From what I saw in biomed only 1/10,000 or so will end up with the time, skills, and appropriately ambitious project to allow a decent job to get done. The rest will be forced to produce BS (literally, most of it is just misinformation at this point) or quit before even graduating.
On the other hand, spending some years pursuing a PhD can be a great time to build skills and introspect in life. I grew a lot in that time in a way I don't think I would have with a job.
Of course, if you have a very stressful advisor, it isn't worth it.
I think something like a basic income would allow the a certain type of person to achieve the same, and do it much faster and cheaper than getting a higher degree. I wonder if that is one reason why most of these problems started when higher education/research became less of an "elite" activity, ie as discussed in subnaughts link in this thread: https://news.ycombinator.com/item?id=12805406
as software dev, I would never, ever, ever want to be in academia compared to the jobs I went through. much more life to be lived, much more money to spend on life to live (ie travelling around the world), and CV/skillset that gets door opened for next jobs.
you can build proper introspect in life when trekking few weeks among 8000m peaks, diving in coral reefs or anything in between, just give it a bit more than 1-2 weeks.
great PhD can be interesting experience, but for me, it would always be subpar to great job
NB That's what I did after ~5 years in academic research at a UK university and never really regretted not finishing by PhD as I had no intention of going back into research.
One needs to have clear expectations going into the PhD about what they are trying to do and they need to have clear communication with their advisor.
The vast majority of PhDs in science and engineering do very well and find jobs, but most don't get professorships at research focused institutions. Having realistic expectations is key, but a career in research is great.
A PhD is stressful and takes 5-6 years, but it can be a great experience with the correct expectations and a solid plan. The main problem is that most PhD students don't seem to think about those things nearly enough and have unreasonable expectations.
Well, here is how I was taught to do research:
1) You come up with a vague idea that one thing should be
positively/negatively correlated with another thing.
2) You measure some proxies for those two things that have only
been half-validated, if at all.
3) You make a bunch of dubious statistical assumptions and check
whether there was *exactly zero* correlation between the two proxies.
4a) Your sample size was too small relative to the noisiness of
the measurements and p>0.05. So you ignore the result.
4b) Your sample size was large enough, or you got lucky and p<0.05.
So you publish saying these results support your theory or are
counter-intuitive, depending on the direction of the correlation.
Honestly, I think the main problem is mostly that PhD students are, on the one hand, full-grown adults supporting themselves, and yet are, on the other hand, paid McDonalds wages to do the most intensive intellectual work our society has for Goldman Sachs hours.
Edit: I see Cyph0n has beaten me by a few minutes. Oh well.
I didn't find the PhD to be too stressful, from the jobs I left to enter education I suppose I was well adapted to stress management. That said I think it's a mix of adviser and individual's personality. I despise the work force (that is an understatement) and love learning. Part of the reason why I browse and lurk on HN as I love coming across stuff that I'm not at all oriented with. I don't think I'll become a programmer or anything..it's just I love learning even things I barely understand (e.g. Quantum anything!)