To maximize the business and societal benefits often associated with artificial intelligence, organizations must democratize AI, according to Ruchir Puri, chief architect at IBM Watson and an IBM Fellow. Puri spoke with SearchCIO at the recent Platform Strategy Summit hosted by the MIT Initiative on the Digital Economy.
In this video, he explained why making AI more accessible and cheaper -- or "more consumable" -- to a general audience will enable AI to live up to its hype. To prove his point, Puri drew on an example of a group of elementary and middle schoolers with basic programming knowledge who are using sophisticated AI technology -- IBM Watson, of course -- for a robotics project. "For me the notion of consumability for Watson and AI is, you don't have to have a PhD in data science," he said.
Puri also discussed IBM's initiative to scale AI on the Watson platform and how making the technology accessible to developers around the world via the cloud is as a step forward in democratizing AI.
The following transcript has been edited for brevity and clarity.
What are the business opportunities associated with AI?
Ruchir Puri: The biggest opportunities associated with AI are in various industry verticals like financial services, retail and healthcare, with healthcare being a prominent one -- and one that has a very large societal impact. But most importantly, the biggest business opportunity lies in democratizing AI, and that leads me to platforms. In the beginning, we at IBM were targeting certain industry sectors like healthcare, retail, consulting and other industries. We have now built a platform around Watson that allows the developer community to consume AI. And that's where democratizing AI comes into play.
If AI is to have the kind of impact it's hyped to have, then it is of utmost importance for AI to become more robust and more consumable. The consumability aspect of AI is extremely important. That aspect is built into the platform, a general purpose platform, which can be absorbed at a much lower price point than delivering software on premises.
For us, AI and the cloud are two sides of the same coin. They're intermixed because that allows us to deliver AI to a wide array of developers across the world in a very consumable way. I coach a team of fourth- to eighth-grade girls in robotics, and they've been working on a project with Watson. They know nothing more than simple Python. For me the notion of consumability for Watson and AI is, you don't have to have a PhD in data science.
The only way we can scale AI on a platform will be when it is consumable by a smart middle school student who doesn't have to know about AI per se. It's about making a couple of AI calls, knowing the application that you are targeting and being able to consume it and have an impact on the world around you.