he system looks at more than 100 elements of a news story

Then Zite picks stories that users might like, including those that overlap in categories, such as a story about ethics in sports, for someone who likes both philosophy and baseball.

That might sound remarkably like a “Push” technology that was so lauded in the late ’90s – including on Wired’s cover in an often-mocked article called “Push! Kiss your browser goodbye”, only to be sworn off by venture capitalists and technologists in the bust.

“Push is a dirty word in the Valley,” Johnson says. “It shouldn’t be. Push is actually a good thing.”

Zite’s fellow travelers in the customized iPad magazine category includes Flipboard, which pulls in content via the links sent out by your social circle, and Pulse, which pulls in stories from publications you follow. Neither attempt the deep-data mining problem of suggesting things for people to read.

So far, Zite doesn’t include any advertising, but Johnson says it’s not very hard to imagine how the app could do very good targeted ads — and do so without angering publishers (read revenue share).

And what of the recent hand-wringing over too much personalization, popularized by Eli Pariser’s new book The Filter Bubble, which argues that personalization threatens to divide us by forcing each of us only see what we want to see.