TL;DR Should I go for an ML master's at a top uni (have offers from CMU MSML, Cambridge MLMI) or continue working at Meta. My short career timeline: - BSc in CS from a top Swiss uni - Joined Meta as a new grad (IC3) SWE, joined an ML team and changed profile to ML Generalist - Mostly data manipulation and backend, no ML modelling - recently promoted to IC4 after 1 year Motivations to go: - Ultimately I would like a research role at an AI lab like DeepMind/FAIR/OpenAI. As I’m not too keen on doing a PhD, I’m aiming for research engineering roles rather than research scientist. - I could be more innovative in creating new ML solutions if I had a strong theoretical background, rather than just knowing what I learned on the job (i.e. data manipulation) - I’m not learning any technical skills in my role anymore, mostly just project management and communication. Motivations to NOT go: - Heard that a master’s degree is useless after some YoE in the industry, but is this true for top AI labs as well? - Possibility to transfer to a FAIR applied research team internally, though I’m not sure if I will have enough skills to be useful in a research team. - Monetary cost What would be the better choice in the long-term? Having a hard time deciding, some advice would be helpful! TC £135K, 1.5YOE #meta #deepmind #fair #openai #masters #machinelearning #research
Hands down Carnegie Mellon university! Best ML and NLP curriculum and professors out there.
PhD or stay in industry.
Don't do it lol.
Why not?
Not worth it. I did it and it was a monumental waste of 80k
CMU since it’s worth the opportunity cost to get access to US salaries.
Fwiw you’d get better responses by asking ml researchers you work with.
Thanks you’re right, I have also been doing that in parallel. But I thought asking here wouldn’t hurt either so why not try.
What you will learn is prompt engineering or tuning, just stay
its worth it to go to CMU just to get into the US. The UK is a terrible place to be for your longterm career.
Go to Cambridge