4 ms·

In general the one regular report I hear from academics is that it was a waste of their time in the first place and that they would have been better off spendin

In general the one regular report I hear from academics is that it was a waste of their time in the first place and that they would have been better off spending those two-six years in industry anyway. Academia is a fucking scam.

epgui 3y ago
I could not disagree more vehemently.
etrautmann 3y ago
this really depends on personal situation, field, and institution. It always makes sense to think carefully about whether you want to do a PhD, but can certainly make sense for the right person in the right place.
- s1artibartfast 3y ago
  Exactly. If you want to study archeology, good luck finding a 500K salary in industry. Academia might be your only option, full stop. They're interesting part is that these things change with time. Academia might have been the best option for a fundamental AI researcher in the 1980s.
  - jcranmer 3y ago
    > Exactly. If you want to study archeology, good luck finding a 500K salary in industry. Academia might be your only option, full stop. I don't know what the salaries are, but there is actually a private archaeological industry which services the need in many countries to do archaeological excavations prior to construction.
reactordev 3y ago
As with everything, it depends. It depends on which school you went to, taught by which professor(s), and that had access to insiders within the broader academia landscape you are in. If you got an ML degree from University of Alabama, You have a vastly different experience than someone who got theirs from Harvard or MIT. It has its place, as with all things, but how many of the AI/ML grads with go work at Google or OpenAI? Not many. A degree does not guarantee you success, but in some cases, it is a barrier to entry (Physicians).
layer8 3y ago
Being constrained by demands of monetization potential isn’t necessarily better, and sometimes much worse.
- __loam 3y ago
  Yes it wouldn't be a hackernews thread without some neoliberal telling us about the harmonious society that the profit motive is creating.
zemo 3y ago
that's definitely what I hear from people who leave academia and go into industry, but that's not a representative sample. Are the people you're referring to as academics all people who left academia before you talked to them?
- __loam 3y ago
  I mean they left academia because it's ass.
- smallmancontrov 3y ago
  What fraction leave vs stay? I thought most left because the population of PIs was relatively constant, so on average one person per lab (not per year, per lab over the lifetime of the lab) "made it." In this case, an unbiased sample would include mostly people who left academia with a very occasional person who stayed.
  - speedgoose 3y ago
    It probably depends on the place, but I know academic institutions that want people to leave eventually because indeed not everyone should or can be a professor. It’s part of their mission to also train people for the industry.
jjtheblunt 3y ago
> Academia is a fucking scam. I think you're describing a system where stagnation is allowed momentum, because perpetual innovation of ideas and the people promoting them can be stymied by "tenure". Is this what you mean?
jedberg 3y ago
If you're getting a PhD with the intention of leaving for commercial work afterwards specifically because you want a high salary, then yes, the PhD program is probably a waste of time. Case in point, all my friends stayed to get PhDs and I left to go to industry. By the time they graduated, my income was already a lot higher than their starting salaries. However, people with PhDs have a much easier time getting promos at commercial places. They get a built in assumption of high competency, especially if the PhD is from a highly regarded institution. But if your intention with a PhD is to become an academic, and you want to spend your time teaching or doing research, besides being necessary it's also very worthwhile. Or if you just really enjoy learning and don't care so much about salary when you're done, then it's extremely fulfilling.
ajcp 3y ago
That logic doesn't check. How did they get their jobs in the first place if not for their academic credentials and research? Did they just apply with a blank resume? Maybe the skills they use at their jobs one can gain outside of academia for nothing, but that still won't get them hired. "Every rich person I've met has said money doesn't matter." Okay, but there's only one way for them to prove it...
- johnnyanmac 3y ago
  If your metric is compensation, which is the metric for many here, then acedemia isn't financially worth it for software engineers. You can check any "entry level" research position and compare it to a position that required 4-5 years of industry experience to compare. Unless you were researching a very hot or soon to be hot field (which requires a bit of foresight), you're generally not going to end up better off by spending those extra years studying such a hyper specific topic. as a CS/SWE, only go for post-grad if convenient (e.g. Some schools had a 4+1 program for a MS. That extra year probably does pay off), if you have a very specific topic you want to work in, or if you want to work in acedemia specifically.
  - ajcp 3y ago
    We're not talking about software engineers, we're talking about research scientist and engineers lured away from academia to work in the private sector. To your point I agree and don't know of any software engineer aspirant who thinks they need even a 4+1. But again, that's not the demographic we're discussing. There's a reason SV is hiring out of academia: because they can't hire the same minds out of a 4-year program, and certainly not off the street. In this case it shows the value of those programs that got them there, so for these former academics to report, as OP is suggesting, that it was a waste of time/money seems to hold no logic.
    - johnnyanmac 3y ago
      >There's a reason SV is hiring out of academia: because they can't hire the same minds out of a 4-year program, and certainly not off the stree I half agree. The degree holds some value (and for some schools they can even influence the curriculum to their needs), but value proposition comes more from a supply of people who 1) have sole semblance of fundamentals 2) an ability to see a long term study through (which is why every job, even many that don't need a degree, put up a Bachelors as a requirement when supply > demand) and most importantly, 3) aren't a training cost for the company itself. I'm sure well chosen apprentice could be trained by a company for a single year and would be comparably productive to any new grad, but no one really wants to train talent directly. With academia right there there is no need to, at best you have a set, 3 month internship to Gauge value. It's the economically best decision (arguably. If everyone wants to make offers of 200k to a top student instead of pay 80k for an apprenticeship, is it truly worth it?), but not necessarily the best from a company quality standpoint. But we know which of those 2 win out 9/10 times.

