I'm interested in knowing how easy or difficult it is to get a role where you publish at the top conferences at FAANG companies. (and other strong AI industrial labs like NVIDIA). Realistically I might go to grad school in a top 20-50 school in the US, and finish with say 100 citations and an average single digit h-index. I know the net-value of being a grad student is negative monetary savings in the short term, and maybe only slightly better in the long term. I think I would enjoy working on more open-ended problems and a research career more. So ultimately - how difficult is it to land an Applied Scientist or Research Scientist role where I can publish regularly at the top conferences? It seems like this is dependent on the network that my advisor has with people in the industry. I'm not sure how research at FB/G works. Are Deepmind/FAIR/Brain the only orgs where research happens? Are there scientists in other orgs who publish? TC: 180k ~ 1 yoe
What’s the h-index? How many published papers?
One has an h-index of n when one has at least n publications with n citations but not at least n+1 publications with n+1 citations
Quality of papers is more important than quantity. More top conference papers make a better profile. If you go to top 10 schools for phd, things are easier. Top 20 to 50, I would say choose the professor and research group wisely. Good luck.
This just have papers at the top conferences of your research field and you’re in the green
Tech Industry
3h
328
Does the market believe in Sundar Pichai?
Tech Industry
Yesterday
1857
The end of Backdoor Roth?!
Layoffs
Yesterday
3528
Google laid off 200+ in core team
Tech Industry
Yesterday
2318
Quitting this Slave life
Working Parents
Yesterday
679
What do you think is wrong with a kid who got rejected by 9 colleges?
FB/G require h-index of 10+ for entry level research scientist positions. I had 15 when graduating from PhD and managed to get interest from HMs. Although it's more competitive in 2020 and it's almost at the competitive level of faculty hiring. Applied scientists don't publish. They simply apply other papers to the company's products. Don't do a PhD for (only) the financial gain. You need to have a strong curiosity to learn. NVIDIA is the easiest to get but AI research at NVIDIA serves a marketing role. They aren't revenue generating. This gets reflected in the prioritization of research projects.
Interesting. Is there a source for this 10+ bar? :) sounds about right that it's as competitive as faculty hiring. But okay, let's argue that I'm able to get to that bar through some combination of work and luck. I'm not aiming for financial gain, but I do think that I prioritize stock options (even as a researcher) over moonshot start ups. I'm still not sure if research at FB/G is limited to the three orgs I mentioned. Are there researchers outside those orgs? Conceivably, Google Assistant or YouTube orgs need researchers too?
There’s no such bar. On my team, we have a hard time finding good candidates who can look past the academic research into practical applications meaningfully. I couldn’t care less what your h index is if you can demonstrate deep, grounded understanding of research.