Now that I'm older (27), it's time to start a real business. The kind that takes 5 years to build, delivers solid value to customers, makes two or three people

Now that I'm older (27), it's time to start a real business. The kind that takes 5 years to build, delivers solid value to customers, makes two or three people independently wealthy, and feeds twenty families. But no VC is ever going to fund that. I've come to believe that self-improving processes (such as economics and technology) exhibit a "natural growth rate", that this could even be applied to a person, and that for programmers it's about 20-40 percent per year. Compared to anything else, that's very high. We can reliably grow a skill base or a company at around that rate, give or take. Of course, starting a new business is risky, but there's not strong evidence that it ought to be riskier than starting a new job. Sure, half of small businesses fail within 5 years. But half of all jobs turn to crap in 5 years. At any rate, the mid-risk / mid-growth space that tends toward 20-40% growth and a non-zero but not obscene rate of failure is... as you've noted, severely overlooked. Bank loans require personal liability and are therefore only viable if there's nearly a zero chance of failure. VC builds disposable companies that have a 1/10 chance of returning 20x and a 1/100 chance of returning 250x. In the mid-risk / mid-growth space? Nothing. I think part of the problem is that such businesses-- even though they add more value in toto than anything else (they're where most open-source software comes from)-- are very difficult to sell. The issue is that VCs are optimizing for their careers rather than their portfolios. You can just as credibly build a portfolio out of mid-risk/mid-growth businesses as these get-rich-quick-or-die gambits. The problem is that you don't become a celebrity investor by building a great portfolio, but by having "been in on" the Facebook or Google that comes every few years.