Disappointment with Watson a Matter of Expectations?

An original Healthcare Research & Analytics survey found that many in healthcare know the name, but not the potential.

IBM again posted disappointing financial returns in the second quarter of 2017. Even before IBM posted another losing quarter, it and its artificial intelligence capstone, Watson, were drawing snipes and being called “a joke” on national television. For the past month, the company and project have received a fair share of derision on social media, with an enormous exploration from Gizmodo last week highlighting many points of criticism.

Recently, Healthcare Research & Analytics, the research arm of Healthcare Analytics News’ parent company, conducted an exclusive survey of roughly 200 physicians in both private and hospital practice to determine their awareness of various AI technologies from major tech companies with healthcare intentions. Overwhelmingly, fewer doctors reported being “Not at all familiar” with Watson than they did with Morpheo, Microsoft’s Hanover, Intel’s Saffron, or Google’s DeepMind (a thorough overview of the results will be published in our September print issue and online).

Across the board, few physicians were currently aware of AI tech being used in their particular workplace, but the most telling group might have been the second option. Of IBM Watson, 30% of the physicians polled responded that they “Know the name, but do not know much” about it. While simultaneously having the best name recognition and understanding metrics in the survey, it also had the most people who had heard of it but didn’t quite understand it.

Watson roared into the national consciousness with a game show victory and star-studded television commercials. The same expectations followed it into its healthcare forays, with health systems still frequently announcing collaborations with Big Blue’s thinking machine.

It’s these inflated expectations that many are pointing to as the root of recent derision. In Healthcare Analytics News’s September cover story on AI expectations, Ambra Health’s CEO concurred: “It’s not likely that by next Christmas machines are going to be curing cancer. I suspect that Watson sort of made it feel like that reality is already here, when there’s still a lot of work to be done,” Morris Panner said.

Other figures are harsher. The recent Gizmodo rundown features Allen Institute for AI CEO Oren Etzioni saying that “the emperor has no clothes” and describing IBM’s PR and marketing as having “run amok.”

Though expectations are high and delivery may still remain ambiguous, confidence remains in many minds. Watson’s health partnerships are extensive, and institutions continue to put trust into the effort and share their data with it. This week, it was reported that the Association of Public-Safety Communications Officials (APCO) would team up to allow Watson to listen in on and analyze 911 calls in hopes of helping call centers to improve call times and outcomes.

The technology still has the potential to awe, as evidenced by a study of its potential in precision medicine from last month, which joined a long line of small but verifiably useful findings produced by the technology. With the AI market becoming more competitive by the week, however, it remains to be seen what IBM Watson and its health partners will produce to quell derision.