I remember reading a Steve Pavlina article a few years back about income. His claim was that it's better to diversify your income in the sense of decentralization. That's why a salary could be a lot more risky than self-employment. If you set up a bunch of automated systems that generate you a $100/month, then a single system failure will only put you out of that $100; on the other hand, if you put all your eggs in one basket, you bankrupt if anything happens. The idea is that once you set up enough of these systems, you can essentially spend only a few hours each week or month tuning them, but otherwise sit back and let them generate income in the background.
Pavlina's example was his game website. He created a site (company) that sold games, and once he optimized his site to be easily found on Google and automatically process transactions by letting users download the game after payment, he could essentially put it on cruise control and let the income flow (and all the games would themselves be little "systems"--if one game was made obsolete by a much better competitor, he would not lose all his income).
I am curious as to examples of such income generating systems. Several blatantly obvious ones come to mind: ecommerce (what Pavlina did), writing a book/song/etc. and collecting the royalties, taking a tiny cut when providing a service (e.g., ATM fees, taxes, etc.). You might be using several of those right now, and that's great. But additionally, I've been hard pressed to find more creative examples that deviate from these formulas. Perhaps it is because they work so well (after all, write code once and sell it a thousand times is pretty damningly attractive), but perhaps there are also many others, simply unpopular due to poor implementation or too obscure for a guy like me to notice.
An easily envisioned future store would completely automate this process with automated transportation (driving AI?) which took the products from the vendors and delivered it to stores, automated day-to-day operations (theft control systems, automated check out counters, etc.), and require only a small amount of tuning (a couple of "real" people in each store to accomodate any edge cases, like an unsatisfied customer who wants to complain to a live person). Such a store would require minimal management (occasional technical fixes and a couple of human managers), but each sale would take a tiny cut of profit for the company. The first company to successfully (without "bugs") engineer such a system would undoubtedly become incredibly rich.