Apocalypse Not

By Chris O''Malley

Just south of Silicon Valley, where he toiled for many years as a computer engineer, Tim May is spending his retirement in the picturesque hills of Corralitos, Calif. But he's not there simply for the view. May believes his spot in this rich agricultural and fishing area might spare him the hardships of a famine ushered in with the new millennium, and he's ordering gold coins and laying in food in bulk just to be sure. He's also buying weapons, adding regularly to his growing gun collection. In the coming months, says May, more and more Americans are going to realize that "we aren't going to make the deadline."

The deadline? Some sort of cultish prophecy, perhaps? Nope. He's talking about the famous Year 2000 computer bug, or Y2K, as the acronym-happy computer industry has dubbed it. On Jan. 1, 2000, May and many other software experts believe, millions of computer systems could go haywire, shutting down life as we know it and turning our information age into a digitally dysfunctional society. Electric and phone service could be lost. Banks and supermarkets shuttered. Life savings could vanish and lives be imperiled.

Really? No, not really. Well, not likely anyway. But that hasn't slowed the mounting angst over the Year 2000 glitch, particularly on the Internet, where the mix of technical savvy and suspicion is proving to be the perfect outlet for dire predictions. "I've never seen such hysterical projections, and I lived through the paranoia of the 1960s," says Nicholas Zvegintzov, president of Software Management Network, a Los Altos, Calif., company specializing in software maintenance.

Still, there are grim reports that corporate America isn't taking the Y2K task seriously enough, a danger that will be highlighted this week at Senate subcommittee hearings. The Senators will be told that an examination of recent Securities and Exchange Commission filings found that 60% of the U.S.'s biggest companies had not completed the first step of assessing whether their systems are ready for the new century--much less begun fixing the problems. "That's a disturbing percentage at this late date," says Steven Hock, president of Triaxsys Research, which studied the SEC filings.

But the flaw in such reports is that they mistake reticence for inaction. Most companies are purposely saying as little as possible about the issue in their public disclosures. Hock concedes that the sec's information request was both voluntary and vague, and that big companies have a natural tendency to clam up about business operations and potential liabilities. The majority of firms, says Hock, do seem to be focusing on their most critical systems--a strategy that should spare us from catastrophic consequences. "There will be some extremely annoying, disruptive failures," says Hock. "But it's not going to be the apocalypse."

The bug at the center of the Year 2000 mess is fairly simple. In what's proving to be a ludicrously shortsighted shortcut, many system programmers set aside only two digits to denote the year in dates, as in 06/15/98 rather than 06/15/1998. Trouble is, when the computer's clock strikes 2000, the math can get screwy. Date-based equations like 98 - 97 = 1 become 00 - 97 = -97. That can prompt some computers to do the wrong thing and stop others from doing anything at all.

Left uncorrected, the Y2K problem could certainly wreak havoc. But for the most part, it is being corrected--at a collective cost estimated at anywhere from $50 billion to $600 billion. And that does not include what the Federal Government, which is dramatically and perhaps dangerously far behind in dealing with the Y2K issue, will spend (see box). However, the question of how widespread and serious Year 2000 failures will be remains a highly contentious one, often pitting those with a Year 2000 fix or book to sell against those with economic turf and reputations to defend.

This much is beginning to emerge from the fog of claims and counterclaims: while there are more than a few fatalists like May, most of the folks responsible for fixing the nation's electronic infrastructure actually think we're going to make it into the next millennium with only minor, if any, disruptions of vital services. There are technical reasons for their saying so, and probably a few public relations ones too. But the primary motivator may be that those companies have survival instincts of their own. "There's no way to overestimate how important this problem is to our customers and to us," says Michael Sansolo, senior vice president at the Food Marketing Institute, which represents most of the country's food wholesalers and retailers. "If the store's not open with food on the shelves, we don't make any money."

Many of the most dire Y2K scenarios are predicated on the assumption that the glitch will KO the country's electric utilities, turning out not only your lights but everything from the pumps at the gas station to the Slurpee machine at the 7-Eleven. It's a plausible theory. The conventional and nuclear power plants that produce our electricity are all controlled to some degree--usually a large degree--by computers, and some of the suspect programs are etched directly onto silicon chips, making them even harder to find and fix. Some utilities have only recently begun the process of ferreting out potentially weak links in their delivery systems. Worse, since most utilities are linked to one another in gridlike fashion, there could be a domino effect, turning local failures into regional blackouts.

But there are reasons to believe that blackouts may be averted or, at worst, short-lived. Most of the larger utility companies have shifted their Y2K efforts into high gear and are speeding their repairs by sharing their findings through an online clearinghouse set up by the Electric Power Research Institute. The utilities are also under regulatory pressure. The Department of Energy has set a July 1 deadline for power companies to provide assurances that Year 2000 problems will be remedied in time.

