In 2011, IBM launched Watson, a sophisticated intelligence platform, with much fanfare, driven in part by its ability to beat humans in the TV game show Jeopardy. However, commercialization of the product thereafter didn’t get the expected traction.
Rather than pull back, IBM recently announced a further US$1 billion investment in the newly named IBM Watson Group. This doubling down on previous investments is intended to re-energize Watson’s mission to exploit the platform to serve data driven businesses such as insurance.
Our question to you: Will platforms like Watson impact the insurance community in a meaningful way in the foreseeable future?
Way back in 2011 …
…. IBM demonstrated the value of its new Watson supercomputing platform by competing with, and beating, humans in the television game show, Jeopardy. There were some high profile announcements about commercial applications, including health insurer, Wellpoint, which would use Watson “to help improve patient care through the delivery of up-to-date, evidence-based health care.”
Expectations were raised. IBM took Watson out to big insurance customers and enthusiasm was generated. During a 2012 world tour, Munich Re said that this type of system could advance self-service by answering broker and customer questions and “automate accurate answers to complex claims questions posed by policyholders or even third-party claimants.”
Then things went a little bit sideways
In spite of significant technical and marketing efforts, Watson has not been a commercial success. According the Wall Street Journal, by the end of 2013, top line revenue was significantly lower than IBM’s targets.
Why? Seems that human intelligence and communication are not as easily replicated by machines as anticipated. Even really smart machines. Antonio Regalado, business editor of MIT Technology Review, writes:
IBM has aggressively marketed the idea that Watson’s eerie smarts will revolutionize everything from cancer care to call centers. But real-world problems aren’t as tidy as the game show … For Watson to perform just as well in new areas, it requires a big effort to train the system and adapt it to new information
Watson may be down, but IBM is hardly counting it out.
You don’t learn to swim without getting wet
The problem in the healthcare example is not related to reasoning alone. Communications presents critical challenges. Regalado reports that the error rates in cancer evaluation were more challenging than Jeopardy questions because “unlike crisp game-show questions, doctors’ case notes are a maze of jargon, abbreviations, and inconsistently used terminology.”
We’ve all seen doctor’s written prescriptions, right? To IBM’s credit, Watson was not unveiled as a fully formed solution. And the decision to increase its investment recognizes that mistakes must be made to achieve results in complex areas.
Regalado quotes Rob High, vice president and chief technology officer of the newly expanded IBM unit on the decision to launch Watson when it did: ““You have to put this stuff into action and refine it. You need examples to work on. You do hit speed bumps, but I don’t think it was premature, it was exactly right.”
High also notes that other features will be added to address these issues. According to high: “We are adding the ability to listen and hear. It will be a Watson system that can hear, see, and talk.”
So what do you think?
There are compelling needs for tools such as Watson in the insurance industry. Big Data will require sophisticated tools to produce truly informed decisions. At the same time, insurers and brokers must optimize the time of their knowledge workers by utilizing tools (like Watson) to address simpler issues and make recommendations on others.
Our questions to you: How far can we expect Watson and its colleagues to go over the next 3, 5, or 10 years? What will this mean for the average insurer or broker?
Let us know what you think.
Note: IBM is a Platinum sponsor at the Insurance-Canada.ca Technology Conference 2014.
I think the emerging analytics capability of cognitive computing, which is what Watson does, will be disruptive as the internet has been and telematics now is. New use cases are emerging at the same time, for example, improving call centre effectiveness. The ability to do this depends on appropriate training of Watson – this means loading large volumes of information for the computer to learn. As cognitive computing matures, I’d expect it to become increasingly available and affordable – like a search engine today – and therefore, change how and where it’s deployed. What this will mean for the average insurer or broker? It will provide new opportunities to solve industry challenges of doing “more with less” by providing an alternative. However, just like you can’t hammer a nail with a screw driver – you can’t yet use Watson on any data to solve any challenge. And the industry still has an opportunity to leverage proven descriptive, predictive and prescriptive analytics software in ways it hasn’t yet fully exploited. Using analytics on new data from telematics, for example, is something multiple industries are beginning to explore, today.