Does it make sense to choose a task from the field of AI as a thesis topic with further development as a PhD thesis?

Question

IT education

Hello!
Tell me, please, does it make sense to choose any task from the field of artificial intelligence as a topic for a thesis with its further development as a Ph.D. work? Or just as a thesis? That is, are there any questions, the study of which, on the one hand, would be at least a little useful and at the same time new, and on the other hand, not too complicated. Difficulty refers to the required amount of time and the availability of materials for study. I would like to complete the task on time, and not get bogged down for a long time.
To a large extent, practical implementation is of interest, i.e. I want not only to engage in theoretical research or ready-made software packages, but to write something on my own.
Sorry if the question is vague and not specific. I will be glad to leading questions in the comments.

Answer 1

It all depends on the degree of deepening into the problem personally by you. If you are just starting to study what AI is, then without fanaticism it will be very problematic to defend a dissertation on time. If you study where this problem is already being dealt with, or if your research guide. deals with this problem, then it will be easier.
As for the choice of task, I can advise you to look at monographs, defended dissertations on the topic of AI, and look for a task in this direction.
In general, the practice of Russian postgraduate studies shows that in order to defend your defense on time, you need to do it already with a practically ready solution.

Answer 2

In general, the practice of Russian postgraduate studies shows that in order to defend your defense on time, you need to do it already with a practically ready solution.

If we are talking about postgraduate studies in IT, then perhaps this is so. If in general - not at all like this: with a practically ready-made solution, you can defend yourself in general in a year. And in three years, and with postgraduate requirements, you can defend three papers. ;)