Artificial intelligence isn't just for elite Manhattan hedge-fund managers. Amateur investors can fashion their own AI war room. AIQ Systems ($99 for monthly subscription and real-time data) sells expert system software for picking individual stocks. Tradetrek.com offers a "Live Stock Commentary Engine" ($39.95 a month), which tracks fundamental and technical characteristics of stocks and then "writes" an analysis. And Ward Systems sells neural-network software ($1,045) that can be programmed for investing. Real-time data from a provider like eSignal or myTrack might cost you an additional $100 a month.

So it's not hard to see how you can easily spend a couple grand a year on this stuff. But that price tag can eliminate one advantage of do-it-yourself investing vs. mutual funds. That $2,000 means you would need to invest at least $147,000 to achieve an expense ratio equal to the average mutual fund's, about 1.4 percent. If you invest half that, your cost of doing business would double--and that comes right out of your return. And don't think expenses are irrelevant compared with the huge profits you will be making. "We're into risk management, not 600 percent returns," says Jim Ackles, president of LBS Capital Management in Clearwater, Fla., which manages $50 million ($25,000 minimum) using AI programs.

Tools for change. Most AI tools available to individuals are really geared to either full-time day-traders or people who invest as a hobby. But Andrew Lo, director of MIT's Laboratory for Financial Engineering, predicts AI tools will soon veer away from such a hands-on approach. "It will be like radio," he says. "Who builds their own radios anymore other than hobbyists?"

For individuals, Lo foresees AI systems that would monitor your portfolio and quietly rebalance it to maintain a desired asset allocation. Or how about "SmartIndexes," which would take into account all your myriad financial goals and alert you if your portfolio returns deviated from them.