[This post was originally published on April 13, 2012.]
In fiction, computers subdue mankind by nuclear war or by despotic control of our very thoughts, but in real life, they will probably just downsize us to death. In February of 2011, in a two-game, nationally televised Jeopardy! match, IBM’s new supercomputer Watson took on Ken Jennings and Brad Rutter, the two most successful contestants of all time. IBM saw the contest as a golden opportunity to market its new “Deep QA” technology, which enabled Watson to answer questions posed in natural language quickly and accurately, the way a Jeopardy! player responds to clues and the way that futuristic computers have long been portrayed in movies like Star Trek and 2001: A Space Odyssey.
I played a small part in the preparation for the contest: I was one of forty Jeopardy! Tournament of Champions participants who was invited to compete against Watson in test matches to help IBM engineers polish their programming. In fact, in one of the last dozen games played before the televised match, I defeated Watson. Sure, I am proud of it now, but when one of Watson’s descendants takes over the world, I’ll be one of the first to get a visit from the cyber-Stasi.
In any case, IBM learned from its mistakes, and in the end Watson trounced Jennings and Rutter. The splashy debut seemed to have the desired effect; over the past year, IBM has successfully sold the use of the Watson technology to entities like Wellpoint and Memorial Sloan-Kettering, to be used for medical diagnosis and treatment planning, and there are plans to extend the use of the technology to such uses as financial analysis and legal research.
It was clear that Watson was indeed pretty good at answering Jeopardy! clues. Not necessarily better than the top players, but certainly better than the average player – which is indeed an amazing feat on IBM’s part. But knowing the answers was of much less consequence than Watson’s inhuman speed in “buzzing” in to answer the question. Once their signaling devices are activated, good human players can buzz in within a tenth of a second or two, but, because of its electronic reflexes, Watson was able to buzz in considerably less than a tenth of a second. Thus, purist fans of Jeopardy! dismissed the Watson match as a gimmick. Any merely competent Jeopardy! player guaranteed first crack at 80% of the clues could beat the likes of Jennings and Rutter. In essence, they argued, the Watson match was not fair. And that’s true, in some sense, and maybe even matters in the quiz show world. But in the business world, you would be wise not to rely on “fairness” to protect your job.
A character in Robert Harris’s new thriller, The Fear Index, points out that in the 1950s, people used to think that robots would take over all the mundane physical aspects of human drudgery, like vacuuming and ditch-digging, but the reality of the 21st century is that machines are really replacing humans in what has traditionally been the mental arena. After all, that is really the area in which our technology advances most regularly: the processing of increasingly huge amounts of data. But this is part of the space most professionals inhabit. We rely on doctors, lawyers, investment advisors, and the like for their comprehensive knowledge and their judgment. We are rapidly coming to the point, however, where no person’s comprehensive knowledge will be comparable to the amount and complexity of the data that an advanced computer will be able to access, sort through, and process.
In the legal world, for example, much of the work done by young associates consists of this kind of data processing — reviewing thousands of documents for substance or privilege, for example, or performing exhaustive research in statutes, regulations, and cases to determine all the law relevant to a particular issue. Computer programs already facilitate this work by giving lawyers access to more information more quickly. But it may not be too long before programs with Watson-like language processing ability are able to take over this kind of work completely — eliminating the need for expansive teams of associates with expensive requirements like food. When that happens, what will become of those young professionals? When companies and firms are deciding whether to employ 20 associates, or one computer at one-tenth the cost, “fair” will go right out the pod bay doors.
We all need to face the fact that any modern worker — not just unskilled laborers but well-educated professionals — may find that some of their skills will quickly become devalued in the workplace as technology advances. It is a good idea to cultivate a new mindfulness, an awareness of which abilities might soon be duplicated and which ones are still uniquely human, and to work to develop the value of the latter. Judgment, creativity, interpersonal skills — these are essential work tools that are not going to be replaced by computer in the foreseeable future (at least, not yet), and they have always been valuable qualities in their own right. While we must all still master the nuts and bolts of knowledge and research in our practice areas, for long-term career fulfillment it is now more important than ever to develop our essential, and essentially, human abilities.