. It would be illegal to refuse to hire an engineer simply on the basis of what college he or she got a degree from. (=all of society must recognize marriage certificates of all kinds, whether for same-sex marriages or not.) We would leave the whole accreditation machinery in place, and universities capable of giving a good accredited education would still be able to do so. (=men and women who want to marry the opposite sex can still do so.)
What do you think would happen to the institution of engineering higher education in this country? Outfits handing out engineering degrees would spring up like newsstands on every corner, and students would flock to them. The average competency level of degree-holding engineers in this country would go into a precipitate decline, and the whole process of engineering education might undergo permanent damage that would take years or decades to repair, if ever. And note: in this hypothetical scenario, we did nothing whatever to the good schools. They were still free to stay accredited and do their good, competent job. We simply forced everyone to recognize the fly-by-night institutions as competent, but they were in fact incompetent.
The adjective "incompetent" often carries negative connotations, but it need not do so. It simply means that the noun modified is incapable of doing something or other. I have no shame in admitting the fact that, being a male, I am incompetent to bear a child. Women are incompetent to beget children without a male being involved somewhere along the line. And two men together, or two women together, are incompetent when it comes to fulfilling the practical duties and responsibilities of marriage, namely: being a biological and social unit that consists of a man as father, a woman as mother, and children who each have the same mother and father.
There are many scientific studies—thousands, in fact—performed by sociologists with all kinds of backgrounds and personal beliefs, which examine the question, "Do children who grow up in a family consisting of one mother and one father who are married and stay married, do better than children raised in any other kind of environment?" To qualify "better" you can look at social adjustment, criminal records, levels of school achievement, early or frequent sex and drug use, rates of depression and suicide, and so on. And the resounding, repeatable, monotonously consistent answer is, "Yes." This is not to say that kids raised by a single mother or two gay men are doomed to failure and a miserable existence. The human spirit can triumph over adversity of whatever kind. But when children are examined in statistically significant numbers, there is no question that the social institution we call conventional intact marriage beats any other way of raising children hands-down. That is not an ideological statement. It is a social-science statement backed up by years of the best kind of research that social science can offer these days. If you don't believe me on that, see David Blankenhorn's The Future of Marriage.
Now for the connection to engineering. It is my subjective impression, which I wish some social scientist would check out with the machinery of their trade, that the better grade of engineering students come from just the kind of stable family background that same-sex marriage will militate against. The National Science Foundation, among other institutions in this country, is concerned that very few students of either sex (and especially few women) choose engineering as an undergraduate degree, and even fewer decide to go on to graduate school. This is why it is increasingly rare to find engineering professors who were born in the U. S., because whatever mysterious factor it is that makes people want engineering graduate degrees is in short supply in this country, but seems to be plentiful abroad.
I will not claim that unstable marriages, divorced and remarried couples, single parents, and same-sex parenting is responsible for the entire decline in interest in engineering among young people in the U. S. But I believe a part of it is. And if we damage the institution of marriage further by insisting that same-sex unions get the same recognition as conventional marriages, I forecast a worsening dearth of U. S. students able to muster the discipline and deferred gratification necessary to pursue careers in engineering. I suspect we will wait a long time before the National Science Foundation comes out in opposition to same-sex marriage. Nevertheless, if I'm right, it might do more good for them to work in that direction than to spend their money on some of the programs they have supported in the past to encourage students to become engineers.
There. I made the connection. Like it, hate it, argue with it as you will. But that is my opinion, and as far as the marriage part goes, I'm on solid ground, not hanging from a tree by a limb.
Sources: Although I have not read the book, David Blankenhorn's The Future of Marriage comes highly recommended as a careful, scientifically reasoned argument written by a person who favors equal rights for homosexuals, but is convinced by scientific evidence that same-sex marriage would be too high a price to pay.
Saturday, May 31, 2008
Sunday, May 25, 2008
Remembering Brian O'Connell
Last Thursday, May 22 brought the sad news of the passing of Brian O'Connell the previous day. Anyone who knew Brian, or met him even once, was not likely to forget him. For those of you who did not have the privilege of meeting him, I would like first to offer you my sympathy. Then I will try to describe one of the most colorful personalities ever to grace the field of engineering ethics.
