Ask HN: Is this the right time to pursue a PhD in Machine Learning?
20 comments
Well if you are getting at the top, go for it. You will build more credibility and trust from top universities rather than working on some startup. If the startup happens to be like OAI, then choose that one. The rate you learn at startups is something else.
A good idea would be to try research in academia for 2 years, see if you like it not, then drop out with a Masters. If you are a citizen of the states, why not do both? Work at startup and pursue PhD as a hobby? Pretty sure you are capable of doing that.
A good idea would be to try research in academia for 2 years, see if you like it not, then drop out with a Masters. If you are a citizen of the states, why not do both? Work at startup and pursue PhD as a hobby? Pretty sure you are capable of doing that.
Appreciate your opinion. Unfortunately, I am a Masters student continuing for a PhD, so dropping out with a Masters is not an option for me. A PhD in general is quite demanding, so, doing both would be next to impossible.
You can enter the program and if you don't like it you can always drop out and join slightly stable(Series B / Series C) startup. You will have amazing network. Startups by very definition in failure mode until they crack PM fit, especially the early stage one.
This. I regret not trying a PhD. It is like turning down a call option :-)
If you are going to into a top 4 school it is something that will look good on your resume unless you already went harvard, stanford, etc. The Research community seems to like phds from prestigious schools.
Yes! I happen to have a masters degree from one of those schools.
https://www.happyschools.com/bachelors-vs-masters-vs-phd/
Generally, advanced degrees are worth it for areas where you can't get the resources to learn yourself.
CS is unique in this case because all you need to learn everything computer science is a laptop.
Generally, advanced degrees are worth it for areas where you can't get the resources to learn yourself.
CS is unique in this case because all you need to learn everything computer science is a laptop.
Stanford and MIT are offers that you shouldn’t turn down.
You can work on an industrial project, as if you are in a company, and make friends with great people!
You can work on an industrial project, as if you are in a company, and make friends with great people!
Generally the industry is in a down turn right now (though AI is pretty hot). It might take a bit for it to recover. If you pursue a PhD you’ll be in good shape to go into corporate research when things are moving again.
Also, public research desperately needs some more material. It’s been consumed by the private space too much.
Also, public research desperately needs some more material. It’s been consumed by the private space too much.
For you yes - you're interested in research and have an offer from a top university - take it.
Do PhD if you want to be professor, otherwise masters is enough imho
Is this specific to machine learning ? I feel industry is slowly eliminating ML research scientist positions and transitioning to engineering only off late.
There are multiple research positions, of course lesser than engineering but the work that researchers do in big tech is no where close to research done in universities
MS might be enough to work in industry but isn't there a consideration between paying(MS) and getting paid(Phd via stipend) as well?
Nope. Want to work at OpenAI and Anthropic as researcher? Or hell even a small betting outfit as a quant? PhD.
Seems like you want to go work on the start-up.
I want to do what's optimal in the long run. although, working on the startup is tempting i feel it will be sub-optimal for a long term career in research.
Of course. If you want to do research you know the path. If you want to make money leave now. If you want even more money leave later.
Start ups will always be there whether you want to be the founder or employee 1 or whatever.
While I love research, I do have a fear of missing out on how hot the AI/ML in industry is right now with excitement in all areas - robotics, web agents, coding assistants etc. I'm looking for advice on choosing between the two options
1. *Ditch PhD*. Is the right time to jump on the industry bandwagon (ditch PhD) ? I'm considering joining an early stage startup in AI/ML as a founding engineer.
2. *Go for PhD* Is continuing for a PhD the better option for the long term ?