3

I have applied for a PhD in Computer Science. I do not however plan to become an academic afterwards. What I want to achieve is to try to commercialize my research by founding a startup or find a job outside of Academia, using the knowledge I will obtain for enhancing or building products/services for my employer.

What I would like ideally, is to achieve Academic as well as Industry expertise in machine learning and AI, through a job or by trying to found a startup with a related thematology. Becoming a professor is completely out of my goals. However I want to be able to study academic literature with ease.

Do you think my goal is reasonable, or a PhD is an overkill that would make me overqualified, loosing precious time in case I change my mind in the future ?

I already have an MSc that provided a machine learing and image analysis course. My Master's thesis was also heavily focused on these subjects. But the reason I want the PhD is to learn more things and receive more hands on experience about Machine Learning and Complex Event Analysis. I believe that this process will also enhance my critical skills and analytical thinking.

I understand this can get difficult and risky but I am willing to try.

Does my goal make any sense or do I need a reality check ? I need hard critique..

Is that a healthy reason for attending graduate school ?

  • 2
    Because this depends very much on personal circumstances - and also subtle differences between courses, two different PhD programmes may have very different structures and serve as very different preparation for the later world - I think answers to this question are going to be prone to subjectivity. I wonder if it would be possible to rephrase this question to focus on a more objective criterion. – Silverfish Oct 22 '15 at 22:46
  • 1
    criterion such as ? –  Oct 22 '15 at 22:49
  • That is something that the asker of the question is in a best position to judge :-) But one of the reasons to close a question here is if, as written, it would generate answers which are primarily opinion-based. – Silverfish Oct 22 '15 at 22:54
  • If you want an example (this is just that, not a recommendation) of an objective criterion: you might ask whether there is evidence that average salaries post-PhD in this subject compensate for both lost earnings during PhD (for someone whose earning capacity is already boosted by an MSc in this discipline) and the cost of taking a PhD, That is clearly evidence-based and somewhat objective. But even such a question clearly varies massively according to circumstances, especially career aspiration and choice of school – Silverfish Oct 22 '15 at 22:54
  • More relevant for http://academia.stackexchange.com/? – rightskewed Oct 22 '15 at 23:20
  • No because academia.stackexchange contains mostly people from universities, while here people from the industry also make posts. Opinions on this subject can be biased. I have also posted this in the workplace.stackexchange for the aforementioned reason. –  Oct 22 '15 at 23:23
  • 5
    [Here's the cross-post at Academia SE](http://academia.stackexchange.com/q/56716) - they have closed with the sensible reasoning ""The answer to this question strongly depends on individual factors such as a certain person’s preferences, a given institution’s regulations, the exact contents of your work or your personal values. Thus only someone familiar can answer this question and it cannot be generalised to apply to others" – Silverfish Oct 23 '15 at 00:17
  • 4
    [Here's another cross-post, at Workplace SE](http://workplace.stackexchange.com/q/56415). Note that cross-posting is generally discouraged. – Silverfish Oct 23 '15 at 00:19
  • 1
    Another one: https://www.reddit.com/r/MachineLearning/comments/3pts4i/how_can_a_phd_in_machine_learning_and_ai_help_you/ – rightskewed Oct 23 '15 at 00:23
  • Is stackexchange affiliated with reddit ? –  Oct 23 '15 at 00:25
  • 1
    To the extent this is a generalisable question, not just one about specific individual circumstances (which are not detailed in depth in the question text), this is a near-duplicate or perhaps even duplicate of [this question about earnings premium of PhD over MSc in machine learning](http://stats.stackexchange.com/q/27495/22228) – Silverfish Oct 23 '15 at 00:30
  • There are definitely jobs for big data looking for a Ph.D. I would say more as a employee with a big firm that does big data. They tend to want people that will learn their data and build some analytics. Clearly a start/up and consulting firm is an option but in my opinion I would rater get hooked up with with a big company on their in house big data team. – paparazzo Oct 23 '15 at 01:02
  • 4
    Cross-posting without immediately explaining what you have done is poor practice. It's true that SE as such has no call to tell you what/what not to do on Reddit and vice versa, but that's a matter of forum rules, which do not describe internet etiquette completely. People feel let down if they post an answer in one place and then discover that someone else wrote something similar earlier somewhere else. You have to show that you care about the time and effort you are asking strangers to give for no evident return. Regardless of that, cross-posting on SE is explicitly discouraged. – Nick Cox Oct 23 '15 at 11:07

1 Answers1

8

I started out on a Ph.D. track. About halfway through I realized that I wanted to be out working in industry more than I wanted to be working towards a Ph.D. But that's me..

A PhD is a bit of a grind, and as with everything else that's hard in life, keeping one's motivation throughout is a function of the intrinsic value you derive from the pursuit. In other words, if you're going to a do a Ph.D., don't do it because of where you think it will get you (although that certainly is a factor in the equation). Do it because you have a passion for learning about a field, for exploring it to the very edge of what we know about it, and then contributing new knowledge to it.

In the end, that was what I thought was most rewarding about graduate work: to know something that few people, possibly no one else in the world knew, and then to contribute that knowledge to make the world a slightly better place.

But to take the pragmatic view, yes, I believe a PhD in Data Science can indeed translate to a good position. Will it translate to a better position, than if you go spend that 5 years working? Probably not.. My hunch is that, six years out, you'll be making more income if you just go to work. If that is your measure of success...

But there are some advantages of doing the graduate work: if you choose a topic that is bleeding edge, and you happen to pick a topic that's getting traction in 5 years (hard to predict...), and/or you make significant connections in your chosen industry, then, yes, there is a chance you'd be a little better off.

TimBaynes
  • 96
  • 1