Scoble and Frank Seide

Scoble dropped by again, this time to see the new demos from Frank Seide, who is visiting from MSRA. Frank’s projects all try to turn speech into a first-class data type that can be manipulated the way you’d expect for data. With text, for example, you can easily apply formatting like capitalization or bold face, you can copy/paste, search, apply analysis like summarization, and on and on. What would it be like to do the same for speech? You can already see some of Frank’s results in the new version of OneNote, which includes a feature for searching audio files. But that's really just the beginning of the cool things we can do.

Robert filmed the whole thing, so hopefully it will be up on Channel 9 at some point. It's fun talking to Scoble because he sees so many other cool demos and hears about other interesting technologies, so we get a sense of how our speech stuff stacks up against the rest of the geek world. Apparently we do pretty well on the coolness scale, so that was good to hear.

By the way, I have no idea how Scoble does what he does. He gets hundreds of emails a day, is constantly appearing at conferences and demos. You might expect that he'd have a whole support team, but no, he does it all by himself.  That said, he did look awfully tired.