Thursday, August 09, 2007

Is There Wisdom in Crowds?

Published: August 8, 2007

NEW YORK

'It's the wisdom of the masses that makes up our front page," says Kevin Rose. He is the founder of a Web news site called Digg that rates stories, videos and other content on the basis of how many people like them. Editors need not apply; Digg is proud to have none.

Of course, the "wisdom of the masses" produced a few 20th-century bummers, not least in Germany, Italy, the Soviet Union and China. Collective wisdom is often an oxymoron.

But the Internet has made the views of crowds more accessible - "crowdsourcing" is the geek jargon - while rendering them less dangerous. The cacophony can be overwhelming and undiscerning. Give me a serious food critic any day over the agglomerated diners' gibberish of the Zagat guides.

I am no doubt in a minority on that. Zagat has proved a global winner, as has American Idol. We live in an age when people love to know what everyone else thinks and the means exist to convey those thoughts instantaneously online.

Rose, whose background is in computer science rather than journalism, started Digg in 2004. It now has 18 million unique visitors a month. The idea behind the site is that it "surfaces the best stuff as voted on by our users."

By clicking to indicate you like or "digg" an item, you propel it toward page one. An algorithm edits. Citizen journalists rule. Rose, 30, tells me he wants "to give power back to the community."

The results are interesting. Among the "world" stories doing well of late - front pages exist for whatever category of news you choose - is an item about a New Zealand pizza chain using President George W. Bush's image for an advertisement.

Other strong performers include a chart comparing U.S. gas consumption to the rest of the world, how Paris Hilton lost her inheritance, and a piece about 2,500 French citizens forming the words "We will never forget" on Omaha beach last month.

"Funny," says a commentary on the latter, "whenever a French person does something anti-American, we hear about it." But when 2,500 French do something pro-American, it's only news on Digg.

A life spent in newspapering does not endear me to outfits that consign editors to the Paleolithic age. Digg needs content to be dug by its community; it provides none itself. Rather, it relies for "stuff" on my colleagues. Journalists once wrote stories. They now provide fodder for cannibalization.

Still, this is the future - communities jabbering rather than edicts falling. Barriers have crumbled, between states and between producer and consumer. That, on balance, is good for the world, if not necessarily for newspaper companies.

The collective wisdom of Digg produces an intriguing distillation. But it only begs the question: What makes stuff - a story, a book, whatever - a hit rather than a flop?

Until recently, Brent Stinski was a student at Cambridge University doing a doctoral thesis on the psychology of art and entertainment. In prosaic terms, he was looking into why Harry Potter boomed but "Superman Returns" was a bust.

That is an elusive quandary. Most entertainment products, be they books or movies, don't work. About 10 percent of what film companies or publishers or recording studios put out accounts for 90 percent of revenue. The big hit salvages myriad failures.

I once covered the publishing industry - or rather, the publishing lottery. Any price for a book could be justified if a suitably inflated hypothetical sales number was used for a prospective profit-and-loss statement.

Now, Stinski wants to do something about that using collective online wisdom in a slightly different way from Rose. His new Web site, called Media Predict, amounts to a virtual stock market for manuscripts, television pilots, rock bands and the like.

Traders with the equivalent of $5,000 in fantasy cash buy shares in the material they believe in. Whatever rises on this prediction market ought in theory to be the things entertainment moguls should buy and back.

"There's a crying need to link major media companies with the user-generated movement," Stinski says. "Going with your gut is inefficient for media and not very satisfying for viewers." Already, Touchstone, an imprint of Simon & Schuster, is working with Media Predict to select a manuscript for publication.

I feel queasy about the wisdom of the masses. But Digg and Media Predict give me pause. I may indeed have been hard on the French. More worrying, I have published three books and was grossly overpaid for all of them.

E-mail: rocohen@iht.com