This business tends to attract people with mixed backgrounds who are both conversant with the intricacies of some technical field and also interested in the human side of things. Brian was no exception. He once told me he was one of the youngest people ever to run a planetarium show at Hartford's Gengras Planetarium, when as a young teenager he was asked to fill in for the regular operator whom Brian had become friends with. But his interest in the depths of the human soul expressed itself soon thereafter when he attended seminary for a while. Deciding he wasn't quite cut out to be a priest, he switched to computer science, and then back to humanities as he took a law degree and practiced law for several years. Eventually he joined Central Connecticut State University and served with distinction in both their computer science and philosophy departments.
I met Brian shortly after he discovered the Society on Social Implications of Technology (SSIT), a society within the Institute of Electrical and Electronics Engineers (IEEE). He was the guy with long blond hair, horn-rim glasses, and a suave and engaging manner, and he saw something humorous in just about everything. Among the more staid, business-suit-clad engineers that often showed up at SSIT meetings, Brian looked like a hippie who had wandered into a Rotary Club meeting by accident. He was the kind of person who could walk into a room and change the whole tone of conversation in five minutes from boredom to excitement, and he often did.
Naturally, not everybody always agreed with Brian's ideas. But he had the ability to see the other person's point of view instinctively, sometimes better than the other person himself. I'm sure that's what made him a good lawyer, and it is also what made him an excellent advocate of engineering ethics in a wide variety of fields, starting with computer ethics and ranging over other areas it would take a detailed study of Brian's writings to determine. As I have said elsewhere, seeing the other person's point of view is an essential first step in good engineering ethics, and Brian could do that better than just about anyone I know. In everything Brian did, there was a foundational joy in living and a desire to see other people blessed by the same joy, not harmed. And technology, since it was such a big part of life nowadays, was something Brian wanted to bless people with, not the other way around.
I think that desire is what drove him to work so energetically on behalf of the SSIT (which he served in many capacities, including President), on behalf of his law clients when he practiced law, and on behalf of his students at CCSU, many of whom he invited to his own basement lab in his house in West Hartford. When I last saw him in July of 2007, he showed me where he pursued robotics projects with his students and we talked about what he could do with robotics and remote control radio links, which he had obtained an amateur radio license to use.
Brian's actions in his chosen professions (and I count at least three: law, computer science, and engineering ethics) all sprang from a view of life that was deeply rooted in his religious and philosophical outlook. We never spoke about it much, but he was familiar with the classics and liked to quote thoughtful people of faith, from St. Augustine to G. K. Chesterton. Like Chesterton, Brian believed life was a thing to be enjoyed with all one's might. Chesterton enjoyed a glass of wine and a cigar, and Brian was partial to tobacco as well (his lung cancer was diagnosed in the spring of 2007). His legacy continues in the lives of the hundreds or thousands of students, colleagues, and fellow professionals who, I hope, will know more about engineering ethics and act on that knowledge because of something Brian did, said, or wrote. His life crossed the paths of the rest of us like a skyrocket shooting up through the trees. Perhaps Edna St. Vincent Millay had someone like Brian in mind when she wrote
My candle burns at both ends;
It will not last the night;
But ah, my foes, and oh, my friends--
It gives a lovely light!
Requiscat in pace, Brian.
This business tends to attract people with mixed backgrounds who are both conversant with the intricacies of some technical field and also interested in the human side of things. Brian was no exception. He once told me he was one of the youngest people ever to run a planetarium show at Hartford's Gengras Planetarium, when as a young teenager he was asked to fill in for the regular operator whom Brian had become friends with. But his interest in the depths of the human soul expressed itself soon thereafter when he attended seminary for a while. Deciding he wasn't quite cut out to be a priest, he switched to computer science, and then back to humanities as he took a law degree and practiced law for several years. Eventually he joined Central Connecticut State University and served with distinction in both their computer science and philosophy departments.
I met Brian shortly after he discovered the Society on Social Implications of Technology (SSIT), a society within the Institute of Electrical and Electronics Engineers (IEEE). He was the guy with long blond hair, horn-rim glasses, and a suave and engaging manner, and he saw something humorous in just about everything. Among the more staid, business-suit-clad engineers that often showed up at SSIT meetings, Brian looked like a hippie who had wandered into a Rotary Club meeting by accident. He was the kind of person who could walk into a room and change the whole tone of conversation in five minutes from boredom to excitement, and he often did.
