Is a PhD worth it?
In 2008, when I started my PhD, data science and machine learning were niche topics - not nearly as popular as today.
I worked on Monte Carlo simulations of differential equations to understand variability in biological systems at the Tidor Lab at MIT. The statistics and programming I had to do paved the way for my Machine Learning career. But was that worth it? Should anyone do a PhD if they don't want to be an academic?
The most relevant counterfactual is to do an MSc, and starting your professional career earlier. Here some reflections about a PhD:
It's financially worse in the short to medium term. An MSc is shorter, and gives you a good (perhaps better!) chance of landing a good job in the field of your choice. You start earning a salary earlier and can learn on the job.
It delays professional experience. Even if the finances balanced out, starting on the job market earlier, gives you professional experience in your early 20s that otherwise you'd only start in your late 20s.
You don't need a PhD to learn something well. If you want to become better at Machine Learning (for example), you don't need a PhD. You can have a job where you practice Machine Learning. Or you can make time for studying on the weekends. The online learning options are enormous and more personalised.
If you do go for a PhD some of the upsides may be:
A good PhD supervisor can be a unique mentor. Sure, you can have a professional mentor, but a PhD supervisor is typically an experienced person dedicated to making you better at your task. Prof. Bruce Tidor improved my focus, rigour and attention to detail and ability to ask "why?" repeatedly.
It is a great time to experiment in a topic. At no other stage in your life you'll have so much time to experiment as during the middle of your PhD. The frequent pace of university exams is behind you, and the short deadlines of business is a few years away.
It can be a unique personal experience. Student visas often allow you to travel and live elsewhere (like I did). If you can do something similar, that has value: it's fun, challenging, and life changing. Moving around after you're employed is not as easy as when you're a student, and has other implications (tax, work visas, etc).
It can teach you to become a better learner. Of course we can be good learners at any stage, but in a PhD you're expected to become one of the few world experts on something - largely on your own. Going though a PhD gives you some grit.
A PhD is a long commitment. We know it's hard to quit a PhD project midway, and most people stay for the duration (I admire those who changed course). Four or more years is a huge opportunity cost to pay.
If you embark on a PhD consider the alternatives, and beware not to do a PhD because you're expected to - I partially did this and I was lucky.
In any case, learning is a lifelong pursuit. A PhD is a unique way of doing it with many advantages, but there's room for all kinds of training in life. Comments are most welcome.