Last night I got a demo of Inkling, an online "prediction market." In this market, you invest fake money (all users start with a bankroll of a pretend $5,000) in verifiable ideas, such as, "Cars will have a U.S. box office opening weekend over $63 million," or "June 2006 will be the hottest on record." Why bother? Because you can win more fake money, which is good for your ego and which may also be redeemable at some point for something physical (such as a stuffed animal, perhaps).
This has been done before. There are several idea markets [see this good rundown on Mashable.com], including specialized ones such as the Hollywood Stock Exchange. It's a game, too, but marketers use it to gauge the potential success or failure of a project; if movie junkies are buying into the Cars stock, theoretically they represent the wisdom of the larger crowd of the moviegoing public, and the movie should do well. The full explanation of this theory can be found in James Surowiecki's The Wisdom of the Crowds.
But with play money, as others have written, you gamble poorly. You need some skin in the game for your fake stock purchases to matter, and to this end, Inkling is trying to sell its prediction market technology to corporations. The thinking here is that employees in a company are likely to be smarter, as a group, than any executive or committee. Will hemlines rise or fall? Which material will be less expensive to build our airliner out of, carbon fiber or aluminum? Which economy is going to expand faster: China or Europe? If employees were to put their reputations on the line by making calls on business issues like these, and then if they were recognized (with bonuses or raises) for accuracy, it would definitely encourage them to think critically about these topics. And it would give a corporation a decent view into the wisdom of its entire consciousness.
Other companies, such as Tacit, are in the business of mining the information stored in the enterprise by evaluating e-mail archives and the communication traffic between employees, then connecting them to each other. That's one way to leverage what people obviously know. Another is to make it worth their while by asking them to pick which questions they want to answer and rewarding them for being right.