epgui 3y ago

I could not disagree more vehemently.

etrautmann 3y ago

this really depends on personal situation, field, and institution. It always makes sense to think carefully about whether you want to do a PhD, but can certainly make sense for the right person in the right place.

s1artibartfast 3y ago
Exactly. If you want to study archeology, good luck finding a 500K salary in industry. Academia might be your only option, full stop. They're interesting part is that these things change with time. Academia might have been the best option for a fundamental AI researcher in the 1980s.
- jcranmer 3y ago
  > Exactly. If you want to study archeology, good luck finding a 500K salary in industry. Academia might be your only option, full stop. I don't know what the salaries are, but there is actually a private archaeological industry which services the need in many countries to do archaeological excavations prior to construction.

reactordev 3y ago

As with everything, it depends. It depends on which school you went to, taught by which professor(s), and that had access to insiders within the broader academia landscape you are in. If you got an ML degree from University of Alabama, You have a vastly different experience than someone who got theirs from Harvard or MIT. It has its place, as with all things, but how many of the AI/ML grads with go work at Google or OpenAI? Not many. A degree does not guarantee you success, but in some cases, it is a barrier to entry (Physicians).

layer8 3y ago

Being constrained by demands of monetization potential isn’t necessarily better, and sometimes much worse.

__loam 3y ago
Yes it wouldn't be a hackernews thread without some neoliberal telling us about the harmonious society that the profit motive is creating.

zemo 3y ago

that's definitely what I hear from people who leave academia and go into industry, but that's not a representative sample. Are the people you're referring to as academics all people who left academia before you talked to them?

__loam 3y ago
I mean they left academia because it's ass.
smallmancontrov 3y ago
What fraction leave vs stay? I thought most left because the population of PIs was relatively constant, so on average one person per lab (not per year, per lab over the lifetime of the lab) "made it." In this case, an unbiased sample would include mostly people who left academia with a very occasional person who stayed.
- speedgoose 3y ago
  It probably depends on the place, but I know academic institutions that want people to leave eventually because indeed not everyone should or can be a professor. It’s part of their mission to also train people for the industry.

jjtheblunt 3y ago

> Academia is a fucking scam. I think you're describing a system where stagnation is allowed momentum, because perpetual innovation of ideas and the people promoting them can be stymied by "tenure". Is this what you mean?

jedberg 3y ago

If you're getting a PhD with the intention of leaving for commercial work afterwards specifically because you want a high salary, then yes, the PhD program is probably a waste of time. Case in point, all my friends stayed to get PhDs and I left to go to industry. By the time they graduated, my income was already a lot higher than their starting salaries. However, people with PhDs have a much easier time getting promos at commercial places. They get a built in assumption of high competency, especially if the PhD is from a highly regarded institution. But if your intention with a PhD is to become an academic, and you want to spend your time teaching or doing research, besides being necessary it's also very worthwhile. Or if you just really enjoy learning and don't care so much about salary when you're done, then it's extremely fulfilling.

ajcp 3y ago

That logic doesn't check. How did they get their jobs in the first place if not for their academic credentials and research? Did they just apply with a blank resume? Maybe the skills they use at their jobs one can gain outside of academia for nothing, but that still won't get them hired. "Every rich person I've met has said money doesn't matter." Okay, but there's only one way for them to prove it...

johnnyanmac 3y ago
If your metric is compensation, which is the metric for many here, then acedemia isn't financially worth it for software engineers. You can check any "entry level" research position and compare it to a position that required 4-5 years of industry experience to compare. Unless you were researching a very hot or soon to be hot field (which requires a bit of foresight), you're generally not going to end up better off by spending those extra years studying such a hyper specific topic. as a CS/SWE, only go for post-grad if convenient (e.g. Some schools had a 4+1 program for a MS. That extra year probably does pay off), if you have a very specific topic you want to work in, or if you want to work in acedemia specifically.
- ajcp 3y ago
  We're not talking about software engineers, we're talking about research scientist and engineers lured away from academia to work in the private sector. To your point I agree and don't know of any software engineer aspirant who thinks they need even a 4+1. But again, that's not the demographic we're discussing. There's a reason SV is hiring out of academia: because they can't hire the same minds out of a 4-year program, and certainly not off the street. In this case it shows the value of those programs that got them there, so for these former academics to report, as OP is suggesting, that it was a waste of time/money seems to hold no logic.
  - johnnyanmac 3y ago
    >There's a reason SV is hiring out of academia: because they can't hire the same minds out of a 4-year program, and certainly not off the stree I half agree. The degree holds some value (and for some schools they can even influence the curriculum to their needs), but value proposition comes more from a supply of people who 1) have sole semblance of fundamentals 2) an ability to see a long term study through (which is why every job, even many that don't need a degree, put up a Bachelors as a requirement when supply > demand) and most importantly, 3) aren't a training cost for the company itself. I'm sure well chosen apprentice could be trained by a company for a single year and would be comparably productive to any new grad, but no one really wants to train talent directly. With academia right there there is no need to, at best you have a set, 3 month internship to Gauge value. It's the economically best decision (arguably. If everyone wants to make offers of 200k to a top student instead of pay 80k for an apprenticeship, is it truly worth it?), but not necessarily the best from a company quality standpoint. But we know which of those 2 win out 9/10 times.