In a keynote talk at a German electronics conference, Google's Eric Schmidt delivered his vision for the future of search.

"The next step of search is doing this automatically. When I walk down the street, I want my smartphone to be doing searches constantly - 'did you know?', 'did you know?', 'did you know?', 'did you know?'" Schmidt said. "This notion of autonomous search - to tell me things I didn't know but am probably interested in, is the next great stage - in my view - of search."

What's fascinating about this is that it's basically the opposite of search now. Search is about finding what you want, not about finding what you are statistically likely to want. I think there is a key difference between those two things.

This is what Google thinks I am statistically likely to search when I type in "religion is":


What we're thinking about is only hazily connected to any of the things that Google can know about me. If I'm sitting in an office in San Francisco around lunchtime, Google may think, "Yes, he probably wants lunchtime recommendations." Or if I type in "religion is," Google's algorithms may be able to make certain suggestions, but our thoughts are not that accessible to the web's spiders.

How could they possibly know when I'm thinking about Manfred Clynes and cyborgs? Or Russia's Kola borehole? Or the mythical luz bone in Hebrew scripture?

That's one reason search is so great: I get to "pull" whatever I want out of the Internet, no matter how old I am or where my house is or what I bought last on Amazon or what anyone else has searched for.

What Schmidt is describing is push-push-push. It sounds like narrowcasting, that awkward phrase for broadcasting in an era without mass audiences. Might that be useful? Absolutely. But I don't think it will ever replace or even be seen as similar to search. Except, that is, for how Google will sell it to advertisers.