Artificial Intelligence (AI) is clearly making inroads in all facets of our society—how we work, live, play, work and plan for the future. Yet it’s reaching a critical juncture, splitting into two paths with very different goals and possibilities. The question is, which one is most likely to dominate in the near future?
The most recognized type of AI, which is grabbing headlines and capturing the attention—and often concerns—of luminaries like Elon Musk and Stephen Hawking—is what has been dubbed “pure AI” or “open AI.” The goal of this approach is to create a robot that is as a human-like as possible—a system that looks, speaks and even reacts like people. You can find it often in the consumer realm, engaging with people.
Siri, Cortana, and other devices are early examples of this AI approach. Hanson Robotics developed a sophisticated example of this with Sophia, which was interviewed by Charlie Rose of CBS’s 60 Minutes. Using algorithms created by data scientists, she can answer spontaneously and refine her future responses as she learns. Sophia’s creator, David Hanson, is hoping that robots like her will be able to fill a social companionship role for the elderly, and thinks that robots will someday surpass human intelligence.
Another, more practical type of AI, pragmatic AI, is delivering critical value today. It takes an entirely different approach: using machine learning to handle discrete tasks and solve specific business problems by augmenting human capabilities. Data scientists “teach” these apps by supplying them with huge amounts of historical data and custom-designed algorithms. Deep learning apps are a type of AI that take it even further, with the ability to teach themselves.
With pragmatic AI, for example, we can help insurers automate the clearance of auto policies, or call centers to interact with customers and answer their basic questions. It can help predict defects in medical products before they go to market, alert commuters to anticipated bus wait times in real time, and identify the latest cancer treatments and research that best meet a patient’s needs, and much more.
While the idea of robots that can become your personal companions is fascinating (and already taking hold to some extent), there are still many obstacles to widespread penetration. It may be many years before pure AI is realized, and it’s uncertain what form it will take.
On the other hand, pragmatic AI is solving critical business problems today, and adding value and aiding humans in all types of industries, from health care to financial services and government. It is providing greater productivity, faster decision making, better customer service, and successful business outcomes.
And where does pragmatic AI goes from here? The possibilities are endless.