Eric Brown of IBM Research, near right, with members
of congress – Jared Polis, Bill Cassidy, Jim Himes,
and Rush Holt, far right.

Over the past several decades, I have faced off against 14 opponents in the game show “Jeopardy!” All were smart; all were tough; all were quick-witted.

Most were human.

The exception, IBM’s Watson — a remarkable machine that is a testament to the promise of scientific research and development — first proved its “Jeopardy!” dominance on national TV one year ago. I was among five members of Congress who, shortly afterward, had a chance to experience the system up close. I even won my round, although the system went on to defeat our congressional team as a whole.

What does it feel like to face off against what is arguably the smartest machine on the planet? Daunting, to be sure — like mounting a footrace against a Ferrari. Yet aside from intimidation, I can remember three other powerful feelings.

The first was awe at the rate of technological progress. When I first appeared on “Jeopardy!” as a graduate student in the 1970s, the fastest commercially available computer in the world was a Cray supercomputer operating at 80 MHz. By comparison, the smartphone I was carrying during last year’s match ran at 600 MHz. Incredible as it sounds, I had more computing power in my pocket in 2011 than many world powers could have obtained a few decades ago.

Watson is far more powerful still: It ran on 90 servers, each with about 3,500 MHz of processing speed. But the ingenuity of its developers was applied not just to how fast the computer operated, but also to how it approached problems — not just looking up facts in a massive database, but coming closer to what I’d call “thinking.”

My second feeling was a sense of vast possibility. Already, Watson is being put to work in the medical community. Imagine you’re a doctor faced with a patient who isn’t very good at describing his symptoms, whose symptoms are not textbook indicators of a particular disease. What do you do? Well, the situation is not unlike Alex Trebek offering up an answer instead of a question, in the form of a pun instead of a direct statement.

Other data-intensive industries such as retail, transportation, and financial services could soon benefit, gaining insights that humans could not have achieved on our own. The prospect also promises new jobs: The United States will need 190,000 more workers with the ability to analyze massive amounts of data in order to achieve future breakthroughs.

My third and most powerful feeling was a sense of urgency. The progress that led to Watson and to my smartphone was not, after all, inevitable. Progress is not a wind blowing always at the back of humankind. Progress is, rather, something we create for ourselves: a plow that we push forward through hard work, imagination, and investment.

Yet big-dream investment has become rare. Over the past 50 years, federal investment in research and development has fallen by two-thirds as a share of our economy. Private investment, such as IBM’s support for Watson, has filled some of the gap, but not nearly all of it.

Cuts in federal support are foolish in light of all of the evidence showing that R&D pays off big. The Human Genome Project, for example, was kick-started by $3.6 billion in federal support and is estimated to produce $800 billion in economic output.

President Obama’s 2013 budget proposal would at least stop the bleeding; federal support for R&D would hold firm. Yet this is not enough to restore research and development to their rightful place as key pillars of the American enterprise. If we, as a nation, fail to make the necessary investments, then we will deprive ourselves and our children of their payoffs — both financial and technological. What a grim prospect!

One year later, Watson remains a bright example of the way that research can make our future better. Now, it is up to us to make that future real.

Rep. Rush Holt, a Democrat, represents New Jersey’s 12th Congressional District. This column appeared originally in the Star Ledger newspaper.