Naturally, not everybody always agreed with Brian's ideas. But he had the ability to see the other person's point of view instinctively, sometimes better than the other person himself. I'm sure that's what made him a good lawyer, and it is also what made him an excellent advocate of engineering ethics in a wide variety of fields, starting with computer ethics and ranging over other areas it would take a detailed study of Brian's writings to determine. As I have said elsewhere, seeing the other person's point of view is an essential first step in good engineering ethics, and Brian could do that better than just about anyone I know. In everything Brian did, there was a foundational joy in living and a desire to see other people blessed by the same joy, not harmed. And technology, since it was such a big part of life nowadays, was something Brian wanted to bless people with, not the other way around.
I think that desire is what drove him to work so energetically on behalf of the SSIT (which he served in many capacities, including President), on behalf of his law clients when he practiced law, and on behalf of his students at CCSU, many of whom he invited to his own basement lab in his house in West Hartford. When I last saw him in July of 2007, he showed me where he pursued robotics projects with his students and we talked about what he could do with robotics and remote control radio links, which he had obtained an amateur radio license to use.
Brian's actions in his chosen professions (and I count at least three: law, computer science, and engineering ethics) all sprang from a view of life that was deeply rooted in his religious and philosophical outlook. We never spoke about it much, but he was familiar with the classics and liked to quote thoughtful people of faith, from St. Augustine to G. K. Chesterton. Like Chesterton, Brian believed life was a thing to be enjoyed with all one's might. Chesterton enjoyed a glass of wine and a cigar, and Brian was partial to tobacco as well (his lung cancer was diagnosed in the spring of 2007). His legacy continues in the lives of the hundreds or thousands of students, colleagues, and fellow professionals who, I hope, will know more about engineering ethics and act on that knowledge because of something Brian did, said, or wrote. His life crossed the paths of the rest of us like a skyrocket shooting up through the trees. Perhaps Edna St. Vincent Millay had someone like Brian in mind when she wrote
My candle burns at both ends;
It will not last the night;
But ah, my foes, and oh, my friends--
It gives a lovely light!
Requiscat in pace, Brian.
Saturday, May 17, 2008
China's Earthquake: What If We Had Known?
On Monday, May 12, the Sichuan region of China was devastated by one of the worst earthquakes in recent memory. At this writing, the death toll stands at over 50,000, and more bad news about the disaster arrives daily. One of the strangest news stories that has come out of region concerns rumors spread on the Internet that scientists working for the Chinese government knew the earthquake was going to happen, and suppressed the information out of fear that making their prediction public would cause panic ahead of the Olympic games.
A news source almost certainly affiliated with the Chinese government (China Radio International) issued a release Wednesday which quoted Zhang Guomin, a research fellow at China's Institute for Earthquake Science, as saying that earthquake forecasts should be based on scientific analysis and not tailored to political requirements. According to him, earthquake forecasts are not possible with our present state of knowledge. However, another researcher, Zhang Xiaodong of the China Earthquake Networks Center, seems to wish that predictions were possible, because he told the reporters, "I feel deeply regretful and sorrowful at the failure to predict the earthquake."
What if we could predict earthquakes with the same accuracy as, say, we can predict tornadoes today? At least one leading authority believes that such predictions may be possible. A NASA researcher named Friedemann Freund has published a series of papers over the years that connect measurable changes in the earth's electromagnetic fields to strong earthquakes that happen shortly after the changes. (My blogs of Feb. 20, 2007 and Apr. 13, 2006 describe more technical details.)
Without taking sides on whether this is in fact possible, let's do a little thought experiment. Suppose after X years of research and development, we assemble the expertise, equipment, and networks needed to predict major deadly earthquakes. Now no prediction system is going to be perfect, so let's say its accuracy can be quantified this way: when the system predicts an earthquake of at least a given magnitude in a given geographic area during a given time window (probably at least a week, and maybe much longer), the prediction is borne out 80% of the time. And let's say false positives and false negatives are equally likely. That is, for the 20% of predictions that come out wrong, 10% are major earthquakes that happen when none was predicted, and 10% are non-events that don't happen when an earthquake was predicted.
Given this imaginary system, what do we do with it? Do we treat the forecasts like hurricane forecasts and order mass evacuations? That's certainly one approach. Originally, Hurricane Katrina was predicted to hit the Houston area, and a graduate student I knew was pretty perturbed when he wasn't able to arrange for transportation out of the city. As it turned out, he was one of the lucky ones—nothing too bad happened to Houston, but everybody who tried to flee had to endure the grandaddy of all traffic jams on the already-clogged Houston freeways.
