IBM is apparently envisioning a future version of its Watson knowledge engine to become available on what we would describe today as future superphones. Like Siri, Watson would work via speech recognition and send requests back to a data center to find an answer. IBM said in a Bloomberg interview that image and video recognition could aid speech recognition for more detailed answers as well as wider application fields.
The goal of Watson 2.0 is to reduce the power consumption needed by the engine. “The power it takes to make Watson work is dropping down like a stone,” Bernie Meyerson, IBM’s vice president of innovation, said in an interview with Bloomberg. “One day, you will have ready access to an incredible engine with a world knowledge base.”
This incredible engine may be able to tap analytics and services we cannot quite imagine today. For example, your phone may be able to provide basic physician services - which could be enabled by combining speech, video and image recognition features, while having access to medical records to an individual as well.
Today, Watson is still very much a large mainframe application and IBM says that the efficiency levels are not quite there for a large-scale, mainstream service. IBM did not say when Watson will be efficient enough to be feasible to be run on a smartphone.