Can relate to all the points mentioned in the article. Had to also make the difficult decision to come out of a pure mathematics PhD and enter industry and that was pretty hard. Just was able to muster enough courage to step out of comfort zone and unskill at the right time of PhD. See many fellow and bright colleagues still languishing the perpetual postdoc cycle with no light at the end of the tunnel, simply because there's not enough jobs in academia and surprisingly its been like this for the couple decades
I have a formula for grad students in the humanities. Tl;Dr if there are more PhD students at an institution than there are tenured faculty members, then it's just a diploma mill and you should run away.
I agree with almost all of your points, but not the final one, because no LLM is currently anywhere near able to produce PhD-level research (whether they can even perform anything that can legitimately be called "research" is questionable, imo). Plenty of people who work on AI will tell you that there's a good chance it will never be capable of the kind of sustained, reflexive, original thought (not to mention the empirical and experimental work that PhDs depend on in many disciplines) that a PhD (ostensibly) trains you to perform. So I'm not worried about AI eliminating the need to conduct PhD-level research. But that might be a moot point in the face of all the other considerations! Being able to conduct advanced research doesn't help anyone if you can't get a job.
I have always regarded PhD as being equivalent to getting a drivers license to do research. In that sense it’s not surprising that the theses are not cited. But you are right that such a cruel system cannot persist
I too can resonate with the personal fulfillment/development argument. Theres some beauty in getting deep into something without caring about the RoI. Something I struggle with in the industry is a lot of time I find myself stressed to meet deadlines/implement decisions made by others and theres a decreased level of agency compared to doing self-driven research. Of course follow the money, but to what extent? I feel like I lose a bit of my soul selling vaporware for snakeoil salesmen. Whats your approach to find meaningful work outside of academia?
This is an excellent article and should be compulsory food-for-thought for anyone considering doing a PhD. Yes: it's still an excellent way to engage deeply with a question that one finds extremely important... and also yes there is no doubt that doctoral students are low-key taken advantage of as underpaid labor conveniently ignorant of the job market dynamics. Many faculty members aren't even really aware of how few jobs there are.
This said, I do find that the skills I learned there are useful in my work as a content creator, coach, and educator (outside of academia) today, but I have a very non-traditional career path that most people would not be uncomfortable with.
I am curious about the studies you're referring to regarding "transfer learning". It's seem difficult to measure the effect of this "treatment" (to use stats lingo), so I am not surprised that the evidence is inconclusive.
I think personal fulfilment (or not quite that, but something related) is actually a pretty reasonable argument to fund science and humanities. I can see how a more libertarian person would disagree.
Can relate to all the points mentioned in the article. Had to also make the difficult decision to come out of a pure mathematics PhD and enter industry and that was pretty hard. Just was able to muster enough courage to step out of comfort zone and unskill at the right time of PhD. See many fellow and bright colleagues still languishing the perpetual postdoc cycle with no light at the end of the tunnel, simply because there's not enough jobs in academia and surprisingly its been like this for the couple decades
I have a formula for grad students in the humanities. Tl;Dr if there are more PhD students at an institution than there are tenured faculty members, then it's just a diploma mill and you should run away.
I agree with almost all of your points, but not the final one, because no LLM is currently anywhere near able to produce PhD-level research (whether they can even perform anything that can legitimately be called "research" is questionable, imo). Plenty of people who work on AI will tell you that there's a good chance it will never be capable of the kind of sustained, reflexive, original thought (not to mention the empirical and experimental work that PhDs depend on in many disciplines) that a PhD (ostensibly) trains you to perform. So I'm not worried about AI eliminating the need to conduct PhD-level research. But that might be a moot point in the face of all the other considerations! Being able to conduct advanced research doesn't help anyone if you can't get a job.
I have always regarded PhD as being equivalent to getting a drivers license to do research. In that sense it’s not surprising that the theses are not cited. But you are right that such a cruel system cannot persist
I too can resonate with the personal fulfillment/development argument. Theres some beauty in getting deep into something without caring about the RoI. Something I struggle with in the industry is a lot of time I find myself stressed to meet deadlines/implement decisions made by others and theres a decreased level of agency compared to doing self-driven research. Of course follow the money, but to what extent? I feel like I lose a bit of my soul selling vaporware for snakeoil salesmen. Whats your approach to find meaningful work outside of academia?
This is an excellent article and should be compulsory food-for-thought for anyone considering doing a PhD. Yes: it's still an excellent way to engage deeply with a question that one finds extremely important... and also yes there is no doubt that doctoral students are low-key taken advantage of as underpaid labor conveniently ignorant of the job market dynamics. Many faculty members aren't even really aware of how few jobs there are.
This said, I do find that the skills I learned there are useful in my work as a content creator, coach, and educator (outside of academia) today, but I have a very non-traditional career path that most people would not be uncomfortable with.
I am curious about the studies you're referring to regarding "transfer learning". It's seem difficult to measure the effect of this "treatment" (to use stats lingo), so I am not surprised that the evidence is inconclusive.
I think personal fulfilment (or not quite that, but something related) is actually a pretty reasonable argument to fund science and humanities. I can see how a more libertarian person would disagree.