Peter Bowditch's Web Site

Advertising policy

Y2K - Is there still time to panic?

This is an adaptation of a presentation given by Peter Bowditch and Brian Watts to the annual conference of the Australian Skeptics on November 5, 1999.

Extraordinary Popular Delusions & the Madness of CrowdsIn a fabulous book called published in 1841, Charles Mackay told us about panics and hysterias of the past. Updating the book to today would just require the addition of a chapter on the Internet stock craze, an update to the witchcraft chapter to include mention of recovered memory syndrome, satanic ritual abuse and alien abductions, and a new chapter about millennial madness.

Talking of the first crusade, Mackay said:

"A strange idea had taken possession of the popular mind at the close of the tenth and commencement of the eleventh century. It was universally believed that the end of the world was at hand; that the thousand years of the Apocalypse were near completion, and that Jesus Christ would descend upon Jerusalem to judge mankind. All Christendom was in commotion. A panic terror seized upon the weak, the credulous, and the guilty, who in those days formed more than nineteen-twentieths of the population."

Half a millennium later, not much had changed.

"A singular instance of the faith in predictions occurred in London in the year 1524. The city swarmed at that time with fortune-tellers and astrologers, who were consulted daily by people of every class in society on the streets of futurity. As early as the month of June 1523, several of them concurred in predicting that, on the 1st day of February 1524, the waters of the Thames would swell to such a height as to overflow the whole city of London, and wash away ten thousand houses. The prophecy met implicit belief.

"It was reiterated with the utmost confidence month after month, until so much alarm was excited that many families packed up their goods, and removed into Kent and Essex. As the time drew nigh, the number of these emigrants increased. In January, droves of workmen might be seen, followed by their wives and children, trudging on foot to the villages within fifteen or twenty miles, to await the catastrophe. ... Bolton, the prior of St. Bartholomew's was so alarmed, that he erected, at a very great expense, a sort of fortress at Harrow-on-the-Hill, which he stocked with provisions for two months. Many wealthy citizens prayed to share his retreat; but the prior, with a prudent forethought, admitted only his personal friends, and those who brought stores of eatables for the blockade."

There are two sets of possible Y2K problems. The first of these is to do with the times themselves, as the predictions in Revelation and the works of Nostradamus and other seers come into effect. It is The Millennium with capital letters. It is the end of time and the start of new time. Societies, communities and countries will be torn apart and rebuilt. Some will be saved - many will not.

The second set of problems is technological. Computers and automated systems will fail. The devices we have come to depend on will no longer be Waiting for Y2Kable to help us, and may even turn on us in a hostile or recalcitrant manner. Our thinking that we can invent intelligent machines will be exposed for the hubris it is, and again some will be saved but many will perish.

Across these runs another dimension of likelihood. This dimension is dichotomous (I would say "binary" but that's what got us into the computer problems). You either believe or you don't, like young-earth creationism or the value of immunisation. When you combine these two dimensions you get a four-way Panic and Pessimism Matrix. There is a fifth alternative, but this will be mentioned later.

The first, and most pessimistic, of the cells in the matrix is that the four horsemen will ride into a landscape already devastated by the loss or failure of computer systems. The computer problems are just part of the overall plan. We have created the tools of our own destruction and, like the wings of Icarus, they will fail when we come face-to-face with reality and our limitations are exposed.

The next cell is where the computers fail but the religious and social pessimism fails to come true. We are left starving in the dark, broke because the banks can't operate, communications gone as the Internet and telephone systems crash. There will, of course, be looting, violence and murder, but it will be generally good-natured as people just provide for their families and readjust to a subsistence society based on village life.

The third cell is where the apocalypse comes but the computers keep on working. You may wonder why this is a less pessimistic scenario than losing the computers, but think of the advantages. Society , law and order might collapse, but those of us with Internet connections will be safe in our houses as we order food, toilet paper, ammunition, pornography and bomb-making instructions. On a personal note, there will also be less demand for the lynching of computer experts.

The fourth scenario is that nothing will happen at all. This is the least interesting, which is why it has not been fashionable (or profitable) to talk about it.

When I set out to write this I was going to concentrate on the hysteria which has come from the computer industry about Y2K. I had originally thought that the hysterical end of the spectrum was a lot like the superstitious or religious millennialist crowd. There was a belief in illogical and contradictory nonsense and a readiness to believe anything. Examples of gullibility were everywhere, and danger was seen where there was no rational possibility of a problem.