The real assurance, however, may be that many utilities aren't counting on complete success. Rather, most plan to have extra people and manual work-arounds in place for critical systems, according to Jon Arnold, chief technology officer at the Edison Electric Institute, which represents the public utilities that generate more than three-quarters of the country's electricity. "People forget that electric utilities have equipment failures and outages all the time," says Arnold. He acknowledges that "it's not going to be a typical New Year's Eve" in 1999. But, he adds, Y2K "is not like a storm or a random failure. We know this one is coming."

Whether or not you see the street-lamps glowing as you head home from your auld lang syne, you will probably be able to call Aunt Margaret the next morning and wish her a Happy New Millennium. AT&T, for one, says it will have fixed and tested most of its heavily computerized network by the end of this year, and will spend next year making sure its systems work with upgraded gear at other phone companies.

It seems increasingly unlikely you will need to convert your mattress into a safe-deposit box either. Many in the numbers-laden banking and securities industries have been pouring money and manpower into the Y2K problem for two or three years. Their megamergers notwithstanding, Citibank and NationsBank each say they will be done with their internal fixes this year, and are scheduling tests to make sure money can flow through the Federal Reserve. The Securities Industry Association, meanwhile, has scheduled a Year 2000 "dress rehearsal" next month for its members, who handle about 90% of the trading volume in the U.S., with a full-blown readiness test scheduled for March 1999.

O.K., you will have money, but will there be anything to buy? You bet, say manufacturers. "Trucks will be rolling on Jan. 1, 2000," insists Bill James, a vice president at the Grocery Manufacturers of America, which represents large food- and consumer-product makers such as Kellogg and Procter & Gamble. James fully expects some computers to crash, but he points out that there's typically a 15- to 20-day supply of food in the distribution pipeline, which should cover any short-term shortages. "That [extra supply] is not something we're normally proud of," he says, "but it might come in handy."

Paying for your purchases may involve a few misadventures, however. Credit cards with "00" expiration dates are already in circulation from Visa, MasterCard and American Express, and there are reports of unwarranted rejections. But most larger retailers have either updated their card-swiping systems or are in the process of doing so.

Not that there won't be critical problems, especially in the health-care industry. Those medical devices we've come to rely on are made by hundreds of different companies, and less than half of all hospitals have coordinated efforts to check their Year 2000 readiness, according to Joel Ackerman, executive director of the Rx2000 Solutions Institute, a nonprofit group that educates health-care providers on the Y2K bug. Then there are the thousands of far-flung clinics, labs and pharmacies that need to check their equipment. "I do believe there are going to be some unnecessary deaths," says Ackerman.

But even here the anxiety may outstrip reality. Tests suggest that only a small percentage of medical devices could be impacted by the date change, and fewer still will directly affect patient care. Despite some predictions, makers of implanted devices such as pacemakers and defibrillators say these devices will not be affected since they use sensors and not date-based calculations to time the delivery of their electrical impulses.

And while taking the stairs may be good for your heart, you won't have to do it for fear the elevator will fall--another dire bit of misinformation making the Y2K rounds. "An elevator doesn't need to know the date to go up and down, so we never put date-sensitive controls in there," says Peter Kowalchuk, a spokesman for Otis Elevator Co. Other major elevator makers have issued similar disclaimers.

In fact, fewer than 5% of the chips embedded into things like elevators and medical devices are at risk for Y2K foul-ups, says Jim Duggan, research director for the Year 2000 program at the Gartner Group. That's still a lot of chips to be checked, but even many of the suspect ones are programmed with manual overrides or "soft-landing" outcomes where safety is an issue. (Nonetheless, the Gartner Group estimates that litigation costs over Y2K service and product failures, both real and imagined, could soar to $1 trillion or more.) Duggan's forecast for the impact of Jan. 1, 2000, sounds like a tolerable weather report: "It's going to be like a couple of inches of snow that stays on the ground for a few days."

So why all the millennial storm clouds? Mainly because the Year 2000 bug is what Paul Saffo at the Institute for the Future calls a "low-probability, high-consequence event." Says Saffo: "There is great danger in our cross-dependencies with computers, but there are also lots of humans in the loop making judgments." In the end, says Saffo,"the most likely scenario by far is that we'll muddle through."

As a futurist, however, Saffo does see an opportunity in all this to help distant generations. "For God's sake, let's go to five digits [for calendar dates in software] so we don't have a Y10K problem," he advises. Amen.

--With reporting by Declan McCullagh/Washington

With reporting by DECLAN MCCULLAGH/WASHINGTON