In '99 I left a relatively standard and easy 9 to 5 type job at Intel to join a startup. My assumption was that the startup would fail, because that's generally what startups do, but it gave me the opportunity to work on systems that I found interesting (specifically Linux, which was still kind of fringe back then) and learn from very smart co-workers. I spent more time at work, but I did so because I was enjoying myself and learning a lot. Personally, I'd much rather spend 12 hours doing something I love than 8 hours doing something I hate.
Just because you are aiming for the stars doesn't mean you can't still enjoy the journey. For me, it's more fun to fail at something audacious then succeed at something mundane. Of course others are free to feel differently, and I would never claim that this is the right decision for everyone.