I got my Ph.D. in CS, dabbled in start-ups (once unsuccessfully, once with marginal success) and then returned to academia as a professor. I’ll contrast these two lifestyles as best I can.
The Ph.D. student experience follows a predictable path:
Phase I is when new Ph.D. students get excited about taking advanced classes, digging their way toward the edge of human knowledge.
Phase II is when students reach the boundary of human knowledge and begin to realize that to escape with a Ph.D., they must meaningfully and significantly extend that boundary. All the advice they receive boils down to, “Just do it.” The path forward is unclear and uncertain. Doubt creeps in. As students flail around, depression begins.
Phase III is when the light bulb goes off. After enough time toiling at the boundary, a moment of insight happens. Suddenly, the path forward is clear. Often times, many paths forward reveal themselves all at once. Once this moment happens, a Ph.D. student never solves problems the same way again. Learning this unteachable style of problem-solving is what Ph.D. school is all about.
Phase IV is when students near graduation and look for academic jobs. Suddenly, they realize that there is one tenure-track professorship for every ten graduating Ph.D. students. Nine in ten leave academia, depressed and bitter, for industry. One in ten restarts, excitedly, at the bottom of the heap as a pre-tenure assistant professor.
As a pre-tenure professor, I’m learning a new unteachable skill: how to get government grants to fund my research program. This consumes close to 60% of my time, and after two years and 15 proposals, I’ve had no success. I’d be depressed if this weren’t a routine scenario these days. I try to make the proposal-writing process fun by treating it like writing plausible science fiction.
As one of my colleagues put it, the benefit to being a professor is that “you get to choose which 90 hours a week you work.” I do set my own schedule. I do set my own research agenda. I often work from home until 11am. I teach three hours a week, and I’ve even come to enjoy it. So, even with the work load, I’m happy because I love what I do.
I think some of the same benefits hold for start-ups. The only difference was that I worked 120 hours a week for start-ups. Again, I loved what I did, so I was happy. I used the problem-solving skill I learned as a Ph.D. student to solve big engineering problems for the start-ups, even though the kinds of problems I was solving were different. In start-ups, a successful exit is roughly the equivalent of tenure.
In the end, I think the difference between start-ups and academia boils down to risk: if you get a tenure-track faculty slot, you will probably get tenure in 6-7 years, and you’ll have a stable, fun job and good pay for life. In contrast, most start-ups fail, but for the few that hit it big, the payoff is outstanding.