HomeIndustry Commentary

Another pat on the back for IT

Like Tweet Pin it Share Share Email

My all-time favorite editing gaffe garbled a column I wrote about Y2K.

What I wrote: “The money saved dwarfed that spent on remediation.”

What InfoWorld printed: “The money saved the dwarfs that spent on remediation.”

I felt like Thorin Oakenshield with a corrupted database.

Speaking of Y2K, my recent column on COVID-19 and what you should do about it (“When Corona isn’t just a beer,” 3/2/2020) included a reminder of the KJR Risk/Response Dictum: Successful prevention is indistinguishable from absence of risk. I used the global, effective response to the H1N1 virus as an example.

Several correspondents reminisced with me about another, even better example: Global IT’s astonishingly effective response to the Y2K bug, and the ensuing certainty among the ignorati that it was all a hoax.

Y2K’s outcome was, in fact, a case study in what David Brin calls self-preventing prophecy. In the case of Y2K the problem of using two digits to represent the year in date fields, with the 19 prefix assumed, was indisputably real. The potential impact should the world fail to correct the problem was, in the aggregate, unknown and probably unknowable. Concerns ranged from the mundane — employees and customers who, according to HR and CRM systems, would have had negative ages — to the alarming but unlikely possibility of computer-controlled elevators plummeting down their shafts.

For a more in-depth account, read “The Lessons of Y2K, 20 Years Later,” Zachary Loeb, Washington Post, 12/30/2019.

Pre-COVID-19 we knew the overall risk of a viral pandemic soon enough to be worth investing in advance preparedness was high. Which virus, exactly when, exactly how contagious and exactly how virulent? Of course not. The Y2K problem was definitive. COVID-19? The lack of in-advance specifics made, for some decision-makers, the fourth risk response (hope) attractive.

About all we know about the risk of future pandemics is that it’s increasing. That isn’t in any doubt because (1) a pandemic only needs one sick person to get things started; (2) every year, Earth has more persons who could become that one sick person; and (3) every year, more and more people travel to more and more destinations, and “more and more” means a higher likelihood that the one sick person could cross borders to spread their disease more widely.

But never mind all that. Observing the global response to COVID-19, we in IT should be busily patting ourselves on the back again … washing our hands before and after we do, of course.

We deserve the back-patting because if it weren’t for IT, and specifically if it weren’t for our investments in: electronic mail; internal chat; file sharing technology; web conferencing systems; secure remote access to business applications; along with, I hope, broadly available training in their use, coupled with, at this stage of our evolution, peer pressure to master at least the basics coupled with peer knowledge-sharing to provide informal support … if the world of commerce hadn’t embraced these technologies and the idea of remote workers they support, your company’s Business Continuity Plan, sub-section Pandemic Response Plan, would be pretty much worthless.

And right now, if it weren’t for these business innovations that quietly took hold over the past decade or so, the current pandemic’s impact on the world economy would be quite a lot worse.

It’s only ten years ago that I wrote “10 sure-fire ways to kill telecommuting” for InfoWorld (3/30/2009). Some readers got the joke. Even those who thought I was serious recognized that telecommuting was far from universally accepted among business leaders and managers.

Among evolutionary theorists, this sort of thing is called a “preadaptation.” It means a species develops some heritable trait or behavior because natural selection favored it for an entirely different reason. Sometime in the distant future the species makes use of it in some entirely different way that gains an entirely different advantage.

For example, fish developed swim bladders to control their buoyancy. Long, long afterward the swim bladders they had as fish evolved into the lungs they needed as amphibians.

Likewise what we used to call telecommuting and now call remote work. Organizations didn’t embrace it because it would make them more resilient in the face of a global pandemic. They embraced the practice because it reduced the cost of business infrastructure, gained access to a broader pool of talent, and let companies construct project teams out of a broader array of employees.

The moral of this story: You can’t predict all the ways a new technology might create value. So don’t let your governance committees stifle experimentation. You never know when an experiment might turn out to be a preadaptation.

What you do know: If you prevent the experiments then they won’t.

Comments (6)

  • One of the ironies of Y2K, is that it came back to bite us quite recently. One of the remediations used for some of my employer’s systems was to assume that all two digit years less than 20 were from the 21st century, and all years of 20 or greater were 20th century. And of course, nobody who made those changes 20+ years ago is still around, so in January, those systems did some strange things.

    A great effort in 1997-99, but less great follow up in 2019.

  • “You can’t predict all the ways a new technology might create value.”
    You also can’t predict all the ways a new technology might dissolve current value or negatively affect other things we’re already used to. I think that’s where a lot of folks come from.

  • Y2k was (at least) two different problems. One was a raft of older programs, many of them in COBOL, that used two digits to represent the year and whose logic would fail without amendment. There was a major effort needed to identify all such programs and rectify them, and this was done and worked pretty well.

    The other problem was the BIOS in older PCs which potentially could have similar issues, but which would be harder to fix (apart from the coding language being abstruse, it was more difficult to update the BIOS in those far off days). Various companies offered solutions and employed salespeople to travel round frightening businesses that all their PCs might fail on 1/1/2000. Recent PCs with a newer BIOS generally didn’t suffer from this problem.

    I was responsible for several thousand PCs spread across Europe. Rigorous testing revealed a small handful that had a real problem – all very old systems that were in need of replacement in any case. A more substantial minority would come up with a wrong date (usually 1900) upon their first reboot after the millennium, and once the date was correctly entered manually would be fine thereafter. The rest (the great majority) had no problem.

    My conclusion was that the ‘COBOL’ (other computer languages are available) problem was a genuine issue; while the PC problem was mostly snakeoil.

    • I think there was more. In addition to systems themselves were supply chain certifications. For example I recall one reader telling me his company, which sold concrete, had to certify that the sand they used was Y2K compliant.

  • Thanks Bob, I’m sure you’re right. Trying to make sense of such a bizarre certification, I guess they may have been asked to certify that they got their sand from a supplier whose systems were Y2k compliant.

Comments are closed.