Great, balanced & honest overview of this very important decision many of us had to make at some point. Thank you! Speaking for the "other side", my only comment would be on the 'commitment-to-topic' aspect. Nobody can predict with certainty how 3+ years of research onward, the state-of-the-art will look like. As a result, the topic of my PhD thesis concerns an area of representation learning which didn't really properly form (certainly not in the way we know it right now!) at the time I started my PhD. Assuming your advisor is open-minded and supportive, it is certainly possible (and often encouraged!) to take regular internships, either industrial or academic, while doing a PhD. This can at the same time expose you to industrial work and offer wide classes of interesting problems that can feed back into your research. I've taken it to the extreme (my PhD thesis is consisted of _only_ work done during internships), but I'm certainly not the only one with a similar story. So a commitment to topic is not always necessary and some degree of wiggle room tends to always be possible. Of course, while this works for migrating across certain subfields of machine learning, this may not be the case for moving between machine learning and type theory (though I have seen some work pop up at the intersection! :) ).

This is a really simple situation: When asked, "Why do you want to do a PhD?", if the hesitation is longer than 5 seconds, skip it. It requires a level of obsession which the student already must have. Some care should be taken when talking about "industry", because the handful of places like Facebook and Google are relatively unique in offering "research" opportunities. Most companies are about a combination of design and implementation. Many will find that exciting, but it should be considered accurately.

A more chill video discussing why I’m not doing a PhD, as requested by you guys. I’ll release the video about my degree later as it’s still being scripted and it’s a video I really want to make a step function better than other videos!

Hi! I'm hoping you're in good health wherever you are. Thank you for the insights you've provided in this video. I'm a Master student and a graduate research student who's currently feeling the same conflicts. I want to be an expert in my field because I'm not that good but on the other hand (in my opinion), researching is kinda tiring because you need to be able to constantly compete with top minds around the world. This video gives me a new perspective. Thank you!

Why does everyone think the choice is binary - EITHER get a PhD OR get industry employment experience. You can do both!!! It’s called an industry emended PhD where you’re a 100% full time employee getting hands on industry experience solving real world problems while simultaneously getting a PhD. This is not new :)

You made really good and fair points. 80 % agree with you but I do have some concerns, One is reputation. Most of the time, thanks to PhD holders we have a profound and solid theriotical framework in our higher education system specially in the area of computer science. At the same time I see DRY(Dont repeat yourself) in the industry where we will limit for the certain frame. Btw, thanks for posting this video. Cheers

Yeah, I found choosing of a topic is really difficult

I did a Physics PhD, went into industry straight after and spent 6 years working up to a £100k+ salary. During those 6 years, I didn't earn up to £40k per annum unless I tutored/lectured on the side. I think I played the long game very well. The UK is horrible for actual STEM and I would not suggest anyone to stay here after having done a STEM PhD. Go overseas instead. If you consider "tech" to just be a bog standard data analyst/scientist role that a PhD could do in his sleep, then of course remain in the UK and hope that the hiring manager doesn't overlook you in favour of someone who just did a BSc in Comp Sci.

Perfect introduction! 😄

Thank you so much for this useful video. I had the same thoughts regarding the industry and academia when I thought about the scenario where I would pursue a PhD.

Hi brother. I have done my UG in computer science (BSc) i am really interested in Deep Learning, Reinforcement Learning and very much interested in learning how the brain works.I am doing self learning on these fields.I figured out that Computational Neuroscience helps in understanding brain functionality with the help of math and computer.I wanted to do research in this field and understand how brain works. So my question is can i get a research level job in any company with the educational ground i have.

You are really great and your contents are soo good.Waiting for the video where you will discuss about your Cambridge degree.Keep it up.We want to see more videos from you like this

Helpful

Do the residency industry research programs require you to be enrolled in a master's program? Or can it just be a full time one year thing? Also, it's a pretty neat little video on a much neglected topic so big up!

Absolutely!

Thanks buddy, I have been having similar train of thought, now I feel I am having confirmation bias. So here is my dilemma, I just have a bachelor's degree in computer science instead of doing a masters I joined an early stage startup where I got exposed to research and topics like privacy preserving ML, a lot of this stuff doesn't work and still in early days. Also a lot papers that I came across is real garbage, I understand that you have to get yourself published once you get into the PhD but many people say that PhD is just another barrier to get FAANG roles. Also to get Applied Scientist roles do you really need a Masters/PhD cause it's mentioned as preferred/mandatory under job requirements but many of the packages I have seen written like Crypten and other PPML are written by typical software engineer folks. So is Master's really mandatory to get into FAANG if you have some work ex in DS/ML? Waiting for more videos to come on this, I guess you will be covering more on this I guess.

Very informative. I left my corporate job to join PhD program and now I regret my decision. I'm still in the exploring phase and haven't chosen a topic yet, so I'm reconsidering my decision for now.

Money >>> Phd

Great video! Often find myself asking this same question! What field did you do your bachelor's dissertation on?

What are maths require for computer vision research scientist?

Hey mukul I have a question. I'm in 2nd year engineering college cs degree. Recently I have applied for fb internship for swe role but my resume wasn't shortlisted (due to no previous internship). So I did an internship at startup, but now the startup is going to be shut down. So my question is that if I get selected for internship/full time position in future. Will they run a background verification check for internship mentioned in my resume, Cause I'm scared if something goes wrong even after getting offer. I know my question is weird but I just want to ask u that do fb run background check verification for internships, did they done bgv for your previous internships? Thank you in advance sir!

Do you have a PhD?

Hlo sir can I talk to u