As I See It: History Makers

In the early 1960s, when IBM abruptly bolted from New York City, Armonk was a remote and barely inhabited woodland. On my only visit to the IBM mothership nearly two decades ago, I remember wondering why a big-city company would select such an out-of-the-way site to house its world headquarters. Like guys with a Rolex, successful companies enjoy showing off their architectural bling-bling. Power and status find their expression in steel and glass and are typically flaunted, not hidden.

Cheap land, I thought, may have been an inducement, but uber-wealthy people don’t buy things just because they’re cheap. There were, I assumed, other compelling reasons for IBM’s move that were simply unknown to me. I thought little more about it until recently when I came across a reference to a March 1977 article by Thomas Mechling published in a trade journal called Computer Decisions. In “Gimme Shelters: Why IBM Fled the City,” Mechling provides an historical context for IBM’s decision to move–a decision that was so controversial at the time it was kept secret for sixteen years in order to avoid public panic.

Here’s Mechling’s explanation: “The real, unwritten and unspoken reason that Thomas J. Watson wanted to get his IBM top management the hell out of mid-Manhattan. . . was to escape and survive a nuclear bombing of New York City, a likelihood seen by the most influential, inside-information sources he was uniquely privy to on the national, state, and scientific levels.”

Like many of his time, Watson was obsessed with the Communist threat. By August 1962, it was publicly known that Soviet medium-range ballistic missiles were being installed in Cuba. In response to the growing inevitability of nuclear confrontation, Watson built a fallout shelter in his Greenwich, Connecticut home and was offering his employees interest-free loans to do the same. New IBM installations, particularly those near large, potentially targeted cities, were also constructed with bomb shelters. It seems somewhat absurd now, but this was during the time that people believed they could survive a nuclear exchange and school kids practiced putting their hands on their heads and ducking under their desks to avoid being vaporized.

Now let’s move the clock ahead almost thirty years. It’s the spring of 1989, and people are no longer under the delusion that atomic warfare is directly survivable or that its impacts will be anything but long-lasting and wide-ranging. In Brazil, a team of mercenaries is preparing to outsource its specialized skills. A group of men led by General Hugo Piva, are about to test the limits of the free market. But Piva is no ordinary gun for hire. He is the former director of Brazil’s Aerospace Technology Center. According to a 1990 New York Times article, prior to retirement, the general “was in charge of converting Brazil’s latest space rocket, the Sonda IV, into a missile big enough to carry a nuclear warhead, and of secretly making nuclear weapon material by enriching uranium by using gas centrifuges.”

That spring seventeen years ago, Piva and his team of scientists were preparing to leave for Iraq.

Again, historical context is helpful. At the time, Saddam Hussein had just concluded an eight-year war with Iran and, never one to tire of slaughter, was making threats against Israel and neighboring Kuwait. Of course, his threats would acquire additional gravity and urgency if he actually had weapons of mass destruction (WMD) to back them up.

The Brazilians wanted to help Saddam in his quest, and had in fact already sold uranium to Iraq in the 1980s. But being a third-world country, Iraq lacked the necessary technology to constitute a nuclear program, and being both violent and unstable it wasn’t likely to acquire such technology legally. But the Brazilians could.

Enter IBM.

Brazil wanted to buy the latest supercomputer from Big Blue, ostensibly to be used by its aircraft manufacturing industry, but the Department of Energy and the Arms Control and Disarmament Agency knew better and objected vehemently. They didn’t want Brazil developing nuclear weapons, plus they knew the Brazilians were selling their knowledge to the Iraqis. There were already Iraqi missiles pointed at Tel Aviv, and a computer that could calculate air flows around aircraft wings could compute the same flows around missiles with fins. And, as the Times reported, “a supercomputer can simulate the implosive shock wave that detonates nuclear warheads, calculate the multiplication of neutrons in a nuclear chain reaction, and model the process of nuclear fission in a hydrogen bomb.”

Not a good toy for the bad guys to have.