Hurricanes generally end up somewhere, so hurricane forecasters are given the benefit of the doubt when they miss on exact predictions of the storm's path. But what if earthquake experts made a prediction that turned out to be a complete bust—that is, everybody evacuates for the full term of the warning and exactly nothing happens? That might sully the reputation of the field indefinitely, and nobody would take them seriously forever after.
To bring the matter closer to home, what if this hypothetical system predicted The Big One for the San Francisco Bay area? If we shut down everything that goes on in Silicon Valley for a week, that would constitute a major economic disaster of its own. You don't just walk up to a huge semiconductor plant and turn off the switch, unless you want to turn it into scrap. Of course, a major earthquake might do that for you, but then you get into the question of how to deal with an evacuation order that would cost billions of dollars to a private company. Lives are more valuable than property, but property isn't negligible. And that's just one example of many problems that we would face in dealing with accurate earthquake forecasts.
The approach California has taken in the absence of reliable earthquake predictions is to mandate earthquake-resistant construction. But that costs more than ordinary construction, and requires a well-functioning regulatory system and a cooperative construction industry, neither of which are always found in other countries. Mass evacuations are simpler, and might be the best path to pursue for countries that can't afford to replace their entire infrastructure with earthquake-resistant structures.
Clearly, even if we had reliable earthquake prediction, we would face a lot of issues in deciding how to act on the knowledge it would provide. But it seems to me that knowledge is always better than ignorance, especially when it comes to earthquakes. And considering the terrible loss of life and property that major earthquakes usually cause, I wish that we spent more intellectual capital on serious efforts to predict earthquakes, and tried to evaluate the predictions in a statistically meaningful way.
Sources: The China Radio International article I quoted appeared at http://english.cri.cn/2946/2008/05/15/48@357631.htm.
A news source almost certainly affiliated with the Chinese government (China Radio International) issued a release Wednesday which quoted Zhang Guomin, a research fellow at China's Institute for Earthquake Science, as saying that earthquake forecasts should be based on scientific analysis and not tailored to political requirements. According to him, earthquake forecasts are not possible with our present state of knowledge. However, another researcher, Zhang Xiaodong of the China Earthquake Networks Center, seems to wish that predictions were possible, because he told the reporters, "I feel deeply regretful and sorrowful at the failure to predict the earthquake."
What if we could predict earthquakes with the same accuracy as, say, we can predict tornadoes today? At least one leading authority believes that such predictions may be possible. A NASA researcher named Friedemann Freund has published a series of papers over the years that connect measurable changes in the earth's electromagnetic fields to strong earthquakes that happen shortly after the changes. (My blogs of Feb. 20, 2007 and Apr. 13, 2006 describe more technical details.)
Without taking sides on whether this is in fact possible, let's do a little thought experiment. Suppose after X years of research and development, we assemble the expertise, equipment, and networks needed to predict major deadly earthquakes. Now no prediction system is going to be perfect, so let's say its accuracy can be quantified this way: when the system predicts an earthquake of at least a given magnitude in a given geographic area during a given time window (probably at least a week, and maybe much longer), the prediction is borne out 80% of the time. And let's say false positives and false negatives are equally likely. That is, for the 20% of predictions that come out wrong, 10% are major earthquakes that happen when none was predicted, and 10% are non-events that don't happen when an earthquake was predicted.
Given this imaginary system, what do we do with it? Do we treat the forecasts like hurricane forecasts and order mass evacuations? That's certainly one approach. Originally, Hurricane Katrina was predicted to hit the Houston area, and a graduate student I knew was pretty perturbed when he wasn't able to arrange for transportation out of the city. As it turned out, he was one of the lucky ones—nothing too bad happened to Houston, but everybody who tried to flee had to endure the grandaddy of all traffic jams on the already-clogged Houston freeways.
Hurricanes generally end up somewhere, so hurricane forecasters are given the benefit of the doubt when they miss on exact predictions of the storm's path. But what if earthquake experts made a prediction that turned out to be a complete bust—that is, everybody evacuates for the full term of the warning and exactly nothing happens? That might sully the reputation of the field indefinitely, and nobody would take them seriously forever after.
