Which company is the most promising for the current and future careers of ML/AI scientist?
Please share the company in your mind among FAANG + Lyft, Uber, LinkedIn, MS, any Unicorn or Promising Startup
- New / Eng undecidedFor research, google brain, microsoft research, facebook fair, deepmind, open ai. Just look at number of publications each have in nips and icml.
- Yeah, thanks. I heard the similar story on Google and MS AI. Google publishes better but the link from research to the project is weak and MS AI tries to focus on the applications. Is pure research publication meaningful without the impact on the real product for a long term view in career path?
- Facebook bTId74I’ve been at 3 of the industry AI labs you mention. I think you should distinguish between research caliber and career growth, they’re not the same.
G Brain is at the top, followed closely by deepmind and FAIR when it comes to research. For career growth, fb (not FAIR), product parts of G and uber/linkedin/netflix are good. You have to understand that if you want career growth, you won’t get to work on latest/greatest/sexiest stuff.
MSR used to be great 10 years ago, but it’s no longer worth considering.
- The tech giants are the best, Alphabet, MS, Amazon, and Apple... they have the greatest resources, out of these four, the best are MS & Alphabet given they both have cloud computing + Quantum AI research putting them at the cutting edge.
Source: Me an AI expert.
- New / Eng undecidedI counted the number of papers by each in last year's nips or icml, don't remember. Google brain was first, Microsoft close second, then deepmind, then facebook fair. Something like that. Alphabet is the highest given both brain and deepmind, but again they're like different companies. MSR has exceptional researchers in the field.May 5 1
- I’m not doubting FB’s team. I know they are good as well, but in the case of resources needed just to run an DL experiment, companies with cloud computing beat them out. Primarily because without certain resources you can’t even train a dl model. I was at apple when they didn’t have enough resources for dl models, and it wasn’t optimal (this was when the field was still new).May 6 0
This comment was deleted by original commenter.