Ray Kurzweil: 'Users Should Be Passionately Involved in Every Stage of an Invention'
00:07 Ray Kurzweil: You're walking along, and on the one hand there's this great vistas ahead of you that you hope to climb, but there's also the ravine right next to you that you hope you won't fall into.
In 1974, inventor Ray Kurzweil developed computer software that could read text--but he wasn't sure what people would do with it.
00:26 Kurzweil: My field and my passion is pattern recognition, teaching computers to recognize patterns. So, I had a technique that could recognize printed letters in any type style. And it seemed like kind of a solution in search of a problem. I wondered what is this good for? I happened to sit next to a blind guy on a plane and said, "Wow, that would be great as a reading machine, but it would need other things. It would need a scanner so you could scan printed documents. It would need synthetic speech so that you could hear the output." Neither of these technologies existed. But I thought, okay, well, we'll develop those two technologies also and develop a reading machine.
Two years later, Ray held the first public demo of the Kurzweil Reading Machine.
He had done limited user testing.
01:14 Kurzweil: We worked to create this demo of being able to put a book on the machine and it would read it out loud. And people were excited about that. It was on the Walter Cronkite show. He used it to read his signature sign-off. But then we actually had people using it and realized it didn't really meet their needs because they couldn't use the user interface and it wasn't giving them feedback they needed to be able to turn the next page, and they didn't know how to put the pages on. And while it worked in principle, there were lots of little details. Well, how do I know what the key does since they can't see the name on the key, and lots of little issues.
01:52 Kurzweil: There's kind of a disconnect between... You have a successful demo, and this thing reads books out loud and that's a big breakthrough and everybody's excited, and then you give it to people to use and it doesn't really meet their needs, and there's something wrong with it. And there's a real disappointment there and you realize you really have to start over and craft this technology differently. I felt I can imagine what it's like to be blind. I can just close my eyes. But I actually learned that there was a lot to learn. Our imagination wasn't good enough. We really had to work with blind people who were experts on being blind.
In 1977, Ray worked with the National Federation of the Blind to test the product with people who can't see.
02:38 Kurzweil: In fact, the blind people we worked with, very quickly came up with strategies and things the machine should do. For example, there's a mode where each key will announce what it does. Simple solutions, but we hadn't though of that. So, I learned that lesson that you actually wanna work with the users, but not just in a beta testing step. The actual users of the technology should be passionately involved in every stage of an invention. So, I'm used to this now, this part of the process, you kind of develop something that shows the promise, but then you have to start over with the actual users and build something that will appeal and work for that marketplace.
To date, hundreds of thousands of blind people have used the Kurzweil Reading Machine.
In 1980, Ray sold his company, Kurzweil Computer Products, to Xerox for $6 million.