Technology Needs to Get a Lot SmarterBy Samuel Greengard | Posted 2016-06-17 Email Print
The reality is that we need devices and systems that can understand conditions, context and behavior at a far broader and deeper level than they can today.
I recently spent a few days talking to entrepreneurs at startup tech companies in France and attending the Connected Conference in Paris. There, I witnessed an amazing array of ideas and products, ranging from 3D printing systems designed to build housing for the economically disadvantaged to an entirely redefined concept of how to help the disabled achieve mobility.
It's clear that incredible technology innovation is taking place all over the world. Yet, while these niche solutions could revolutionize everything from housing to health care and beyond, the reality is that all IT systems need to get a lot smarter.
One of the biggest problems is that smartphones really aren't all that smart. Despite all the whiz-bang features and capabilities, they ring during meetings; they can't display a reservation when you step into a restaurant; you have to dig through emails to find your concert tickets and manually place them in Apple Wallet or Google Wallet; and speech recognition is still rather primitive.
For example, if you ask Siri to display your upcoming trips or vacations, you will receive the following response: "That may be beyond my abilities at the moment." If you ask Siri, "Who do I call most often?" it will reply, "I'd rather not say." If you ask how much money you have in your checking account, you will likely hear: "Who, me?" or "I can't answer that."
The reality is that we need devices and systems—and this certainly isn't limited to smartphones—that can understand conditions, context and behavior at a far broader and deeper level than they can achieve today. We need devices that can change the form of input and notification based on the immediate situation and needs.
It would be helpful if the phone would know when I'm driving or cooking, and, if my hands are not available, switch to biometrically triggered speech mode when a text or call comes in. It would be great if the phone would know when I'm running for a plane and automatically hold the call. It would also be extremely useful to have the phone understand the types of restaurants I like and then, when I'm in another city, make recommendations based on my previous behavior.
This same type of intelligence would redefine enterprise computing, cyber-security and many other areas. "We are still in the very early stages of deep learning and AI," says Rand Hindi, founder and CEO of deep learning firm Snips. "The end goal is to create technology that is far more practical and makes the interface disappear."