Facebook AI Research Expands with New Academic Collaborations • Hacker News Board

Re: Facebook AI Research Expands with New Academic Collaborations

#51

Post by eachro » Fri, Jul 20, 2018, 4:58 PM UTC

dharlow wrote: post #40

visarga wrote: post #27

Earlier quoted context omitted.

That might be true, but don't confuse FAIR (their DL research lab) with FB. FAIR employs good people doing interesting and useful research. They promoted pytorch as a framework, and it is perhaps the best framework for non industrial applications. On a parallel line of thought, I prefer FB's React to Google's Angular. In both React and Pytorch I see the same elegant design. TensorFlow and Angular on the other part ar…

FAIR is part of FB. The reason why FB invests billions of dollars into FAIR is because it supports its business model and its democracy-wrecking product. FAIR researchers are complicit in the societal damage perpetrated by FB. They're cashing the (multi-million $) checks, and in exchange they work on making FB more powerful, by giving it better AI. When you talk to these guys (they're almost all guys), you realize th…

The $$ they can get from anywhere else in industry given how hot research level ML experience is (and how scarce this talent is). I imagine the appeal of FAIR is much more about the academic freedom. But sure the $$ doesnt hurt.

Re: Facebook AI Research Expands with New Academic Collaborations

#52

Post by username_123 » Fri, Jul 20, 2018, 4:59 PM UTC

edhu2017 wrote: post #44

username_123 wrote: post #28

What level are these professors being hired at within Facebook? What sort of base salaries, stock options do they command? Any idea on total compensation numbers here? With them being part-time professors what are their university salaries like? In total compensation between industry and university I'd imagine this would be a good deal. Lastly, how does recruiting work here? I assume they don’t make these people go t…

Total compensation: Definitely a wide range, but a few hundred K would be the median. Recruiting: one of my labmates flat out refused to do algorithmic interviews for a research position and the company still gave him an offer. Your projects, papers, and ability to demonstrate your understanding of research topics are far more important then your ability to solve a small, defined problem in 45 minutes.

I assume there is a distinction between software engineering and research applicants in terms of their interviews at the Big 4s? I know two people with PhDs that were given algorithmic questions and had to go through the same process as regular software developers (this was at the same Big 4 company). The only exception is one had expertise in ML and during one of his interviews they brought in a ML expert and interviewed them on their dissertation, in addition to their 4-5 algorithmic interviews. Both of these people were interviewing for software developer positions despite having expertise in research. Wasn't sure if the research interviews would be different. Good on your lab mate.

Re: Facebook AI Research Expands with New Academic Collaborations

#53

Post by electricslpnsld » Fri, Jul 20, 2018, 5:46 PM UTC

username_123 wrote: post #52

edhu2017 wrote: post #44

Earlier quoted context omitted.

Total compensation: Definitely a wide range, but a few hundred K would be the median. Recruiting: one of my labmates flat out refused to do algorithmic interviews for a research position and the company still gave him an offer. Your projects, papers, and ability to demonstrate your understanding of research topics are far more important then your ability to solve a small, defined problem in 45 minutes.

I assume there is a distinction between software engineering and research applicants in terms of their interviews at the Big 4s? I know two people with PhDs that were given algorithmic questions and had to go through the same process as regular software developers (this was at the same Big 4 company). The only exception is one had expertise in ML and during one of his interviews they brought in a ML expert and interv…

I interviewed for a research position at Facebook reality labs, and it was the same process as an interview for an academic job. Give an hour talk about your research, meet with 6 researchers in hour chunks for the rest of the day. The process was similar for my previous industry research job as well. I can't speak to Facebook's other research arms or Google.

Re: Facebook AI Research Expands with New Academic Collaborations

#54

Post by gaius » Fri, Jul 20, 2018, 5:51 PM UTC

taneq wrote: post #43

naturalgradient wrote: post #4

I understand the call of money but I cannot help but feel very negatively towards academics doing this with facebook of all organisations. After recent events, they cannot pretend not to know the impact and damage their work may have here. Excusing yourself with "I am just a researcher, I don't have anything to do with how my work is used" is just not good enough any more. I would categorically reject any collaborati…

It's a bit of a poisoned grail, though. On the one hand, you're selling your soul. On the other hand, they have data beyond your wildest dreams and you can use it all for anything as long as it might make money. Would you take that offer? Would I? Probably not. But I can see the appeal, and plenty of people wouldn't hesitate.

But I can see the appeal, and plenty of people wouldn't hesitate.

It will keep happening until these companies are a black mark on your CV, like say a tobacco company might be. Will people be so eager if it means they will be shunned by the wider research and engineering communities?

