Stephen Hawking Gets A Voice Upgrade
RACHEL MARTIN, HOST:
The sound of Stephen Hawking's voice is iconic.
(SOUNDBITE OF ARCHIVED RECORDING)
STEPHEN HAWKING: Where did we come from? How did the universe come into being?
MARTIN: The theoretical physicist has ALS, which has made it impossible for him to speak. Hawking has used the same computerized voice system to communicate for more than 20 years. But recently, Hawking unveiled a brand-new communication system. The technology was created especially for him. Horst Haussecker is the director of Intel's Computational Imaging Lab. He worked closely with Stephen Hawking on this project. Mr. Haussecker joins us now. Thanks so much for being with us.
HORST HAUSSECKER: Hi, how are you? It's a pleasure to be with here.
MARTIN: Doing well, thanks. So the big question, I suppose, is Stephen Hawking's voice going to sound different?
HAUSSECKER: No, absolutely not. In fact, this was one of the things he did not want to have changed because his voice has become so iconic that he considers that his own personal voice. And in fact, there is only one system in the world that can generate that voice, which is the very computer board that Professor Hawking is still using to date. It's based on, you know, slightly outdated technology, but it makes it very unique and you couldn't even copy it if you wanted to.
MARTIN: So what was wrong with the old version of his voice? What wasn't working?
HAUSSECKER: It was a very slow process for him to do anything on the computer. In the early days when he was still able to use his fingers, there was a hand-switch that he clicked. And then later as his disease progressed, he had to switch to a simple optical sensor that's mounted to his temple that he activates by twitching the cheek of his face. Over the years as his disease progressed, it got harder and harder for him to do this very precisely. And he had to, you know, eventually difficulty to type even one word per minute. And so what we did is we allowed him to switch over to a more modern technology by creating a system that handles, you know, typing really well - something that would even go as far as predicting words that he most likely will use.
MARTIN: So the technology is so sophisticated that it could learn his language, his cadence so if he types the letter P, based on the context the computer system would be able to predict perhaps what word he wanted to write.
HAUSSECKER: That is correct. And then also, once he has typed one word, it would predict the most likely, you know, consecutive word. And just to give you an example, for Stephen, it meant that if he types the word the, the most likely following word would be universe.
HAUSSECKER: Which is of course, very unique to him and I doubt it would work for anyone else.
MARTIN: (Laughter). He gave you and your team unprecedented access to his life when you were trying to put this program together. What was he like to work with?
HAUSSECKER: He's a very funny person. And he was very easy to work with because he got excited working with us. We went in and we told him that we would like to treat him like a scientific experiment and observe him and study him and then build a working solution. And of course, being a scientist himself he really liked the idea. And he would get excited.
MARTIN: He was OK with that?
HAUSSECKER: Oh absolutely. You know, he found that very funny and actually interesting and he spent more time than we ever expected.
MARTIN: A lot of time and effort and money was spent developing this system. Will it have broader applications, or is this something that just will work for Stephen Hawking?
HAUSSECKER: Absolutely. So it will have broader application. We are going to open up the tool books we build and we are offering that as an open-source library. And that would allow others to build solutions for people who have similar disabilities and to help them to, you know, increase their independence and quality of life.
MARTIN: Horst Haussecker is the director of Intel's Computational Imaging Lab. He worked with Stephen Hawking on his new voice technology. Thanks so much for talking with us about it.
HAUSSECKER: You're welcome. It was a pleasure to talk to you.
NPR transcripts are created on a rush deadline by an NPR contractor. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.