News for hackers. Stuff that matters.
Re:AI researcher Timnit Gebru resigns from Google by tkfu on Thursday December 03, 2020 @02:21PM (#25293295)
Ok, but how do you do that? It was easy in classical music; they just started doing auditions behind a curtain, without any verbal Q&A, and suddenly a field that used to be dominated by men became pretty much 50/50, and all the people who said that men were better at the highest echelons of musical achievement had to shut up.

But tech companies want to talk to the humans they're hiring, so race and gender are (almost) always available as information inputs during the hiring process. There's no equivalent to the curtain to remove the possibility of bias, so you need to look at alternate methods. Setting hiring targets is one way to do it. If you're going to complain that the targets are unfair, you can't just say "Well shucks, guess we have to stick with the system that produces massively unequal outcomes." You've gotta propose a system that's less unfair than the status quo.

Re:AI researcher Timnit Gebru resigns from Google by inglor_cz on Thursday December 03, 2020 @04:05PM (#25294909)
"It was easy in classical music; they just started doing auditions behind a curtain, without any verbal Q&A, and suddenly a field that used to be dominated by men became pretty much 50/50"

Be careful here, the wind has changed and blind auditions are now problematic.

https://www.nytimes.com/2020/07/16/arts/music/blind-audition...

Re:AI researcher Timnit Gebru resigns from Google by dragonwriter on Thursday December 03, 2020 @05:17PM (#25295979)
> Be careful here, the wind has changed and blind auditions are now problematic.

The argument made in the article you cite is ludicrous, blind auditions cannot possibly hurt diversity. If there is a problem when blind auditions are used in hiring for professional orchestras, the problem is elsewhere in the pipeline, and the way to fix it isn't replacing blind auditions, its fixing issue further up the pipeline.

There may be a pro-diversity case for using something other than current skill as a factor at other stages of the pipeline which are intended as, in whole or in part, educational and where current skill is being used not as a measure of current professional competence, but as a proxy for potential. But even there that's largely a poor substitute for identifying and addressing the factors producing a skewed pool.

Re:AI researcher Timnit Gebru resigns from Google by calf on Thursday December 03, 2020 @04:14PM (#25295057)
That's a good point; the original experiment is a good story that for many was proof that prejudice exists, but it doesn't mean blind screening is a panacea.
Re:AI researcher Timnit Gebru resigns from Google by jeffbee on Thursday December 03, 2020 @02:57PM (#25293834)
There are some fairly easy steps you can take that other, more-diverse organizations have taken. First, you have to get rid of your referral pipeline, because hiring people your employees already know is the opposite of diversity. Second, instead of recruiting from organizations you've already heard of, like Stanford where this person came from, you park a full-time recruiter at institutions with practical diversity, like Houston, Georgia, UNLV, CUNY. Finally, you go for diversity of thought, by hiring philosophers and carpenters instead of CS grads. I know for a fact that were I to hire into my AI Ethics department, the last person I would recruit would be a Stanford AI researcher. I'd go directly to the philosophy department.
Re:AI researcher Timnit Gebru resigns from Google by robertlagrant on Thursday December 03, 2020 @04:04PM (#25294906)
> First, you have to get rid of your referral pipeline, because hiring people your employees already know is the opposite of diversity

That's assuming people only know their clones.

Re:AI researcher Timnit Gebru resigns from Google by Karunamon on Thursday December 03, 2020 @02:43PM (#25293634)
>There's no equivalent to the curtain to remove the possibility of bias

Other companies have tried. Names and vitals deleted from resumes before any people involved in the hiring process see them, remote interviews conducted over IM or with voice masking. That's one benefit I see with the much-maligned leetcode-style interviews, is that your identity is completely isolated from the process until the very end.

Re:AI researcher Timnit Gebru resigns from Google by parliament32 on Thursday December 03, 2020 @03:32PM (#25294362)
>There's no equivalent to the curtain

There's no reason you can't do an interview over text-chat instead of in person -- hell, with most of us working remote for the foreseeable future, that's how the majority of on-the-job communication will happen anyway.

Re:AI researcher Timnit Gebru resigns from Google by jpab on Thursday December 03, 2020 @03:17PM (#25294143)
> But tech companies want to talk to the humans they're hiring, so race and gender are (almost) always available as information inputs during the hiring process.

This is actually kind of funny to me, because in the open source world plenty of people have extremely productive working relationships without ever seeing or hearing each other, and in many cases without even knowing what country the other person lives in, how old they are, or what their legal name is.

So while there are definitely arguments that face to face communication is "higher bandwidth" or has other advantages, it doesn't seem out of the question to me that the hiring process could be "blinded" to a similar extent to orchestra auditions, without any significant reduction in hiring accuracy.

(Ok, maybe not quite the same extent; language fluency and style of speech are still significant signals even if everything is done over text)