This week, machine challenged man, and machine won.

“Watson”, IBM’s new ‘question answering machine’ beat Jeopardy’s longest reigning champion, Ken Jennings, handily. After three days of Jeopardy-style trivia, Watson had amassed $77,147, compared to Jennings’ $24,000.

The scene was reminiscent of ‘Big Blue’s’ victory over World Chess Champion Garry Kasparov in 1997. However, Watson did more than just compute probabilities faster than a human being – a defeat we’ve grown accustomed to over the years. Answering jeopardy questions requires a more nuanced knowledge – the ability to solve riddles, read between the lines, identify subtle cultural references, and make educated guesses. In short, it requires the ability to interpret, listen, judge…more than just the computation of ones and zeros.

The implications for Watson are obvious, and perhaps endless. The day after Watson’s jeopardy victory, IBM executives announced that they are partnering with Columbia University and the University of Maryland to build a ‘physician’s assistant’ that will provide doctors with consultative advice on patient diagnoses. A company called Nuance Communications will add voice recognition to the tool, and it could be available to hospitals within 18 months.

For nearly three millennia, humans have searched for unique traits that render them unique among the animals. Among these have been reason (think Descartes, Kant, etc.) and language (the later Wittgenstein, Rorty, etc.). Watson, however, pokes holes in both of these ideas. In a sense, Watson has proved itself to be a reasoning animal that can aggregate information, make instinctual judgments based on sophisticated weighing of tradeoffs, and communicate those intentionally and intelligibly.

But there is more to the human experience, and Watson is far from replacing us. In Watson, I see an opportunity to be more thoughtful about what exactly it is that we do well as human beings and what it is we ought to own, cherish, and cultivate as uniquely ours.

As a first attempt, I see three areas where Watson falls short of its human competitors:

1)    Watson does not have the power to pose the question. Humans are still the inquiring animal – the inventors of the scientific method and the written language. As one story on Watson concluded, “The essence of being human involves asking questions, not answering them.”

2)    Watson does not have the power to join with others in the collective pursuit of answers. Watson acts alone – aggregating and judging against its own set of data. Humans, however, can connect, organize, collaborate, and produce answers that are greater than the sum of their individual efforts.

3)    Watson does not have the power to make ethical judgments about its knowledge. It can tell you what the best answer to a factual question is, but it cannot tell you what the good answer is to an ethical quandary. Ethics is not the domain of humans alone (studies show that even dogs see injustice in the unequal distribution of treats), but its reasonable to assert that we still do it best. Perhaps Aristotle was right when he wrote in the Nichomachean Ethics that, “Human beings…choose the good.”

So, my dear Watson, you have a long way to go. You may have joined us as a reasoner and communicator, and we congratulate you. But until you can inquire, organize, and pursue the ethical good on your own terms, you will remain just what you are – a machine.

Advertisements