Today’s fear that AI might wipe out vast swaths of jobs across the economy—not in some misty future, but in just a few years—is a relatively new phenomenon. But 12 years ago, it was becoming clear that technology was starting to outperform humans in some categories—as rather painfully demonstrated to me when I was roped into competing with IBM’s Watson, the cognitive computing system, in a game of Jeopardy! at a retail industry conference. The early A.I. software trounced me.
My 2014 Fortune feature did not foresee the AI revolution, instead focusing on the rapid advancement of computer processing power, itself an astonishing and alarming phenomenon. “In a decade, [computers] will be 32 times more powerful,” I wrote. Today, AI engines double their power in a few months. If we say their power doubles every three months, then five years from now they will be about 1,000,000 times more powerful. Truly, the mind boggles.
The very concept of near-universal job destruction was a mind bender in 2014 because history had disproved it. The new technologies of the Industrial Revolution eliminated vast numbers of jobs in various industries over the decades, but in the long run the system created more jobs in new industries, and those jobs overall were more productive than the lost jobs. The result was an economic miracle of rising living standards.
The fundamental issues identified in the article remain the same, but more pressing. The article ended on a question that is arguably still the bottom line: “As computers begin to acquire some of the most advanced cognitive and physical human skills, we confront a new reality. In a way that has not been true before, the central issue for the economy and for all of us who work in it will be the answer to the question: What will people do better than computers?”
