June 17, 2014
Devesh Raj, Partner at Boston Consulting Group and Solomon Assefa, Director at IBM Research invited a group of Davos Young Global Leaders to learn about its Watson technology. Named after Thomas J. Watson, the founder of IBM, Watson first premiered on Jeopardy in February 2011. It's considered a groundbreaking advancement in computing and represents a new era of cognitive computing.
What is cognitive computing?
We have historically programmed computers with very specific instructions to perform a limited set of tasks. Watson improves on current paradigms of computing by interacting in natural language, learning from data, and uncovering insights to complex questions by rapidly analyzing very large amounts of unstructured data. Watson is possibly the first and currently the most advanced in an era of computing systems that will have cognitive learning capabilities.
Dr. Dario Gil, Director of the IBM Cognitive Enterprise Lab, began his talk with the history of computing. He indicated that we are moving from an era of automation shaped largely by tabulating systems and programming systems to an era of learning or cognitive systems.
Watson processes information by understanding natural language, generating hypotheses based on evidence and learning as it goes. The technology functions more like a human than a computer. "Watson makes brains bigger and more effective," says Dr. Gil.
Several people expressed concern that Watson would eliminate the role of humans in decision making. Dr. Gil explained that Watson was designed to enhance human cognition not to replace the human brain. For one, it takes a team of engineers to enter data into Watson. Also there are still things that the human brain does better, such as image processing. Cognitive technologies like Watson should be viewed as a tool not a replacement.
Post-Jeopardy, researchers at IBM are working on new solutions including a Watson for healthcare. The goal is to help health practitioners make more informed, better and accurate decisions. Watson will provide real-time context by presenting health practitioners with probabilistic outcomes based upon an overwhelming and often unstructured sets of data. Watson is presently too expensive for most use cases, however, researchers plan to continue developing real-world applications as the technology improves and costs decrease. NATALIA
Illustration by Sachin Teng