March 03, 2002
Afghanistan and the lessons of Y2K

After reading Gary Farber's thorough takedown of The Guardian's resident idiot, I got to thinking how a presumably intelligent person could be so utterly off-base. It's not so much that Mary Riddell and those like her are guaranteed to be wrong about their conclusions - the US-Iraq situation could very well turn into the kind of disaster that she seems to be rooting for - it's the way they get to those conclusions.

Alarmists in general seem to be extremely adept at tailoring logic to fit their conclusions. They ignore or shout down opposing views, and when the facts come in against them, they shunt them aside and hunt for any dark cloud they can point to as proof of their inevitable correctness. We saw this behavior quite a bit in the weeks leading up to the successful invasion of Afghanistan, and I daresay we'll see more screeds like Mary's in the future if and when the US decides to act against Iraq.

This line of thinking led me to another event for which doom was frequently, loudly, and incorrectly predicted. I'm speaking of the once-feared Y2K problem, in which outdated computer systems would melt down on January 1, 2000, causing worldwide chaos and destruction. I think it's instructive to look at what some of the Y2K doomsayers were saying then, and what they said afterwards when it became clear that they weren't just wrong but not even close.

The Y2K doomsayers made claims like the following:

  • The scope of the problem and its potential for catastophe are bigger than the mainstream media and its anointed experts would have you think.

  • There are zillions of lines of code which may have the Y2K bug in them. Fixing this will take jillions of person hours and cost gazillions of dollars.

  • Government and business have a vested interest in downplaying the problem. Those of us who dare to question their claims and tell the truth about what's going on will be harassed and silenced.

Sound familiar? Even after January 1, 2000 came without civilization ending, most of them refused to back off their predictions. They never said that all problems would manifest themselves right away, after all. Wait and see, there will be problems later.

Ed Yourdon was one of the big Y2K doomsayers. If you poke around his website, you can find some hindsight from him on this issue. On this index page, he admits he was wrong, in a roundabout way, anyway:

I was wrong about Y2K. Not about the magnitude and pervasive nature of the problem, and not about the likely consequences if millions of computer systems and embedded chips around the world had not been repaired or replaced. But I was wrong about the likelihood that enough of the repair/remediation would be finished in order to prevent serious disruptions. Indeed, it has gradually become apparent during the first few weeks and months of 2000 that Y2K has caused a number of moderate-to-serious problems in various parts of the world -- but it has not turned out to be the crisis that some of us had anticipated.

Italics his. This last sentence is a masterpiece of understatement, since the "crisis" Ed Yourdon and others anticipated was in fact the fall of the United States government amidst worldwide panic and economic catastrophe. In this essay Yourdon wrote responding to President Clinton's speech at the National Academy of Sciences in 1998, Yourdon gives a fairly clear idea of what he had anticipated:

If the lights are out, if the phones are dead, if the banks are closed, if the airplanes are not flying, and if the hospitals are not accepting patients, I can assure you that the average citizen will not be spending much time wondering whether he still has more computer power than MIT.
[P]lease tell the American people that several of the oil companies are terrified of the problem they face here, because many of the chips are down on the bottom of the ocean floor -- and thus incredibly difficult to find, fix, or replace. If the oil rigs shut down, our supply of oil is threatened; if we don't get enough oil, you're going to be faced with the alternative of rationing gasoline or shutting down the oil-fired electric utilities.

Yourdon goes on in the introductory essay to downplay his erroneous vision and to point to any and every glitch that did occur as vindication. It's in this essay where Yourdon shows that he truly did not and does not understand why he was wrong about Y2K. This was a postmortem look at accusations that he had been "shouting fire in a crowded theater" with his dire predictions:

In any case, we didn't know how Y2K would work out in 1998. Perhaps there are a few who can honestly say that they absolutely, positively knew that Y2K would be a non-event, but the vast majority of us had to admit, if only in private, that we wouldn't really be absolutely sure. And, in the context of this postmortem, I'm not sure if it would work out the same way -- i.e., a Y2K non-event -- if we had it all to do over again. Indeed, my mental image of the whole situation is that God flipped a coin to determine whether to decide whether to make Y2K a disaster or a non-event. This time, the coin came up heads, and God shrugged and let the world off with only a few glitches. But if we rolled the calendar back a couple years and went through the whole process again, that same coin-toss might come up tails -- in which case, God might have decided to let a few electric grids shut down, a few banks collapse, and maybe even a few airplanes fall from the sky. I know that this is an area of intense debate and controversy, and that many people are deeply convinced that there was no possible way -- no way, no-how -- for the incipient Y2K bugs to have caused a serious disaster. But there are others of us, myself included, who feel that we were incredibly lucky, and that the outcome could easily have been much, much different.

