IBM insists that its revised AI strategy – a smaller, less world-changing ambition – is working. The job of reinvigorating growth was given to Arvind Krishna, a computer scientist who became chief executive officer last year, after leading a recent overhaul of IBM’s cloud and AI businesses.
But the great visions of the past are gone. Today, rather than being an acronym for technological prowess, Watson stands out as a sobering example of the pitfalls of tech hype and arrogance surrounding AI.
The journey of artificial intelligence through the mainstream economy will, it turns out, be a step-by-step evolution rather than a cataclysmic revolution.
A new wave to go
Many times throughout its 110-year history, IBM has opened up new technology and sold it to corporations. The company dominated the mainframe market to the point of being the target of a federal antitrust lawsuit. PC sales really skyrocketed after IBM entered the market in 1981, confirming these small machines as essential tools in corporate offices. During the 1990s, IBM helped adapt its traditional business customers to the internet.
IBM executives have seen AI as the next wave to push.
Mr. Ferrucci first presented the idea of Watson to his bosses at IBM’s research lab in 2006. He thinks building a computer to solve a question-answering game could boost science. The first learning in the field of AI is known as natural language processing, in which scientists program computers to recognize and analyze words. Another research goal is to improve automatic question-answering techniques.
After overcoming initial skepticism, Mr. Ferrucci assembled a team of scientists — eventually more than two dozen — who worked out of the company’s lab in Yorktown Heights, NY, about 20 miles north of IBM’s headquarters in Armonk.
The Watson they built was a supercomputer the size of a room with thousands of processors running millions of lines of code. Its archives are filled with digitized references, Wikipedia entries, and e-books. Smart computing is a rough job, and the processing machine claims 85,000 watts of power. In contrast, the human brain runs the equivalent of 20 watts.