Q: You invented ways to put more computing power behind deep learning. Now you lead a lab in Silicon Valley for the Chinese search company Baidu. Why did it need a lab there?
A: They spin up new projects very fast. It’s partly driven by the dynamism in China—tech companies have to go quickly from having nothing to having state-of-the-art something. My lab’s mission is to create technology that will have an impact on at least 100 million people; it is intended to move rapidly, like a startup. We’re recruiting AI researchers and many people in Silicon Valley who have amazing skills from working on products and haven’t thought they could use that to make progress on artificial intelligence.
Q: What is the lab working on?
A: The first technology that we are focusing on is speech recognition. Touch screens on phones are fine for some things but really awful for others, and there are all kinds of other devices that are crying out for better interfaces. People don’t use speech today because it doesn’t work well enough. Our goal is to get it to a level where it’s as easy to talk to your devices as it is to talk to the person next to you. In December we hit our first milestone with DeepSpeech, a speech engine we built quickly from scratch using deep learning. When there’s a lot of background noise it’s dramatically better.
Q: Why would that have an impact on 100 million people?
A: In rapidly developing economies like in China, there are many people who will be connecting to the Internet for the first time through a mobile phone. Having a way to interact with a device or get the answer to a question as easily as talking to a person is even more powerful to them. I think of Baidu’s customers as having a greater need for artificial intelligence than myself.
—Tom Simonite
Watch this Innovator at EmTech 2015
Meet the Innovators Under 35