To bring the matter closer to home, what if this hypothetical system predicted The Big One for the San Francisco Bay area? If we shut down everything that goes on in Silicon Valley for a week, that would constitute a major economic disaster of its own. You don't just walk up to a huge semiconductor plant and turn off the switch, unless you want to turn it into scrap. Of course, a major earthquake might do that for you, but then you get into the question of how to deal with an evacuation order that would cost billions of dollars to a private company. Lives are more valuable than property, but property isn't negligible. And that's just one example of many problems that we would face in dealing with accurate earthquake forecasts.
The approach California has taken in the absence of reliable earthquake predictions is to mandate earthquake-resistant construction. But that costs more than ordinary construction, and requires a well-functioning regulatory system and a cooperative construction industry, neither of which are always found in other countries. Mass evacuations are simpler, and might be the best path to pursue for countries that can't afford to replace their entire infrastructure with earthquake-resistant structures.
Clearly, even if we had reliable earthquake prediction, we would face a lot of issues in deciding how to act on the knowledge it would provide. But it seems to me that knowledge is always better than ignorance, especially when it comes to earthquakes. And considering the terrible loss of life and property that major earthquakes usually cause, I wish that we spent more intellectual capital on serious efforts to predict earthquakes, and tried to evaluate the predictions in a statistically meaningful way.
Sources: The China Radio International article I quoted appeared at http://english.cri.cn/2946/2008/05/15/48@357631.htm.
Saturday, May 10, 2008
Ethics of the Smart Car
The relationship between drivers and their cars has always been a complex one, fraught with emotional and moral overtones. Maybe that was why some television writers with more enthusiasm than judgment came up with the concept of "My Mother the Car." I'm old enough to remember watching that show, which aired on U. S. television back in 1965. The basic idea was that this guy buys an antique car, only to discover that somehow his deceased mother's spirit has taken up residence in it. The radio dial flashed whenever she spoke to him, I guess so TV viewers could tell that it really was the car and not some hallucinogen-inspired inner voice. The show lasted only one season and is remembered, if at all, for being one of the worst TV series of all time. But if Prof. Clifford Nass of Stanford University has his way, we all may be talking with our cars in the future—and the cars may talk back in tones to match our emotions.
A recent Wired article profiled Prof. Nass's research on the future of the human-automobile interface, and how smart cars may be used. Smart in what sense? Well, with current GPS (global positioning system) technology and computer power, coupled with broadband wireless networks that will be ubiquitous soon, you can imagine driving down the street and saying to your car, "Hey, I'd like a pizza. Any good places within a couple of blocks?" Advertisers and automakers would like your car to reply, "Well, there's Gino's in the next block and Papa's one block over—they're having a lunch special today. What shall it be?"
Of course, the same smarts that lets your car give you dining advice will also empower it to remember how you drive. Auto insurance companies currently give discounts to good drivers and raise rates on poor ones, but the quality of your driving is determined mainly by very coarse measures: the number of accidents and traffic violations. Suppose every week your insurer could download and process (by software, of course) hundreds of details about how you drive: how fast you pulled out after a light changed, whether you were speeding and by how much, and whether you ran red lights without getting caught. Most of the technology's there, it's just a matter of developing it.
Some people would think this amounts to turning one's private car into a spy. The matter gets even more complex if we move to cars which partially or totally take over many of the functions of driving. (See my column "The Human Side of Automated Driving" Dec. 10, 2007). Clearly, if you take your hands completely away from the controls and let the car do everything, your responsibility for accidents that ensue is limited, if not absent entirely. But many plans for computer-assisted driving don't go that far. Nass imagines a heavy-footed driver negotiating with his car for permission to step on the gas after a stop light changes. "Aw, c'mon, just this once?" "No, you're wasting gas, and at five dollars a gallon!" Nass says that changing the car's tone of voice to match the driver's mood may help the situation, but I'm not so sure.
Right after it was economically feasible to put computer-generated voices in cars, some time in the early 1990s, a few manufacturers experimented with it. The idea proved to be almost universally unpopular, as the mechanical female tone reminded everybody of their worst nagging nightmares of school librarians and mothers (there it is again), and the feature disappeared in a model year or so.
