I left my PhD program in CS (specifically, compilers).
I began working after graduation with a job at a (very) small Chicago start-up. I was the first engineer. I enjoyed a lot of freedom, and I was paid OK. The job had many ups and downs. The opportunity arose to pursue a PhD when my advisor-to-be recruited me at a weeklong PL research retreat over the summer. I thought: "well, I will regret it if I don't at least try and see." Not to mention: how could I have a better opportunity, than for my advisor to recruit me?
I was part-time for a little while. (My advisor was very flexible.) Work got worse, mostly for unrelated reasons, and it was time to quit the job. I was spending 3 days a week on research for the PhD, and not enjoying it. However, I was very much enjoying my compilers research seminar course (read a paper and discuss every week). So, I quit the job. I figured: "maybe I am not enjoying research because I haven't been able to focus on it."
It turned out I really didn't enjoy the work I was doing. I wanted to do research, but this wasn't it.
My advisor and I discussed my dissatisfaction with the PhD at length. I felt like I came to the program with a lot of excitement for compilers research, but I was spending all my time on tedious engineering work (writing a `#pragma`-parsing extension in Clang). I wanted a promise that there was a plan to switch gears, to get creative, to do (what I believed) I had signed up to do: to write a totally new kind of compiler!
My advisor was flummoxed. "This is it," they said. I was confused, and a little angry.
It's a common story to start a PhD on the promise that you'll get to do a certain interesting piece of research, but spend 2 years (in some cases, more!) doing other work for your advisor (possibly due to funding). I knew a PhD candidate in this very position, who one day finally became so frustrated they said aloud: "I'm sick and f*cking tired of this! This is not what I signed up for." They had been there 2 years, and still weren't working on the research they began the PhD out of an interest in pursuing.
Finally, my advisor issued an ultimatum: decide whether to stay at least 2 years through quals, or quit.
I told my advisor I couldn't commit to 2 more years. Not when I was unhappy with the work, and when they were openly promising me more of the same.
My advisor was angry. They started to say some unflattering things: sometimes things like, "since you're an engineer, you should understand that the foundation work comes first." Other times, they were just openly rude, and non-constructive: "in my opinion, you are a bad presenter." They prided theirself on radical honesty, but failed to value empathy. Eventually, I just stopped listening to them.
I told my advisor I was going to explore my options and look for work. At first, they were encouraging: "I expect nothing less." In retrospect, I think they were confident at the time that I would stay. After it became clear I was planning to find a job instead of continuing in the program, however, my advisor began to make arguments to me like:
"Nothing in industry is interesting. Only people with PhDs get to do the interesting work."
"You will get paid more with the PhD."
"You will never become a self-directed learner if you quit."
All of these points are blatantly false. (The last one in particular was really weird to hear.) But, finally, I was hearing it said: I was hearing my advisor spread the fear. All along I picked up on the trepidation my peers in the research lab felt toward "industry work," which they all seemed to imagine as a form of cubicle hell in the bowels of Oracle's terrifying UFO office complex in California. My advisor finally gave a voice to that fear. They tried to spread that fear to me.
I am extremely glad I had already worked in industry. I knew my advisor was wrong. I think, however, that they believed their own arguments. They've never worked in industry, after all. And maybe they needed to believe it was worse, in order to justify the stress and overwork of academia.
For my part, I ended up with an excellent job in New York City working for better.com. I have no regrets. I still have a fascination with compilers, and I occasionally read papers for fun. I work on my own self-directed projects in my free time, which I enjoy far more than I enjoyed working for my advisor. I also get paid very well, and my work peers are some of the smartest and most talented people I've ever met - PhDs included. The best thing my advisor ever did for me was fire me. I got a raise (even over my first job); I got to move to New York City; I got to work with incredible new people (and lots of them); and I now know exactly what I'm missing out on, having left the program.
By the way: we're hiring at better.com. :)