SpokenLayer Aims to Give the Web the Power of Speech

I remember when I discovered "Read It Later" (now called Pocket). This is great, I thought. It's like a DVR for reading. Unfortunately, like a DVR, the Read It Later queue can become a dead letter office for good intentions. But what if instead of tagging material for later reading, you could tag it for later listening? Instead of having to set aside time to read an article in the future, you could layer it over another activity, like the time spent commuting to work or walking the dog? That may be possible soon if the concept underlying a startup called SpokenLayer catches on. SpokenLayer's premise is simple but ambitious: make Web content as friendly to the ears as it is to the eyes. SpokenLayer plans to turn web content into speech by human beings. While some people don't mind listening to synthetic speech for long periods -- say, more than a minute -- many more people do. That's not to say SpokenLayer is turning its back on synthtalk entirely. For content with a pressing time element involved, it says it may produce a synth version for immediacy and replace it later with a human voice edition. The startup is currently using professional talent and the authors of articles for its audio content. The obvious drawback to that approach is it doesn't scale very well. SpokenLayer has an solution to that problem, however. It's preparing a self-service platform to supplement its contingent of professional readers. It even has a section of its website dedicated to giving would-be contributors tips and tricks for creating audition recordings for the service.

As a first step in its strategy to make the Web more aural, SpokenLayer launched an iPhone app last week. The app contains audio articles from publishers who have agreed to join SpokenLayer on its maiden voyage. They include the Associated Press, The Atlantic, the National Journal, TechCrunch, Engadget, TUAW (The Unofficlal Apple Web Site), and Joystiq. In the future, SpokenLayer plans to tag its publishers' written content that has an audio analog with a "listen" button. Although it doesn't mention a "listen later" option, it would be wise to somehow connect the button with the app. SpokenLayer faces some obvious challenges in the near future. For example, how will it manage the labor intensive task of creation and quality control for the vast amounts of audio it will be dealing with? Rights issues may also be an issue. Someday, synthetic speech will be able to emulate our own voices. Better yet, it will be able to emulate famous voices. Think of what it would be like to have the voice of Winston Churchill read A History of the English-Speaking Peoples to you? Or to have James Earl Jones read something you've written? While those possibilities remain in the distant horizon, SpokenLayer, by adding an aural layer to the Web, could be a baby step toward that future.