I recently interviewed for a job in Boston, in which I would get to work with Shae Erisson (shapr) and Edward Kmett, which would be a wonderful treat, of course. But I turned it down after it was described to me. My reasoning at the time was that it was too formal and there were not enough hard problems for my taste. But after a little reflecting, I realize that that’s not the reason at all:
I believe now that I am deeply and fundamentally an artist. At a job like that, I’m just an engineer building products for customers. I am not cut out for engineering as an industry.
I’m currently making a game with my friend Max about gravity and the universe, which we plan to enter into the independent games festival at GDC. In the process of making this game, we written and scrapped three large-scale gravitational simulation algorithms, and are developing our fourth. We have learned and are simulating a fair amount of cosmology. We even tried to use some general relativity to solve the simulation problems. Our gameplay will probably not be intuitive to the naive player, but that’s because the naive player doesn’t really understand gravity, and we believe that showing the true nature of the universe trumps pandering to ignorance to flatten the learning curve. Clearly, our hearts are in this: this isn’t just a game to take to the bank. We’re trying to show people something about the beauty of the universe.
Dana is also a work of art. That’s why I’m having trouble understanding what it is as a clearly-stated engineering goal. It is evading the usual engineering practices, because it’s not trying to solve a specific problem. Really what Dana is trying to be is advanced technology. I want it to do whatever it does, and when the code pop culture sees it they have the sense that they are looking at alien technology: it is compact, unfamiliar, but obviously very elegant, whatever it is. When you dig into its subproblems and put in the effort understand them, the response will be “Oh! Of course that’s how you do that!” The realization that Dana is art has convinced me to try to reduce it’s scope. It is more important that it do one thing in the most beautiful way, than for it to be a complete foundation for arbitrary software.
And I have always been an artist with my music. My friend Eric Shapiro bought a piano for me to play on the Pearl Street mall, in exchange for 25% of my earnings. It is going great: I often attract crowds of 20 or more watching me improvise and make about $40/hr doing it. Eric helped me see that my music can be valued. So I’ve decided to be more industrious about it — after all, the music industry is about supporting artists (even if it is mostly corrupt); in contrast, I don’t know anyone who would pay for something like Dana. I’m developing a solo album to sell on the mall and approach record companies with. I’m trying to think of a pseudonym (my name is pretty lame). During the winter I’ll make my money with teaching and shows. Essentially, my plan is curiously “I’ll use music to support myself, so I have time for my art.”
It’s harder to make good money with music than with engineering. But I think this plan will lead me to a happier lifestyle than wasting my half my life on an engineering project that I couldn’t give a rats ass about, just so I can have a surplus of money in the other half.