I am funded by the university for my PhD, though I had approached the supervisor myself, particularly because he had decent publications in good journals. I am an international student. I have finished exactly 1.5 years of my PhD and my funding is only for 3 years. I am working on an Artificial Intelligence/Machine Learning problem.

I work on a topic which is more academically exciting than of much use to industry. I did not know that before I started my PhD. Let is call this area A. I confess that I was told I will be working on this area. Though I must mention that along with A, B is another topic which was mentioned and I am working on both, though A is where my supervisors "specialize".

Now my problem: 1. A is not really going to get me a job as a respected scientist in industry after my PhD. So I MUST consider A with other ideas from other popular methods -- and I take the burden to prove that my idea is correct.

  1. A has been overused in my group and I am already getting reviews for my paper that it is very similar to other works from this group. I get to believe that I am doing a shadow PhD in previous student's work who has graduated with "flying colors".

  2. I am not a fan of A myself, particularly because it lacks mathematical foundation and my supervisors have no intention of letting me pursue such a direction, nor are they capable of providing me guidance in this area.

I am not looking for an academic career, but wish to be a researcher in industry. So I need to make my research relevant.

Now, when I approach my supervisors (three of them, one of them added by my primary without even asking me, to help him!) telling them that I want to work in B more than A, and also wish to include C in my research I am met with stiff resistance. The argument being this is not relevant to the conferences in A. If presented in a conference of C, people will ask too many questions etc. I would appreciate a proper discussion in being shown that C is not technically useful or viable or something like that.

This clearly is not scientific research. So I am considering raising this issue in my half-monthly report. But this will surely make the supervisors work against me and they would naturally not provide help whenever they are required to. What are your suggestions?

Note: My supervisor has friends in all the journals and conferences he publishes. He does not send a paper to a new conference where he has no friends!

  • 1
    I'm not sure what you mean by "raise a red flag against my supervisors". Do you just mean you want to tell your supervisors again that you don't want to work on A? Or are you referring to some other action on your part? (Note that while you don't have to work on A, your supervisors are probably free to say that they're only prepared to supervise work on A. So you may just have to find a new supervisor, if you feel that strongly about changing topics.)
    – ff524
    Nov 4, 2016 at 5:23
  • @ff524 has raised a good question. I have guessed at what you're trying to say and have edited your title accordingly. Please roll it back if I guessed wrong. Nov 4, 2016 at 5:28
  • 1
    What academic system are you in? That is going to affect how supervision relationships are expected to operate. Nov 4, 2016 at 6:05
  • 2
    Suppose you got a new supervisor and topic. You would probably need more than 1.5 years to get to the point of having some papers and a dissertation in that topic. Do you have an alternative source of funding to continue beyond three years total? Nov 4, 2016 at 6:08
  • 1
    Who is going to read the report? Nov 4, 2016 at 12:36

1 Answer 1


Getting your PhD in good time is probably the most important for your career, so you should focus on that. It would almost certainly delay you to change topics now, and adding extra subjects "on the side" may also endanger your timely completion of the core work. And plus, as you are aware, you are highly dependent on the goodwill of your supervisors in this situation.

Assuming a change of topics would introduce a one-year delay in your studies, there are alternative ways you could use that year to build your CV:

  • Work in a postdoctoral research position on a more "practical" topic.
  • Work in industry for a year in a lower-level position.

Both of these would likely be more effective for your long-term goal than switching topics now. And, in both cases you would be paid and wouldn't have to worry about your funding running out. Of course, you would have to find such a job, but I still think it is lower risk than trying to change topics now against the will of your supervisors.

Also, your fears about not being able to get a job because the topic is too theoretical may be unwarranted. Generally, most of the specific things you do in a job are learned on the job. Employers will look at your education for:

  1. Evidence of your ability to learn and succeed. Simply completing your PhD (and doing it well) demonstrates this.
  2. A good theoretical background in the field. Working even on an impractical "Artificial Intelligence/Machine Learning problem" should show this.

In short, focus on doing an excellent job with your current PhD topic; there are still logical steps from there to your goal.

  • 2
    If machine learning continues the way it has done for at least the last 40 years, whichever method(s) the OP uses during PhD research will not be what the OP will be using in a few years time. Doing early career work with a different method from prior PhD research is a good way of demonstrating flexibility and ability to learn on the job. Nov 4, 2016 at 13:05

You must log in to answer this question.

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