I'm now going to get to the point of this essay, in which I tell you why the Y2K bug was in fact a nonproblem, why Ed Yourdon fails miserably at understanding this, and what it all has to do with all those equally wrong doomsayers about our mission in Afghanistan.

The people who forecast doom in Y2K were actually right about a few things. They were right in that the problem was widespread, and that fixing it would be a massive and expensive undertaking whose success would be in doubt. They were wrong in assuming that the problem had to be fixed, or more to the point that it had to be fixed on the problem's terms. From my experience inside the IT department of a large multinational company, the problem was largely fixed by getting rid of the problematic pieces.

For example, rather than upgrade our mainframe systems that ran a non-Y2K-compliant version of VM, we migrated all the applications off it and onto client-server systems, then retired VM a good six months before 2000. This also obviated the need to edit those oft-cited millions of lines of COBOL code, since most if not all of that code lived on VM. Rather than upgrade BIOS chips on older PCs, we threw them out and installed new PCs. What's more, a lot of this work was done well before "Y2K" became part of the national consciousness. It was done as part of our normal cycle of upgrading old systems and installing new technologies. By the time we got around to creating a Y2K team (in 1997 or 98, I forget), most of the problem had already been solved.

Given that, it's pretty clear that while no one could truly say what Y2K would be like, the vision of God tossing a coin is ridiculous. Ed and his ilk never clued in to the fact that they were looking at a vastly different problem than the rest of us were.

Now think about the people who predicted doom when the US was preparing to invade Afghanistan and take out the Taliban. They threw around words and phrases like "quagmire", "VietNam", "failed Soviet invasion", and so forth. They talked about how the Taliban troops were master guerrila warriors who could hide in the mountains and inflict massive casualties on ground troops. They dismissed air attacks as being ineffective and vulnerable to ground-based missiles. They warned about how the locals would be against us because of the number of civilians we'd kill with our bombs. They scared us with visions of the "Arab Street" rising up to take arms around the world.

In other words, they saw the strengths of the enemy and assumed we would have to fight them on their terms. They didn't give the people whose job it is to solve these problems any credit for thinking of ways to use our strengths and to fight these battles on our terms. They drew on our failures of the past without realizing that we did in fact learn from them. That their predictions were invariably wrong should surprise no one.

Now I'm not saying that any future invasion of Iraq will be as quick or successful as the invasion of Afghanistan was. Nor am I saying that we've licked the whole problem in Afghanistan - we're still fighting, and we will be for the forseeable future. Many people, such as Steven den Beste and Sgt. Stryker have written intelligently and in depth about how things could go in Iraq. I just want to point out that the people who raise the loudest alarms are not necessarily the best sources for how to resolve the problems we face, usually because they're not talking about the problems we are actually facing.

(By the way, Ed is working on a new book called Byte Wars, in which he discusses security, risk management, and the "strategic implications of September 11". You can read the introductory chapter here. Ed also has a blog, which he supposedly updates "most every day", though the latest entry is February 10.)

There's one final thing to consider about doomsayers in general, which is that some of them are not making an honest attempt to predict the future but are instead merely rooting for the outcome they wanted. This was especially true with Y2K doomsaying, as many people interpreted the beginning of a new millennium as the beginning of the end times described in the Book of Revelations. Some people saw the possible breakdown of technology, commerce, and government as being necessary to restore God's rule on earth. Surely some people who predicted dire consequences for the United States if it invaded Afghanistan did so because they hoped to see the imperialist oppressor humbled. For these people, what was true about Afghanistan will be true about Iraq and wherever else the war on terror may lead.

So what happens to these prognosticators whose forecasts turn out to be so wrong? Well, most of them seem to just keep going. No one ever remember these things, right? And surely, they figure, one of these days God's coin toss is bound to come up favorably for them. So they go on with business as usual. As this Wired article notes about Gary North, someone Yourdon frequently cited and whose web site is very different now than it was three years ago:

What will the Internet's best-known doomsayer do if Y2K results in just minor disruptions? "A few years later he'll reappear with another apocalyptic scenario," Berlet predicts.

Surely the same is also true about the Mary Riddells of the world. We would be well advised to keep that in mind.

Posted by Charles Kuffner on March 03, 2002 to Skepticism