Mike_Kiev - Fotolia
IBM Watson holds the attention of many avid technology watchers these days, but a view has yet to emerge from the fog. Navigating the complex package of services and modules that now comprise Watson can be daunting.This episode of the Talking Data podcast searches for a clearer view on Watson and the cognitive computing technology it seeks to employ.
To that end, guests Nicole Laskowski, senior news writer for SearchCIO, and Ed Scannell, senior executive editor for the Virtualization and DataCenter Media Group within TechTarget, joined Talking Data podcast regulars Ed Burns and Jack Vaughan for a special edition.
Cognitive computing can be viewed as a way of changing the role of the computer so it becomes a partner to the business user, Laskowski said. A big part of that is natural language processing, which leads to new ways of interacting with systems. Such speech-driven systems are becoming more familiar, but, she said, it is important to distinguish the far-reaching Watson effort from more narrowly directed speech-savvy personal assistant technology, such as in Apple's Siri.
Ed Burns, who now serves as editor for SearchBusinessAnalytics, said he first began to cover Watson when he covered the health IT beat, and he has seen changes since then. The initial excitement that Watson generated in that critical segment has somewhat abated, he said. But he suggested that IBM may find easier tracking, as it appears to have scaled back from some of the first ambitious Watson showcases, to pursue improved versions of familiar enterprise applications. That effort means going after call center tasks and travel-planning improvements, Burns said.
For his part, Ed Scannell said IBM Watson will find increasing competition, especially in machine learning, a subset of cognitive computing that is of interest to Google, Microsoft and others. IBM's future approach for Watson is more technically sophisticated and could prove better in the long run, he said. But, he added, faster-moving, less ambitious approaches to business problems often win out.
Finally, Vaughan, editor for SearchDataManagement, spoke with Bri Connelly, a University of Texas at Austin student, Watson user, and member of a team that recently won an IBM-sponsored competition that highlighted Watson application building.
Connelly said that since its first heady days as the winner on the Jeopardy! quiz show, IBM has turned Watson into a collection of readily programmable cloud-based services. Watson is not difficult to train, she said, using the parlance of cognitive computing.
Read about Watson's roots in AI
Learn how a U of Texas team is training Watson to help the United Way
Hear how IBM is opening up cloud APIs to Watson in a podcast