Where is engineering ethics in all this? The first responsibility of engineers who are working on these things is to make sure they don't make driving more dangerous. Of course, that doesn't mean things can't ever go wrong occasionally, but tests will have to show a general improvement in safety before new features can be adopted. As for insurance companies and driving information, there is a public-policy aspect which has not been debated yet. It's the same kind of question that arises when health insurers want to use a person's genetic information to restrict health coverage, except in that case you can't help what genes you were born with, but you presumably have some control over how you drive. But should a taxi driver in New York pay higher rates than the legendary little old lady from Pasadena who only drives to church on Sundays? These are questions that involve technology as well as issues of fairness, economics, and what insurers like to call "moral hazard"—that is, the idea that you should not be exempt from all the consequences of your own voluntary bad behavior.
For my part, I'll be content to drive my old, dumb cars (dumb in two senses) until the wheels fall off. And maybe by then I can buy a car named James and commute by saying, "Home, James," and just enjoy the scenery while the car worries about the congestion on IH-35.
A Note To Readers
For the next two to four weeks I will be pursuing some research in a rather remote location where Internet access is not as reliable as it could be. So I apologize in advance for any delays in my weekly postings, which I will try to keep current as much as possible. For more information about the subject of my research, see www.nightorbs.net.
Sources: The Wired article appeared on May 9, 2008 at http://blog.wired.com/cars/2008/05/a-data-mining-c.html. And Wikipedia has an article that will tell you more than you will ever need to know about the show "My Mother the Car."
A recent Wired article profiled Prof. Nass's research on the future of the human-automobile interface, and how smart cars may be used. Smart in what sense? Well, with current GPS (global positioning system) technology and computer power, coupled with broadband wireless networks that will be ubiquitous soon, you can imagine driving down the street and saying to your car, "Hey, I'd like a pizza. Any good places within a couple of blocks?" Advertisers and automakers would like your car to reply, "Well, there's Gino's in the next block and Papa's one block over—they're having a lunch special today. What shall it be?"
Of course, the same smarts that lets your car give you dining advice will also empower it to remember how you drive. Auto insurance companies currently give discounts to good drivers and raise rates on poor ones, but the quality of your driving is determined mainly by very coarse measures: the number of accidents and traffic violations. Suppose every week your insurer could download and process (by software, of course) hundreds of details about how you drive: how fast you pulled out after a light changed, whether you were speeding and by how much, and whether you ran red lights without getting caught. Most of the technology's there, it's just a matter of developing it.
Some people would think this amounts to turning one's private car into a spy. The matter gets even more complex if we move to cars which partially or totally take over many of the functions of driving. (See my column "The Human Side of Automated Driving" Dec. 10, 2007). Clearly, if you take your hands completely away from the controls and let the car do everything, your responsibility for accidents that ensue is limited, if not absent entirely. But many plans for computer-assisted driving don't go that far. Nass imagines a heavy-footed driver negotiating with his car for permission to step on the gas after a stop light changes. "Aw, c'mon, just this once?" "No, you're wasting gas, and at five dollars a gallon!" Nass says that changing the car's tone of voice to match the driver's mood may help the situation, but I'm not so sure.
Right after it was economically feasible to put computer-generated voices in cars, some time in the early 1990s, a few manufacturers experimented with it. The idea proved to be almost universally unpopular, as the mechanical female tone reminded everybody of their worst nagging nightmares of school librarians and mothers (there it is again), and the feature disappeared in a model year or so.
Where is engineering ethics in all this? The first responsibility of engineers who are working on these things is to make sure they don't make driving more dangerous. Of course, that doesn't mean things can't ever go wrong occasionally, but tests will have to show a general improvement in safety before new features can be adopted. As for insurance companies and driving information, there is a public-policy aspect which has not been debated yet. It's the same kind of question that arises when health insurers want to use a person's genetic information to restrict health coverage, except in that case you can't help what genes you were born with, but you presumably have some control over how you drive. But should a taxi driver in New York pay higher rates than the legendary little old lady from Pasadena who only drives to church on Sundays? These are questions that involve technology as well as issues of fairness, economics, and what insurers like to call "moral hazard"—that is, the idea that you should not be exempt from all the consequences of your own voluntary bad behavior.
For my part, I'll be content to drive my old, dumb cars (dumb in two senses) until the wheels fall off. And maybe by then I can buy a car named James and commute by saying, "Home, James," and just enjoy the scenery while the car worries about the congestion on IH-35.
A Note To Readers
For the next two to four weeks I will be pursuing some research in a rather remote location where Internet access is not as reliable as it could be. So I apologize in advance for any delays in my weekly postings, which I will try to keep current as much as possible. For more information about the subject of my research, see www.nightorbs.net.