Re: Facebook AI Research Expands with New Academic Collaborations

#55

Post by traek » Fri, Jul 20, 2018, 6:01 PM UTC

dharlow wrote: post #40

visarga wrote: post #27

Earlier quoted context omitted.

That might be true, but don't confuse FAIR (their DL research lab) with FB. FAIR employs good people doing interesting and useful research. They promoted pytorch as a framework, and it is perhaps the best framework for non industrial applications. On a parallel line of thought, I prefer FB's React to Google's Angular. In both React and Pytorch I see the same elegant design. TensorFlow and Angular on the other part ar…

FAIR is part of FB. The reason why FB invests billions of dollars into FAIR is because it supports its business model and its democracy-wrecking product. FAIR researchers are complicit in the societal damage perpetrated by FB. They're cashing the (multi-million $) checks, and in exchange they work on making FB more powerful, by giving it better AI. When you talk to these guys (they're almost all guys), you realize th…

At that level, they could be making the same money at any of the other tech companies, with the same freedom and same caliber of coworkers.

I highly doubt most of them believe that FB and the work they’re doing is evil. They would have no incentive to work there if they did.

Re: Facebook AI Research Expands with New Academic Collaborations

#56

Post by naturalgradient » Fri, Jul 20, 2018, 6:02 PM UTC

throwawayjava wrote: post #26

I think Facebook hate is clouding out reasonability in this thread. E.g., there's even a comment asserting that people who collaborate with facebook (a.k.a., Jessica Hodgins, Andrea Vedaldi, and Jitendra Malik) are "not worth their salt" (?!?!?!). IDK. Maybe -- just possibly -- there do exist researchers who have their choice of funding spigots and are choosing to work with Facebook. Either that or HN has some damn h…

You are referring to my comment about 'worth their salt' in a gross misreading. I said anyone worth their salt has other choices, so going to FB is a deliberate choice on these professors.

Re: Facebook AI Research Expands with New Academic Collaborations

#57

Post by mliker » Fri, Jul 20, 2018, 6:05 PM UTC

naturalgradient wrote: post #4

I understand the call of money but I cannot help but feel very negatively towards academics doing this with facebook of all organisations. After recent events, they cannot pretend not to know the impact and damage their work may have here. Excusing yourself with "I am just a researcher, I don't have anything to do with how my work is used" is just not good enough any more. I would categorically reject any collaborati…

Fb's research in computer vision has produced works like training a neutral net on large datasets in 1 hour, instance segmentation neural nets, and many more, and made the code and research public with nonrestrictive licenses. These are pushing the state of art! Check out the work yourself and then come back and criticize these scientists if you feel that their work is a net negative on the world.

Re: Facebook AI Research Expands with New Academic Collaborations

#58

Post by electricslpnsld » Fri, Jul 20, 2018, 6:11 PM UTC

Kyragem wrote: post #22

As usual FB and other companies only collaborates with the "elite/Ivy League" schools. IMHO there are plenty of great researchers at non-elite schools and it would have been great if FB would start collaborating with those to reduce this vast concentration of academic research that has been happening for the past decade.

Half the schools on the list are publics, and none of them are Ivies.

Re: Facebook AI Research Expands with New Academic Collaborations

#59

Post by username_123 » Fri, Jul 20, 2018, 6:11 PM UTC

electricslpnsld wrote: post #53

username_123 wrote: post #52

Earlier quoted context omitted.

I assume there is a distinction between software engineering and research applicants in terms of their interviews at the Big 4s? I know two people with PhDs that were given algorithmic questions and had to go through the same process as regular software developers (this was at the same Big 4 company). The only exception is one had expertise in ML and during one of his interviews they brought in a ML expert and interv…

I interviewed for a research position at Facebook reality labs, and it was the same process as an interview for an academic job. Give an hour talk about your research, meet with 6 researchers in hour chunks for the rest of the day. The process was similar for my previous industry research job as well. I can't speak to Facebook's other research arms or Google.

Good to know. So you you weren't asked any algorithmic leetcode-style questions or questions testing that you have basic knowledge of your field (e.g. tell me what a random forest is)? Doesn't sound like it. It was more focused on your research?

Re: Facebook AI Research Expands with New Academic Collaborations

#60

Post by fatjokes » Fri, Jul 20, 2018, 6:15 PM UTC

Kyragem wrote: post #22

As usual FB and other companies only collaborates with the "elite/Ivy League" schools. IMHO there are plenty of great researchers at non-elite schools and it would have been great if FB would start collaborating with those to reduce this vast concentration of academic research that has been happening for the past decade.

It's a bit chicken-and-egg. These collaborators are also in cities where FB already has offices that they can lead teams out of. Of course, these offices were started because of the prestigious schools that were already there.