IBM is seeking more business uses for Watson’s artificial intelligence
IBM needs Watson to be more than a marketing curiosity as Big Blue tries to stay relevant in the digital age.
McDonald’s has Ronald. Geico has the gecko. And Target has Bullseye the dog.
The closest thing IBM has to a corporate mascot is Watson.
Most people remember Watson as the “thinking” supercomputer that defeated two human champions on the trivia game show “Jeopardy” in 2011. (Deep Blue, its predecessor, was famous for defeating world chess champion Gary Kasparov more than a decade earlier.)
Watson wasn’t actually a single computer but rather a collection of technologies that worked together to crunch data and solve problems. In any case, Watson not only symbolized the potential of artificial intelligence but also
the research and development power of IBM, itself a longtime symbol of American innovation.
But IBM needs Watson to be more than a marketing curiosity as Big Blue tries to stay relevant in the digital age. The company said that it will open an office at 505 Howard St. in San Francisco to aggressively sell Watson to potential customers in industries like health, finance and retail. Whereas Watson was once an experimental showpiece, IBM wants to generate real sales and profit from the technology.
‘Vision is real’
“Last year, we had a vision of what (Watson) would become,” said Stephen Gold, chief marketing officer for IBM’s Watson Group. “Today, the vision is real.”
Like Hewlett-Packard, Oracle and Dell, IBM has been trying to refashion itself from a maker of hardware into a provider of cloud software and services. Last year, IBM reported sales decreases in nearly all major business units, including a steep 23 percent decline in hardware.
Those numbers don’t bode well, especially at a company that likes to brag about spending $6 billion a year on research and development.
But there is a bright spot: Revenue from business analytics — finding insights by crunching large amounts of data — jumped 7 percent last year to nearly $17 billion. The service now makes up 18 percent (and growing) of IBM’s total revenue of $93 billion.
It’s no wonder, then, that IBM has invested more than $26 billion in big data and analytics, including $17 billion in related acquisitions, according to documents filed with the Securities and Exchange Commission. About one third of its R&D budget now goes to data, analytics and cognitive computing that can think, reason and learn.
That’s why Watson — which can now help companies do things like analyze images from social media and identify trends and patterns — is so crucial to IBM’s transformation. Last year, the company established a unit to commercialize Watson and committed $1 billion to rolling out the technology to the broader business world.
“Through Watson, IBM is extending analytics to the end user, not just the data scientist,” the company said in SEC filings. “IBM’s goal is to give every business professional access to advanced cognitive-powered predictive analytics, coupled with new forms of data.”
Venture investment
But is Watson still relevant at a time when Silicon Valley dominates the tech universe?
Today, companies ranging from Google to Adobe are rapidly building out their analytics businesses. From June 2014 to July 2015, venture capitalists poured $6.4 billion into big data startups, including $4 billion for infrastructure and machine learning, according to Quid, a San Francisco analytics company.
CB Insights recently reported that investors including Google and Khosla Ventures dumped $309 million into artificial intelligence startups last year, a 300 percent increase from 2013.
In an ideal world, IBM would have found commercial uses for Watson after its resounding victory on “Jeopardy” four years ago, which attracted enormous attention.
“It would have been great to flip on a switch after ‘Jeopardy,’ ” Gold said, but the technology wasn’t ready.
Since then, though, “Watson has really turned a corner.”
For example, Watson has greatly improved its tools to synthesize natural-sounding speech from text. Because the technology lets developers build products and applications that understand intent and meaning, it can find answers for users even when questions are asked in varying ways.
Since last year, more than 100 companies that have worked with IBM on Watson have introduced cognitive-enabled apps, products and services into the market.
“I like to think we are emphasizing our leadership and commitment to this space with our investments,” including the new office in San Francisco, Gold said. “Silicon Valley is pivotal to the work we are doing.”