Supply and demand. We produce more PhDs than needed by either academia or industry.
Hence, the burden of post-docs just grow (I was talking to a biology PhD who said 5-6 years of post-doc was typical, for chemistry I've seen people do 2).
Hence, the burden of post-docs just grow (I was talking to a biology PhD who said 5-6 years of post-doc was typical, for chemistry I've seen people do 2).
I don't think so - industry can't hire enough PhDs. In my field (programming languages) shortage of qualified people is a bit of a problem at the moment! We wish there were more being produced.
Yeah... hence why you need a PhD. 90% of domain knowledge in this field isn't even written down - you have to sit down with a wide range of people to learn from them - which you get to do on a PhD.
> Many major companies that maintain programming languages do not require PhDs in order to work on their compiler teams
They don't require it... but look at who's actually there and how many of them in practice have a PhD.
> if you are biased towards people with PhDs and preexisting experience for junior hires
As I've said elsewhere I've taken on people with no high-school diploma and zero experience in the past - I'm not personally gatekeeping.
Like I get the schooling you do prior is supposed to allow you to hyperfocus something, but at the same time, I've seen jobs asking for postdocs with experience in highly niche areas, often cross disciplinary. One I saw was looking for someone with encyclopedic knowledge of some crazy domain specific algorithm at the intersection of two sciences.
Wild guess, but I if I were to bet, I'd bet most people with a PhD or even professional experience in the filed would not have heard of this algorithm.
I guess I've seen similar software jobs, but at least with software I can usually afford the equipment to learn things myself.
I think both are true. Academia is producing too many PhDs in fields that industry isn't interested in and too few with a background that industry wants/needs.
I've worked with people without even a high-school diploma in my compilers team, so I'm not snobby, just realistic that to succeed you need things and that a PhD is a great way to get them.
Easy, you just have your ATS automatically reject any resumes that don't appear to have a PhD
I'm a (relatively early on) grad student at a chair that does PL and ML, and it's been pretty depressing to look up alumni or authors of papers more on the PL/FM side and seeing those that left academia at 'normal' jobs that seem unrelated to PL, whereas the vast majority of people more on the ML side work directly with it. I like PLT a lot, but especially if you're not at a top school it seems like quite a gamble.