Some examples:

  • I remember starting the pacemaker rumour about two years ago when I told some earnest person that my only Y2K worry was that the software in my pacemaker would decide that too much time had passed since the battery was changed and then turn the thing off. I thought the notion was just too absurd for even the silliest Y2K fanatic to believe, but the person I told it to just nodded. A few weeks ago this came up in a discussion between computer professionals and someone told me that perhaps we should ask an electrical engineer "who might know what he is talking about". I gave him the web addresses of six pacemaker manufacturers who were sick of telling people that pacemakers do not care at all about the year.
  • The application form on the web to join the Australian Computer Society asks for date of birth. Because Americans kept putting the date backwards, I added a note to the online form asking for dates in dd/mm/yy form. Someone complained that this was not Y2K compliant and the ACS should set an example. I refused to change it on principle, because I felt that the ACS staff would have little difficulty working out which century members were born in.
  • One expert reported that a network hub was not compliant and needed to be replaced with one that was. As a hub contains no logic and is just fancy sort of double adapter, one would have to doubt this expert's honesty as well as sanity.
  •  After adverse publicity, Microsoft had to issue a modification to Windows 98 to address a condition which was so improbable that nobody with any grasp of numbers or statistics could imagine it happening. I will talk about this example in more detail later.

There was also a tendency to treat real problems which have trivial solutions or work-arounds as though the problems were huge:

  • Credit cards with two-digit expiry years - as credit cards are never more than five years old the arithmetic to check dates is pretty simple
  • Cases where what matters is the pattern of days, not the actual year - traffic light systems, air conditioners, video recorders where using 1972 as the year gives the same pattern of weekend days.

As I was thinking about this, another thought kept coming to me. Maybe the way computer professionals were behaving was not like the survivalists or the religious millennialists. I started to think about the term "computer science" and then it struck me. The analogy here isn't with end-times fundamentalists, it's with creationists. I was looking at a classic pseudoscience. "Computer science" has the same relationship to real science as "creation science" has.

Every university in the country has a computer science department. Maybe they should be renamed "computer studies" to indicate the danger of influences like postmodernism and the presence of gurus, ogres and absolute truths. If computer science really is pseudoscience, we may have more worries about Y2K than we thought.

Let's look at some of the pseudoscientific (or anti-scientific) aspects of computer science. The definition I am using for real science is "an open system based on sceptical enquiry, with its ultimate appeal being to evidence".

There is a belief among computer people that bug-free software is possible, and this belief causes supposed experts to gloat that large software companies must be incompetent because they keep issuing updates. Unfortunately, writing software is extremely complex and there is no theoretical way to prove that any program (other than the most trivial) does what it is supposed to do in all circumstances. The method employed is to test a program to see if it works without considering why it might work (or not work), so Y2K testing consists of putting in some dates and seeing what comes out. Often, the criteria for accepting the correctness of a program is that it runs to the end without crashing. This is the source of the aphorism "All's well that ends". This is a lot like the alternative medicine quacks who know what works so clinical trials are a waste of time and money. No theory, just results. (I have had some correspondence recently with Dr Jacques Benveniste, he of transmission of water memory by email fame. I said that my knowledge of science made me sceptical of his theories, and he replied that he had no theory, just replicated results.)

Another pseudoscience indicator is the total inability to grasp probability and statistics. The creationists are always quoting the probability of a wind through the Boeing warehouse blowing a 747 out the other end. The Windows 98 problem I mentioned earlier is something like this, where an almost inevitable outcome is claimed for a series of temporally related but unpredictable actions following a precise (but highly improbable) set of initial conditions. On the other hand, with in excess of 300 million personal computers in the world, it doesn't take a very high natural problem rate to produce very large numbers of problem cases. One problem per year on average (which is very optimistic) gives about a million problems per day, so I predict about 30 million problems in January 2000 which will be blamed on Y2K even though they have nothing to do with it. If there is only a million-to-one chance of something happening to you, it could be happening 18 times a day in Australia. (You may think that I am contradicting myself here saying that highly improbable things both do and don't happen. Stats 101. The Windows 98 case is the a priori prediction of a specific event; the natural problem rate is random variation exhibited in a huge population.)

More evidence of pseudoscience is the ignorance of the history, fundamentals and literature of the discipline. Magnetic healers, perpetual motionists and mind readers all act as if what they are doing hasn't been tried before and they all reject or ignore any other opinion.

