IBM exploring new feats for Watson
IBM’s Watson beat “Jeopardy” champions two years ago. But can it whip up something tasty in the kitchen?
That is just one of the questions that IBM is asking as it tries to expand its artificial intelligence technology and turn Watson into something that actually makes commercial sense.
The company is betting that it can build a big business by taking the Watson technology into new fields. The uses it plans include helping to develop drugs, predicting when industrial machines need maintenance and even coming up with novel recipes for tasty foods.
The new Watson projects – some on the cusp of commercialisation, others still research initiatives – are at the leading edge of a much larger business for IBM and other technology companies. That market involves helping corporations, government agencies and science laboratories find useful insights in a rising flood of data from many sources – Web pages, social network messages, sensor signals, medical images, patent filings, location data from cellphones and others.
Advances in several computing technologies have opened this market, now called Big Data, and a key one is the software techniques of artificial intelligence like machine learning. Watson initiatives, analysts say, represent pioneering work. With some of those applications, like suggesting innovative recipes, Watson is starting to move beyond producing “Jeopardy”-style answers to investigating the edges of human knowledge to guide discovery.
IBM’s Watson projects are not yet big money-makers. But the projects, according to Frank Gens, chief analyst for IDC, make the case that IBM has the advanced technology and deep industry expertise other suppliers cannot match.
John Baldoni, senior vice president for technology and science at GlaxoSmithKline, got in touch with IBM shortly after watching Watson’s “Jeopardy” triumph. He was struck that Watson frequently had the right answer, he said, “but what really impressed me was that it so quickly sifted out so many wrong answers.”
That is a huge challenge in drug discovery, which amounts to making a high-stakes bet, over years of testing, on the success of a chemical compound. The failure rate is high. Improving the odds, Baldoni said, could have a huge payoff economically and medically.
Glaxo and IBM researchers put Watson through a test run. They fed it all the literature on malaria, known anti-malarial drugs and other chemical compounds. Watson correctly identified known anti-malarial drugs and suggested 15 other compounds as potential drugs to combat malaria. The two companies are now discussing other projects.
“It doesn’t just answer questions; it encourages you to think more widely,” said Catherine E Peishoff, vice president for computational and structural chemistry at Glaxo. “It essentially says, Look over here, think about this. That’s one of the exciting things about this technology.”
IBM researchers began working with Thiess, a large contract mining company, in Australia last year. Thiess operates an equipment fleet worth $3 billion, mostly very large machines. Its hauling trucks, for example, have 12-foot diameter tires and carry 250 tons in a single load.
Technology-enhanced predictive maintenance on machinery, like jet engines, has been developing for years. But the Thiess project seems to push things further to cover mine operations as a whole. The data include not only information from the 200 sensors on a truck that monitor trips, load weights, speed and driving styles but also weather, terrain and economic models of mine operations.
Watson has been able to deliver complex predictive analytics, said Michael Wright, an executive vice president at Thiess. Data-driven changes in operations are being implemented, and savings will be measured over the next six months, he said.
In a San Jose meeting with analysts, IBM plans to serve them a breakfast pastry devised by Watson, called a “Spanish crescent.” It is a collaboration of Watson’s software and James Briscione, a chef instructor at the Institute of Culinary Education in Manhattan.
IBM researchers have watched and talked to Briscione as he works, selecting ingredients and building dishes. Watson has read those notes, 20,000 recipes, data on the chemistry of food ingredients and ratings of flavors people like in categories like “olfactory pleasantness.”
Watson’s assignment has been to come up with recipes that both are novel and taste good. In the case of the breakfast pastry, Watson was told to come up with something inspired by Spanish cuisine but unusual and healthy. The computer-ordered ingredients include cocoa, saffron, black pepper, almonds and honey – but no butter, Watson’s apparent nod to healthier eating.
Then, Briscione, working with those ingredients, had to adjust portions and make the pastry. Michael Karasick, director of IBM’s Almaden lab, had one of the Spanish crescents for breakfast recently. “Pretty good” was his scientific judgment.