Paul Graham is at it again; this time he is defending the practice of funding startups on the cheap. The usual amount of investment he does (through Y Combinator is about US $2000/person-month, which is (roughly) the amount of money an MIT CS grad student gets in a stipend.

He has an interesting, yet controversial take on funding. If the company fails for some reason, it's not that big a deal, because not that much money was invested. Also, being bought by a larger company is not an unreasonable strategy (as opposed to growing the company).

Some people have a problem with this, claiming that no one who's really serious about forming a company should take a few thousand US$/person-month in exchange for a percentage of the company. The founders should be able to scrape up the money themselves; that is a sign that they're committed to the company. While I can understand this sentiment, there are a lot of kids coming out of school with a lot of debt, so perhaps they don't have a few thousand US$ to start a company, although they have ideas that might catch on.

I did some rough calculations and determined that starting up a three-person company for three months would cost me the amount I currently spend for four years of dance or piano lessons.