  • The editor of one of Australia's most prestigious computer publications once ridiculed Byte magazine for giving a lifetime service award to Commander Grace Hopper of the US Navy, saying that he could see no way anyone from the Navy could have anything worthwhile to do with computers. Grace Hopper was one of the people responsible for the invention of high-level programming languages (and she popularised the use of the term "bug"). Not knowing about her was like a physicist not knowing about Newton. The editor got a promotion to an even more prestigious paper.
  • I did a Unix course at a well-known university once and the lecturer made some absurd statement about a capability which he said was first implemented in the "C" programming language. I pointed out that it was in COBOL, where it was mandatory, and he told me that I was wrong because he didn't know that, and COBOL wasn't used any more, and anything that old could not have any relevance to today, and
  • I don't know how many times I have heard that the Y2K problem is caused by saving space when writing in COBOL, but COBOL has had the capability since 1963 of storing a four-digit year in two character positions. The technique may not have been used, but that does not excuse experts from knowing about it. I wonder what else they don't know.
  • There have been techniques around since the 1970s for structuring programs and managing development so that it is possible to know where particular pieces of data are used and you can find, for example, dates and change how they are stored or acted upon.
  • One of the most-heard complaints from members of the Australian Computer Society is that the Society wastes money publishing the quarterly Australian Computer Journal, which contains academic articles which "have no relevance to the real world of computing".
  • There are no barriers to entry into the computer business and age and experience are devalued. To become a computer expert you simply have to declare yourself to be one. To test this, go to any computer shop and ask the youthful expert behind the counter to define the term "megabyte" and see if he (invariably a "he") can answer without using the word "byte".

I mentioned postmodernism, and you may wonder what that could possibly have to do with computers. About a year ago I became involved in an argument about people writing Java programs which could crash someone's web browser. I was told by someone (from a university computer science department) that Java programs were just streams of bits like any other streams of bits, that bad results were caused by misinterpretation of the bits, that a bad program could not be written, and that it was Microsoft's fault if its browser could not execute the programs. When I suggested that not all streams of bits were equivalent, I was ridiculed for not understanding the problem. When I said that I had two CDs in front of me, one with Windows NT4 on it and the other with Beethoven's 9th Symphony and could someone please explain how my CD-ROM drive knew the difference, I was ignored. Another person told me that viruses were caused by Windows, apparently because Windows runs programs written for Windows! (As an aside, Java is the newest and most fashionable programming language. It was not until two years into its development that someone noticed that its inbuilt date functions allowed two-digit years to be stored. Some people are slow to learn.)

You would think that all this confusion and ignorance would make me pessimistic about computers and Y2K. The computer people may be telling us that everything is in hand, but this is like finding out that chiropractors have taken over a major teaching hospital. The people we rely on don't know what the real problem is and don't know how to fix it. But they believe they do. Despite all this I am optimistic, simply because most of the disaster scenarios I have seen have been highly unlikely anyway and a large amount of money has been spent, much of it to prove that the fears were baseless.

There is actually an upside to all this. Even if the whole thing turns out to be a complete furphy, businesses have been forced to look at systems and procedures and to implement risk analysis and disaster plans. Hopefully, computer people have learned something about the value of planning and documentation, although, as I have mentioned, we knew about these 20 years ago. I am not too worried that the world is going to end, but I just might buy a couple of extra rolls of toilet paper and some beans.

So, to answer the question "Is there still time to panic?", the answer is, as expected, "yes and no". There is always time to panic, even when everything possible has been done, yet there is no benefit to panic when inevitability is showing itself above the horizon.

I mentioned a fifth scenario. All of the above may become moot when the planets and the sun all line up in early May 2000 and the world is torn apart by tidal forces. Now, there's something to really panic about.

To return to London in 1524:

"At last the morn, big with the fate of London, appeared in the east. The wondering crowds were astir at an early hour to watch the rising of the waters. The inundation, it was predicted, would be gradual, not sudden; so that they expected to have plenty of time to escape as soon as they saw the bosom of old Thames heave beyond the usual mark. But the majority were too much alarmed to trust to this, and thought themselves safer ten or twenty miles off. The Thames, unmindful of the foolish crowds upon its banks, flowed on quietly as of yore. The tide ebbed at its usual hour, flowed to its usual height, and then ebbed again, just as if twenty astrologers had not pledged their words to the contrary. Blank were their faces as evening approached, and as blank grew the faces of the citizens to think that they had made such fools of themselves. At last night set in, and the obstinate river would not lift its waters to sweep away even one house out of ten thousand. Still, however, the people were afraid to go to sleep. Many hundreds remained up till dawn of the next day, lest the deluge should come upon them like a thief in the night.

"On the morrow, it was seriously discussed whether it would not be advisable to duck the false prophets in the river. Luckily for them, they thought of an expedient which allayed the popular fury. They asserted that, by an error (a very slight one,) of a little figure, they had fixed the date of this awful inundation a whole century too early. The stars were right after all, and they, erring mortals, were wrong. The present generation of cockneys was safe, and London would be washed away, not in 1524, but in 1624. At this announcement, Bolton the prior dismantled his fortress, and the weary emigrants came back."

A remarkable coincidence happened with the thoughts in this article and you can read about it here.

I was asked to give a follow-up talk after Armageddon had or had not happened. You can read it here.

You can buy a

Copyright 1998- Peter Bowditch

Logos and trademarks belong to whoever owns them