Sources: The Wired article appeared on May 9, 2008 at http://blog.wired.com/cars/2008/05/a-data-mining-c.html. And Wikipedia has an article that will tell you more than you will ever need to know about the show "My Mother the Car."
Monday, May 05, 2008
I Got the Botts About Bots
My father, God rest his soul, had enough South Texas German in him to be subject to occasional fits of Teutonic depression. He had enough self-awareness to know what was going on when these moods hit him. When we asked him what was bothering him, he'd generally say, "Aw, I've got the botts." (I never saw him write the word down, but for some reason I think it's spelled with two t's.) He passed on many years before the Internet was more than a gleam in a few researchers' eyes, but if he were alive now, he might well have the botts about bots.
A bot is a piece of malevolent software (malware) that infects your computer with the purpose of controlling it to do things that the bot tells it to do. These things are generally not nice. In the case of one of the worst bots, Storm Worm, some observers say that over a million computers took orders from some people who apparently went on the black market to offer denial-of-service attacks to the highest bidder. If a criminal takes up the offer, the victim's website is likely to be inundated with many millions of emails or other automated requests for service, whereupon the target website immediately gets overwhelmed and becomes inaccessible to legitimate users. Creators of botnets have progressed in the last few years from random vandalism to coordinated criminal activity, which is why computer security firms and software providers from Microsoft on down have lately spent so much time and effort combating the problem.
Until recently, people such as myself who use Macintosh computers could ignore bots, since up to 2004 or so no one had bothered to write a bot for Macs. Since only a relatively small percentage of all computers online at a given time are Macs, a malware writer who wants access to the largest number of computers in the shortest time is probably not going to bother writing two different bot programs, one for Macs and one for PCs. (Most legitimate software companies don't either, but that's another story.) But this supposed invulnerability has evidently come to an end. The other day I received a message from the IT division of a university where I do research. It informed me that a Mac on a network node in the lab I was working in was being remotely controlled by a bot. I was alarmed until I called the people and checked the Ethernet ID address, or whatever it's called—an identifying number unique to my computer. The number didn't match mine, so my computer must not have been the one that was zombified. Still, it means there could be a problem in the future.
It turns out that bots tend to use something called IRC, which stands for Internet Relay Chat. This is the old original protocol that enabled the first internet-based chats, before companies started selling proprietary versions. I am not a computer scientist and I don't know why this particular protocol is so useful to botnet masterminds, but it is.
Wouldn't it be nice if we could rewind to the day when the first wide-eyed innocent programmer came up with the neat idea of the IRC in the first place? "Hey, kids, let's make it so we can chat over the Internet in real time." Sounds great. But apparently, there is something fundamentally flawed about that IRC protocol that makes it able to take over people's computers.
I'm sure that was the last thing in the programmer's mind, to put in a built-in flaw that would later be exploited by criminal elements to the harm of thousands of victims, and to the possible legal compromise of millions of people who unknowingly participate in these crimes simply because their computers are hosting bots and follow the orders of their evil digital masters. But hey—with opportunity comes responsibility.
There is an idea in the engineering ethics world called the precautionary principle. Wikipedia defines it this way: "If there is a risk that an action could cause harm, and there is a lack of scientific consensus on the matter, the burden of proof is on those who would support taking the action." You hear more about it in European ethics discussions than in the U. S. Taking it seriously would severely hamper development of new technologies of all kinds. I wonder, though, if the people who developed the early Internet protocols had taken a more cynical view of human nature, and tried to think of all the evil things ill-willed programmers could do with the neat tools they were putting out there, if we might not have some of the problems we struggle with today.
If, for example, the developers of the IRC had taken a prototype version to some creative young bucks who spent their days trying to devise malevolent uses for new software, they might have discovered the extreme usefulness of IRC in botnets. And who knows?—they might have fixed it in a way that stayed permanently embedded in the Internet as it grew faster than almost anyone expected.
It's obviously too late to close the barn door on that particular horse. Now that Macs can harbor bots, I'll just have to be careful and try to make sure I follow good computer hygiene, for whatever good that will do. But people are writing new software all the time, and some of it is destined to be as influential and ubiquitous as the infamous IRC protocol is now. Surely we have learned a lesson about the depths of depravity to which some programmers will stoop. I just hope that people who write software these days take some thought as to how what they develop could be misused in the future, and even twist their minds around to be creative about it—and then fix it so it can't be used that way.
