With the inclusion of the personal assistant, Siri, in the newly launched iPhone 4S, Artificial Intelligence (AI) has captured the interest of smartphone users. However, this latest iteration of Siri is just one example of what’s possible using today’s technologies. From the navigation system in Google’s autonomous car, the flight assistants in planes, and Canon’s stablized camera lenses, to EKGs and facial recognition systems, AI has been creeping into our culture for some time. It’s just now that these systems are at a point where the general public has begun consciously interacting with them via natural user interfaces—motion, gestures and sounds—rather than through the languages computers have traditionally understood.
Another development essential to the greater adoption of AI has been processing power: for example, IBM’s Watson required a huge amount of data, 90 servers with 2280 processors, and 16 terabytes of RAM to win at Jeopardy! Google’s self-driving car used a similarly large amount of resources in order for it to navigate streets while being mindful of pedestrians and bicyclists. But by storing needed data in the cloud, the more modest devices we carry in our pockets are able to take advantage of AI. As a result, we’re poised to see these technologies permeate a large number of industries.
As with Siri, one application of AI is the “personal assistant.” Eric Horvitz, a scientist at Microsoft Research, has built a digital assistant over the course of eight years, programming to to observe his daily work routines. Now with the understanding it’s gleaned, the assistant is able to interact with him and his coworkers, giving the appropriate responses to various workplace events: it can sense when he can’t be interrupted or when he’s running late, adjusting his schedule and interactions with his coworkers appropriately. In the near future, similar AIs will be able tailor their interactions with people based on their perceived emotional states as understood by users’ tone of voice or spacing of words.
In fact, in the future, if AIs are working properly, we won’t even recognize them. The goal of ubiquitous computing, of which AI is a component, is having information processing integrated into everyday objects and activities. For instance, a domestic system might recognize an individual by their voice or gait, and automatically adjust the room’s lighting and temperature. These developments will not only help make consumers’ lives easier, but also provide more specific, detailed information to brands about consumer behavior. By monitoring and analyzing individual habits, brands will be able to customize offers and reminders to potential customers at the appropriate moment and location. While this will undoubtedly raise privacy concerns, the potential for brands should not be ignored, and we expect that the most forward-thinking companies will begin to capitalize on this technology in the very near future.