I've been reading a lot about startups lately, specifically in the essays of Paul Graham. There is a part of me that regrets not having ever tried to start my own startup - unfortunately the bursting of the bubble and the intricacies of US immigration policy meant that there was never really a practical time. While being filty rich does have a certain appeal, the more compelling reason is the one at the end of this essay - getting the business of getting enough money for life over with as quickly as possible in order that I can work on whatever I want to. The summer/autumn before I started at Microsoft, when I had several months of mostly uninterrupted time to myself to work on whatever I wanted to was one of the most intellectually satisfying (not to mention relaxing) periods of my life.
I have always thought of myself as having progressive politics but this is the best argument I have seen yet against taxing the rich. I guess one possible counterargument is that maybe taking very big risks isn't really necessary to develop new technologies (nor necessarily particularly beneficial to society as a whole) - the only reason that this has been true in the past is that developers and funders of new technologies have very little information about whether these new technologies are likely to be successful or not. With better information, perhaps we will be able to pick the winners before a lot of money has been spent, meaning that the big risks will be unnecessary.
One other thought: as startups become cheaper than ever to create and run, perhaps we will see more of them following the model of craigslist - staying small (in the financial sense) and private and concentrating on what users want instead of aiming for explosive growth, trying to become as profitable as possible and then either go public or get aquired. While the latter is more lucrative, I suspect the former might be more intellectually and spiritually satisfying.