Sources: Slate has a good introduction to the subject of bots at http://www.slate.com/id/2190275/. A recent overview of the subject from a technical perspective can be found at http://8e6labs.com/2007/11/02/overview-of-the-threats-posed-by-bots/.
A bot is a piece of malevolent software (malware) that infects your computer with the purpose of controlling it to do things that the bot tells it to do. These things are generally not nice. In the case of one of the worst bots, Storm Worm, some observers say that over a million computers took orders from some people who apparently went on the black market to offer denial-of-service attacks to the highest bidder. If a criminal takes up the offer, the victim's website is likely to be inundated with many millions of emails or other automated requests for service, whereupon the target website immediately gets overwhelmed and becomes inaccessible to legitimate users. Creators of botnets have progressed in the last few years from random vandalism to coordinated criminal activity, which is why computer security firms and software providers from Microsoft on down have lately spent so much time and effort combating the problem.
Until recently, people such as myself who use Macintosh computers could ignore bots, since up to 2004 or so no one had bothered to write a bot for Macs. Since only a relatively small percentage of all computers online at a given time are Macs, a malware writer who wants access to the largest number of computers in the shortest time is probably not going to bother writing two different bot programs, one for Macs and one for PCs. (Most legitimate software companies don't either, but that's another story.) But this supposed invulnerability has evidently come to an end. The other day I received a message from the IT division of a university where I do research. It informed me that a Mac on a network node in the lab I was working in was being remotely controlled by a bot. I was alarmed until I called the people and checked the Ethernet ID address, or whatever it's called—an identifying number unique to my computer. The number didn't match mine, so my computer must not have been the one that was zombified. Still, it means there could be a problem in the future.
It turns out that bots tend to use something called IRC, which stands for Internet Relay Chat. This is the old original protocol that enabled the first internet-based chats, before companies started selling proprietary versions. I am not a computer scientist and I don't know why this particular protocol is so useful to botnet masterminds, but it is.
Wouldn't it be nice if we could rewind to the day when the first wide-eyed innocent programmer came up with the neat idea of the IRC in the first place? "Hey, kids, let's make it so we can chat over the Internet in real time." Sounds great. But apparently, there is something fundamentally flawed about that IRC protocol that makes it able to take over people's computers.
I'm sure that was the last thing in the programmer's mind, to put in a built-in flaw that would later be exploited by criminal elements to the harm of thousands of victims, and to the possible legal compromise of millions of people who unknowingly participate in these crimes simply because their computers are hosting bots and follow the orders of their evil digital masters. But hey—with opportunity comes responsibility.
There is an idea in the engineering ethics world called the precautionary principle. Wikipedia defines it this way: "If there is a risk that an action could cause harm, and there is a lack of scientific consensus on the matter, the burden of proof is on those who would support taking the action." You hear more about it in European ethics discussions than in the U. S. Taking it seriously would severely hamper development of new technologies of all kinds. I wonder, though, if the people who developed the early Internet protocols had taken a more cynical view of human nature, and tried to think of all the evil things ill-willed programmers could do with the neat tools they were putting out there, if we might not have some of the problems we struggle with today.
If, for example, the developers of the IRC had taken a prototype version to some creative young bucks who spent their days trying to devise malevolent uses for new software, they might have discovered the extreme usefulness of IRC in botnets. And who knows?—they might have fixed it in a way that stayed permanently embedded in the Internet as it grew faster than almost anyone expected.
It's obviously too late to close the barn door on that particular horse. Now that Macs can harbor bots, I'll just have to be careful and try to make sure I follow good computer hygiene, for whatever good that will do. But people are writing new software all the time, and some of it is destined to be as influential and ubiquitous as the infamous IRC protocol is now. Surely we have learned a lesson about the depths of depravity to which some programmers will stoop. I just hope that people who write software these days take some thought as to how what they develop could be misused in the future, and even twist their minds around to be creative about it—and then fix it so it can't be used that way.
Sources: Slate has a good introduction to the subject of bots at http://www.slate.com/id/2190275/. A recent overview of the subject from a technical perspective can be found at http://8e6labs.com/2007/11/02/overview-of-the-threats-posed-by-bots/.
Subscribe to:
Posts (Atom)
