I want to get a general idea on how in-demand various CS subfields are (going to be in 6 years), so that I pursue my phd research in an employable area.

How do I go about doing that?

  • 8
    By the time you graduate, they will be using something else. – Solar Mike Dec 28 '19 at 19:20
  • @SolarMike True, but some general trends are discernible, no? E.g., reinforcement learning seems to be a much better choice than bioinformatics. – HappyFace Dec 28 '19 at 19:22
  • 2
    Who knows there wont be some bioinformatics breakthrough in 6 years while reinforcement learning fades into legacy applications and niche research due to its limitations beyond gamified problems? Goal of the PhD is the research itself, why not use those years gaining industry experience instead? – Victor S Dec 28 '19 at 20:09
  • 3
    You need to have informational interviews with current PhDs and other industry experts at the top of their field. For instance, if I were you, I would try to speak with someone currently researching quantum computing, or neural networks, or bioinformatics, or nanotechnology, or someone in general hardware, etc. Those experts tend to know what's coming ahead. That being said, 6 years is a freaking long time! Be sure to enroll in a PhD program that can convert itself to a Masters just in case you want to get off the train before it gets to its final destination. – Stephan Branczyk Dec 28 '19 at 21:39
  • 3
    Unless you plan to work in a university or research, I believe that all a PhD in computer science is good for is showing off with a "Dr" in front of your name, and maybe getting a slightly higher salary. So I'd do the PhD in a field you are actually interested in. – ThiefMaster Dec 29 '19 at 21:02

When I was in school, the thought of a man-portable device that can communicate with any other human on the planet was in the same league as plasma beam weapons, space dreadnoughts, mind-reading aliens and "The Force". 25 years later and it's a 100$ device kids get for Christmas.

Moral of the story? Don't try to predict the future. It's not worth the time. It will be different anyway. Do what you like best now and adapt to whatever happens.

  • I have trouble identifying what subfields are in demand even now ... – HappyFace Dec 28 '19 at 19:50
  • Whatever is in demand is whatever you are comfortable with. Top demand areas tend to become hyper-competitive beyond their worth since everyone swarms them. Did you try identifying something that attracts your interest in your own professional network? – Victor S Dec 28 '19 at 20:13
  • 1
    smart phones were far from science-fiction 20 years ago. The internet was in full swing and portable devices had been around for 20+ years already. Smart phones did nothing new, they just provided it in a new format. Cell phones had already existed for decades. The internet and portable devices already existed for a long time at that point. If you were even moderately interested in CS 20 years ago then smart phones shouldn't have been too far outside of the realm of your imagination. Are you sure you didn't mean 40 years ago? 20 years ago was only 2000, not exactly the stone age. – Josh Dec 30 '19 at 20:03
  • You are right, make it 25. – nvoigt Dec 31 '19 at 7:46
  • But "Cell phones had already existed for decades." is just wrong. Not for the normal guy. I got my first one in 1998 and the models before had been clunky to the point where you needed a battery suitcase. It was cool, but it did not fit in a jeans pocket and it would cost 1$ a minute (!) to call someone with a different provider, not even talking about an international call. Having a big enough screen to video call someone for free on the other side of the planet? Not on anyone's mind. – nvoigt Dec 31 '19 at 7:52

Don't do a PhD expecting it to be relevant to getting a job in industry.

A PhD is fundamentally a research degree; it is a piece of paper that says "this person is qualified to do research". While the highly specialized domain knowledge you'll acquire in the process of doing that research would also be valuable, it's entirely possible to get that domain knowledge much more rapidly through things like gaining experience in industry or a Master's Degree that focuses less on original research and more on doing coursework that focuses on mastering the current state-of-the-art rather than finding an opening to expand that state-of-the-art.

A PhD would, however, be quite valuable if you want to go into academia, or a research-focused role in industry like the R&D department of a large corporation.

  • I know the drill, there are some special circumstances for me specifically that ultimately make a phd worth it. (One being I want to know how to do research even if I don’t do that as a career.) I just want to find a field that is both somewhat in demand and that I like. Unfortunately, with all this superstitious resistance of people in giving me any useful info, I’ve decided I might as well go into what’s hyped. – HappyFace Jan 2 at 9:56

Not the answer you're looking for? Browse other questions tagged or ask your own question.