But IBM, having learned nothing from the concerns of its former leader, was determined to sell the machine and found allies at the State Department and the Department of Commerce. Commerce was concerned with dwindling U.S. exports and apparently didn’t care who sold what to whom as long as the numbers went up. State, according to the Times, wanted “to win favor in third world countries,” perhaps particularly with Saddam Hussein, who was by then our ally and had been killing Iranians with our blessing and assistance for nearly a decade.

The decision was so controversial that it eventually got kicked upstairs to the first President Bush. As New Scientist reported in August of 1991: “President George Bush approved the export of the IBM supercomputer in December, overriding concern in the Arms Control and Disarmament Agency, the Department of Energy, and Congress that the machine would boost Brazil’s ability to design nuclear missiles. To complicate matters, Brazil then refused to guarantee that it would not use the computer for military projects or re-export the machine to another country.”

Oops!

Before the current Iraqi war, the joke was that we knew Iraq had WMD because we had the receipts to prove it. It’s hard not to wonder whether the dreadful consequences that followed were not, at least in part, predicated on the fact that we knew Iraq had access to the computer technology necessary to create our worst nightmares.

In any event, move the clock ahead again some ten years. IBM unveils the world’s fastest supercomputer capable of 12.3 trillion calculations per second. It’s linguistically tortured name is Advanced Strategic Computing Initiative White, or ASCI White, and it is a marvel of human engineering. It boasts 160 trillion bytes of storage and is powered by no fewer than 8,192 high-end Power processors. Spanning some two tennis courts and weighing over 100 tons, it can do more calculations in a second than a person could make in 10 million years using a calculator. And what great benefit will this wonder provide humankind? It has been delivered to the Lawrence Livermore National Laboratory, where it will be used by the government for virtual testing of nuclear weapons.

Move the clock ahead again to the present day. As wonders go, ASCI White is obsolete. IBM is set to deliver a 1.6 petaflops hybrid supercomputer (one petaflops being equal to 1,000 teraflops). The speed, power, and potential of this machine is almost beyond comprehension. (For an excellent explanation of the technology involved, see our September 12, 2006 article, IBM to Build 1.6 Petaflops Super.) And where will this work of genius go? What grand purpose will it serve? It will be delivered to the Los Alamos labs, where it will simulate aging atomic weapons and assist in the development of new ones.

Although we seldom think of them in these terms, such commercial transactions are potential history makers, even though they go largely unnoticed and its principals are wholly unaccountable for the outcomes they create.

When will we ever learn? It has been well-documented by Edwin Black in his troubling book, IBM and the Holocaust, that Thomas Watson Sr. happily supplied the Nazis with state-of-the-art tabulation equipment, which was knowingly used to track and identify Jews as the Germans swept across Europe. Nor is Iraq the only repressive country to have ever benefited, directly or indirectly, from IBM’s technological prowess. How long, one has to wonder, will it be before “old” technology such as ASCI White is sold to another unstable or hostile regime in the name of free trade or placation or simply to pad the bottom line.

The poet and philosopher George Santayana said: “Those who cannot remember the past are condemned to repeat it.” It would be funny if it wasn’t so sad to think of IBM executives sitting in Armonk atop their monument to cold-war paranoia, feeling the same illusion of safety and control once provided by backyard bomb shelters, willing and eager to sign agreements delivering technology to people who will create tomorrow’s nightmares. It is no secret that our government wants to develop a new generation of tactical nuclear weapons. When the bill for such recklessness comes due, where will they hide? Indeed, where will any of us hide?

It was Santayana who also said “A country without a memory is a country of madmen.” Given what’s going on in the world, it makes you wonder, doesn’t it?

[Editor’s historical note: IT Jungle’s own Mad Dog columnist, Hesh Wiener, was formerly editor of Computer Decisions before he started his own newsletter publishing company and created The Four Hundred, among other publications. Victor was unaware of this when he wrote this story. –TPM]