thinkpad20

I'm currently taking the introductory computer science class on MIT's OpenCourseWare, which is a Python class. I'm about halfway through and I'm understanding the material just fine, at least, enough to know that I'd be getting a very good grade if I were actually sitting the course. For me the big issue is finding a good program that can put me on a PhD track - once I'm in said program, I'm not so worried about the material being too difficult, because I know I can work hard and learn it. How quickly I earn the degree is only relevant in a practical sense, as speed of learning will affect both how soon I can go to graduate school (in a funded PhD program), and perhaps more importantly, how much debt I accrue over the course of my study. I wouldn't particularly mind going to college for another 4 years, but my bank account would.

I'm 27 years old. I graduated with a physics degree about 5 years ago, but now I'm interested in a career in computer science. Specifically, I'm most interested in researching artificial intelligence and/or computer vision. I don't have any computer science classes in college or industry experience, so I'm starting basically from scratch, outside of general computer savvy. What's a good path for a person in my position to take? Living in the "real world" means that I have to worry about bills and getting old, two things that don't get along well with going back to college. But at the same time, even if I could get into a good master's program (which, maybe I couldn't) I'd be probably far behind my classmates at least initially, and even with such a degree I'm not sure I could get into a PhD program on that alone, which is what I'd really like to do. What would people suggest? Would a second trip to college be a feasible option? Might I be able to get it done fairly quickly with crossover credits, and possibly even save money compared to a master's? There's lots of things to think about.