I'm in my last year of my BSc, and I have a couple of things that I am really interested in, and I am hoping to do a MSc and then a PhD. My adviser explained that a PhD is very focused, and it's likely that my future academic career will be mainly in that general topic. The thing is I like two very different topics. One is high level (machine learning), and the other is low level/theoretical CS (systems, parallelism, quantum) or at least that's kind of how these fields are grouped together at my university. I wouldn't mind doing a PhD in either of these fields, but I would like to have opportunities to work with the other field at some point in the future. And the way I see it, these are very different spheres of CS.

So my question is how broad or narrow is your research area after your PhD. Do you get opportunities to work in different areas? I know there isn't a physical barrier stopping me from exploring different fields, but how difficult is it to get such positions when your PhD is in a different topic? How far can you deviate from your initial research?

At present I have better research experience with low level computing, as my dissertation is on parallelism and quantum programming. And I expect it would be easier for me to get admitted to a PhD with a proposal in this field, since I can demonstrate experience. But I like to think that at some point after the PhD, I can also do ML research (like in a postdoc or as an RA).

For context, I intend to continue my education in the UK.

  • Many postdocs are hired to work on a given project but it's completely accepted and understood that you can pursue other interests and collaborations outside that particular project (provided your time management is good). So, as you say, there's nothing really stopping you. Jan 29, 2022 at 14:11

3 Answers 3


There's a Piled Higher and Deeper comic that addresses this:


The PhD tends to be the narrowest area you work on in your academic career; if you stay in academia your area of focus will tend to broaden, and senior academics often have a remarkable breadth of interest. However, part of this breadth often comes from an ability to leverage skills in one area to apply to others. The more interdisciplinary your area of research, the easier it is to find opportunities to apply your knowledge to distinct areas.

I did my PhD research focused on one specific brain slice preparation, mostly studying a connection between one part of the mouse brain (auditory thalamus) and another (auditory cortex), using pretty much one family of techniques: patch clamp electrophysiology and a bit of calcium imaging. As a post doc I continued this work, and added some mechanisms of anesthesia. Now I'm an academic scientist and study cortical networks in human subjects, consciousness, sleep, anesthesia, neuroinflammation, psychedelics, clinical anesthesiology and critical care; I guess I don't do brain slice studies of mouse auditory cortex anymore, though. I'm unusual in that I've stayed in the same lab most the entire time, which has both broadened and narrowed my interests (gives me more freedom to get my toes in everything going on; though I'm also more constrained to my PI's interests than if I were independent).

I do think you are likely correct that it's easier to start low-level and expand to high; AI/ML is a very hot area, but also very saturated. If you can find applications for ML in the lower-level projects you work on, you may be able to find a lucrative niche. If you go for future jobs in industry, you're going to find a lot of opportunities in AI/ML even if your specific background isn't in that area, as long as you maintain some competency there. If you go for a future in academia, your limits at the professor level are pretty much bounded by what you can justify in a grant (of course, you should be aware that a minority of people who enter academia aiming for a PhD will reach a professorship).


Most doctoral dissertations are very narrow and very deep. There are exceptions such as, perhaps, in the philosophy of mathematics and a few others. There are some interdisciplinary, often applied, dissertations, but most are on a single topic in a single field.

If you study CS, for example, and get a doctorate there, then your first employment is most likely to be in CS and people will expect that you do good work in that field. But even CS has a lot of room to move around in, so you don't need to get stuck in the deep and narrow hole it was necessary dig to get the degree.

A better metaphor, actually is that you extend the range of the known in a narrow but spiky way, pushing out into the unknown as far as you can, but usually over a narrow front.

But, once you have a secure position, the world is yours, and sometimes you are forced to switch fields. I studied math but needed to switch immediately to CS for want of any opportunities in math. My dissertation was so narrow that only about half a dozen people in the world were interested or could easily follow it. I finished my doctorate so long ago that it was unusual for a mathematician to know how to program, and I had to learn that.

But even in CS, I had broad interests that morphed over the years. At one point I was a language maven and usually taught the compiler course. But I was also interested in human aspects of computing and wrote in that area.

One of the nice things about study, especially graduate study, is that you learn how to learn. You can always push that button so that you don't get bored.


PhD is quite narrowly focused
A good PhD usually involves publishing several research papers in major journals. Working on a paper with an original subject as the first author is quite a lot of work - in terms of studying the relevant literature, doing calculations or measurements, writing this up, etc.; and takes from a few months to a few years. In this respect one is pretty narrowly focused. And, even if one manages to write several papers during the allotted time (whether it is doable largely depends on your field), they are likely to be on related subjects or subjects falling withing the same sub-field (or sub-sub-field, or sub-sub-sub-field...).

After PhD the world is wide open
There are different reasons why it does not restrict your future choices:

  • Your future postdoctoral or industrial employment will likely have nothing to do with your PhD thesis, even if in the general area where you specialized. (there are exceptions, where people continue to work on the same subject for decades. Some subjects do require this, but one have to be sure to start in the right field, where such focus would not be boring and would not impede the career.)
  • You can actually seek postdoctoral jobs in different sub-fields or even different fields of science - the drawback is that you will be a novice, with no prior experience, but the excitement for learning something new may worth it.
  • Switching between fields is actually rather common. There are also many interdisciplinary fields, such as bioinformatics - where people with biology background learn some math and programming, people with computer science or physics/math background learn some biology. Among other such fields one could mention quantitative finance (always hot and very demanding in terms of math) and environmental science.

Transferable skills
Note also that the most important things that one learns during a PhD are not the knowledge related to the subject, but general skills like:

  • conducting research
  • studying scientific literature
  • creating figures and presentations for communicating your results
  • writing scientific papers.

Many PhDs and their supervisors fail to see the importance of those, but this is actually what makes people very valuable outside of the academia (where your narrow research subject may be of little interest). In some countries the PhD is actually geared towards acquiring more transferable skills, rather than doing groundbreaking and/or independent research (although much always depends on the specific research group or supervisor's personal approach).

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .