Cognitive Computing: What Can You Do With It? Two Satisfying Use Cases

Cognitive Computing: What Can You Do With It? Two Satisfying Use Cases

December 10

In my last post I introduced you to cognitive computing, mostly via IBM’s Watson, which defeated two all-time Jeopardy! champions through its ability to take terabytes of random data and formulate the correct questions—along with its uncanny ability at wagering down to the last dollar.

But what else can a system like Watson do besides being an ideal partner for Trivial Pursuit? Here are a couple of use cases, one obvious and one not so obvious, for you to chew on.

Medical Evidence

Back in April, I wrote several posts about EHR (Electronic Health Records), including How EHR came to be (or why doctors aren’t like airline pilots). In that post I discussed how EHR, at least in an ideal scenario, would give doctors complete visibility into a patient’s overall health. But even a single patient’s EHR may contain dozens of tests, diagnoses from multiple doctors, treatment plans, and various prescriptions over that patient’s lifetime. How does a doctor map out the best treatment for that patient based on the latest research, let alone do so for all of her patients?

Watson can help. According to a 2013 Forbes article written by Bruce Upbin, Watson had analyzed “605,000 pieces of medical evidence, 2 million pages of text, 25,000 training cases and had the assist of 14,700 clinician hours fine-tuning its decision accuracy.”

Upbin goes on to write:

Watson doesn’t tell a doctor what to do, it provides several options with degrees of confidence for each, along with the supporting evidence it used to arrive at the optimal treatment. Doctors can enter on an iPad a new bit of information in plain text, such as “my patient has blood in her phlegm,” and Watson within half a minute will come back with an entirely different drug regimen that suits the individual.

IBM’s Watson Group already has established several partnerships with medical groups, most notably MD Anderson Cancer Center and WellPoint.

Chef Watson

Would you ever consider using green peas, sweet pepper, pork belly, and cottage cheese in a dish? Me neither. Watson recommended combining these four ingredients to James Biscione, director of culinary development at New York’s Institute of Culinary Education (ICE), and Biscione says the moussaka he and his colleagues made from these ingredients is one of the best things he’s ever tasted.

IBM research scientist Dr. Lev Varshney says that most chefs can “reason” two ingredients, with the best chefs capable of reasoning three that can be combined for a good dish. However, virtually no one can put together four ingredients. But Watson can. In this same video Biscione shows how he gives Watson an ingredient like lobster, sets up the parameters of the dish (style of cooking, types of ingredients to include.), and “spins the wheel” to see what ingredients the Watson database comes up with. According to Dr. Varshney, Watson can put together “quintillion” combinations that are based on such factors as what ingredients go well together, regional cuisines, degrees of pleasantness and surprise, and other things modeled on both food chemistry and human taste perception.

However, Watson doesn’t come up with the actual recipe. It’s up to the chef to figure out what to do with these ingredients. In a blog post Watson group head Michael Rhodin discusses why this type of collaboration is so groundbreaking:

Chef Watson is so much fun that it may seem trivial to some, but it demonstrates a powerful new force that’s being unleashed. Think of it as a discovery engine. And think of such engines as essential players in a new age of discovery.


Photo credit: Rev314159 via Flickr