Strong/Weak Reasons to Do a PhD in Computer Science
raymondcheng.net> if I wanted to do a career change and start working in computational biology, would I do another PhD? Of course not! I would read papers, find good mentors and colleagues in the field, and start doing work in the space.
I mean I’m sure that’s all good and fine for the author who if I understand correctly already has academic accomplishments and connections to lean into with.
But what if I or John Nobody wanted to work in computational biology? Am I supposed to just start reading and than reach out to randos working in the field and beg them to let me work for/with them?
The best way is to apply for a job in a computational biology lab or research center. Maybe read a few papers too. Most of them need software developers. If you do good work you'll be a co-author.
Another path is to join one of the many many NGS startups where you too can analyze Nextseq data.
That will give you sufficient credentials to get into a PhD program if you want to do that.
Also there's an ex-hedge fund guy D.E. Shaw who runs a comp bio group, also Wolfram and all the big tech companies have comp bio groups.
Work there, see if you like it, that may give you an opportunity to enroll in a PhD if you want.
It will probably pay less than you'd make in a regular software job but its enough to live on and likely interesting work.
No matter where you go find a lab that is generating data. Data is the key.
Great advice! I did exactly this and realized a PhD is not for me, yet enjoyed being part of the scene as a developer very much!
It'll be harder for you than it would be for him but I still wouldn't discount it! Considering the alternative is (1) trying to get into a PhD program (might require a year or two of remedial classes with no guarantee of success) and then (2) spending 4-6 years doing a PhD, reaching out to the randos may well be much faster AND lowest risk.
If you're a software engineer, I'd venture a claim and say that the rando route will be easier. You have a valuable skillset researchers need -- you can code at a professional level. Use that to get your foot in the door!
Great post, Ray!
I particularly like the mention that a CS PhD isn't about "novelty" or "freedom", and I've found this to be a common misconception that new CS PhD students make. Another common one I will bundle in there is "making an impact on the world"; I cringe when a new PhD student tells me they chose to do a PhD because they wanted to change the world. There are so many better ways....
This is a great point. There are many avenues where one can invest time+money to make a positive impact on the world, yet "doing a Ph.D" is unfairly held up on a pedestal, even though there is no hard science showing that grad school is a prudent way to attempt such an endeavor.
Edit: adding examples of ways that are less sexy, yet way more likely to be measurably impactful (in a positive way)
1. Teach high school,
2. Volunteer at your local soup kitchen (do it for 2 years -- less than half the time you'd spend in classes for quals),
3. Mentor an undergrad for a decade,
etc.
So I am someone who has done a masters in CS but it was completely coursework and project-based. I have regretted not doing any "research" or having any papers published. I guess, I just want to experience the whole process of diving deep into a topic and coming up with something new. Now, I'm about 3 years into my career as software engineer working on data infrastructure for a data science org in a large non-tech company. Could I potentially volunteer my coding skills to a professor who has a lab in some field I am interested in and join in their work in my personal time. I never thought that is possible. I'd actually like that. Can someone who has any experience or seen something like this comment?
I'm biased here since I know Ray, but I thought this post was super informative & helpful! I love how short & sweet it is yet so chock full of meaningful points. Thanks Ray!
Thanks! If only I could teleport back in time and tell myself some things
Here's an interesting notion -- that a lot of wisdom is simply bound up in the integral over time of personal life experience.
Thus, the value of wisdom is individual, and non-transferrable -- it can neither be carried "backwards over time", nor handed over to another human via a "brain dump".
The best one can hope for, is to distill personal lessons from our own experience (compute that integral), and invite others to sample it (no guarantees that it'll stick in any way, but their journey might be similar, allowing them to compute their own integrals faster, by seeing our own).
This is why you read books about people. Success leaves clues.
Thanks for writing and sharing this!
Guilty of weak reason #2. This post puts things into perspective. Thanks!