Live data from Hacker News

AI researcher Timnit Gebru resigns from Google

platformer.news

411–420 of 926 posts

Re: AI researcher Timnit Gebru resigns from Google

#411

It looks like the paper was critical of large language models: https://www.fastcompany.com/90582375/timnit-gebru-google-fir... The original article has been updated with Jeff Dean's response, while a member of Gebru's team (who still works at Google Brain) states that Dean's email is misleading and inaccurate: https://twitter.com/alexhanna/status/1334579764573691904 As I understand it: The paper had already been subm…

From Jeff Dean's email:

> Timnit co-authored a paper with four fellow Googlers as well as some external collaborators that needed to go through our review process (as is the case with all externally submitted papers). We’ve approved dozens of papers that Timnit and/or the other Googlers have authored and then published, but as you know, papers often require changes during the internal review process (or are even deemed unsuitable for submission). Unfortunately, this particular paper was only shared with a day’s notice before its deadline — we require two weeks for this sort of review — and then instead of awaiting reviewer feedback, it was approved for submission and submitted.

> A cross functional team then reviewed the paper as part of our regular process and the authors were informed that it didn’t meet our bar for publication and were given feedback about why. It ignored too much relevant research — for example, it talked about the environmental impact of large models, but disregarded subsequent research showing much greater efficiencies. Similarly, it raised concerns about bias in language models, but didn’t take into account recent research to mitigate these issues. We acknowledge that the authors were extremely disappointed with the decision that Megan and I ultimately made, especially as they’d already submitted the paper.

> Timnit responded with an email requiring that a number of conditions be met in order for her to continue working at Google, including revealing the identities of every person who Megan and I had spoken to and consulted as part of the review of the paper and the exact feedback. Timnit wrote that if we didn’t meet these demands, she would leave Google and work on an end date. We accept and respect her decision to resign from Google.

Re: AI researcher Timnit Gebru resigns from Google

#412
post #56

I don't think "fired" is the right choice of words -- she gave an ultimatum offering to resign, and they accepted her resignation. Is accepting a resignation the same as firing someone? > Thanks for making your conditions clear. We cannot agree to #1 and #2 as you are requesting. We respect your decision to leave Google as a result, and we are accepting your resignation. https://twitter.com/timnitGebru/status/1334364…

Have “#1 and #2” been published? I don’t see them at the link above or anywhere else.

One of them is in the updated article with Jeff Dean's response:

> Timnit responded with an email requiring that a number of conditions be met in order for her to continue working at Google, including revealing the identities of every person who Megan and I had spoken to and consulted as part of the review of the paper and the exact feedback.

Re: AI researcher Timnit Gebru resigns from Google

#413

It looks like the paper was critical of large language models: https://www.fastcompany.com/90582375/timnit-gebru-google-fir... The original article has been updated with Jeff Dean's response, while a member of Gebru's team (who still works at Google Brain) states that Dean's email is misleading and inaccurate: https://twitter.com/alexhanna/status/1334579764573691904 As I understand it: The paper had already been subm…

This. Failure to cite relevant research is a reason to reject a paper from a conference but that's not the job of an internal review. Research independence is fundamental and why professors are given tenure. Sure companies are companies and don't play by those rules, but you can't claim to be making a fair academic criticism if you're not adhering to academic norms. If the paper was bad it should have been rejected b…

>> Failure to cite relevant research is a reason to reject a paper from a conference but that's not the job of an internal review.

The done thing is generally for the reviewers to suggest additional references to improve the paper. If any mention of related work is entirely missing, then that's another matter, but a reviewer thinking that the authors should cite a particular piece of work (e.g. the reviewer's own paper) is not normally something that could lead to outright rejection.

Re: AI researcher Timnit Gebru resigns from Google

#414
post #348

Earlier quoted context omitted.

> I read her email. Is this now posted somewhere publicly? I feel like most of this conversation is pretty speculative until the commentariat (self included) can read it! EDIT: duh

It’s in the article?

This is not the email in which she offers a resignation ultimatum, an exchange referred to in her Twitter thread as occurring between Timnit, (presumably) her boss (Megan?), and possibly others.

Re: AI researcher Timnit Gebru resigns from Google

#415
post #336

Earlier quoted context omitted.

The negative publicity is a result of Google only providing feedback on her research paper, and forcing its retraction, through an HR mediated and anonymized process. That is toxic.

I don't understand this argument. Why do the providers of the feedback need to be unmasked? So they can be disregarded because of their race or gender or "position of privilege"? From what I can tell, she doesn't seem to address the points raised at all and instead complains they won't tell her exactly who said what.

The anonymity of the reviewers might be somewhat detrimental, but it might not.

https://www.nature.com/articles/nn0399_197

I think the process being opaque is more detrimental, because there's no way for anyone outside of the process to know it happens. If it wasn't opaque, then everyone could see when it happens.

Not being able to engage in a debate about the merits of something is also sketchy. It seems like an order being dressed-up as peer feedback.

Re: AI researcher Timnit Gebru resigns from Google

#416

Earlier quoted context omitted.

Her point is more than that. I think that she was trying to say that ML systems have an inherent bias that is independent of their data. This is a consequence of the fact that it's not possible to do generalization, and thus learning, without bias to begin with. What an AI researcher does when deciding which architecture is essentially tuning the bias in the generalization system in order to produce better results, w…

> Her point is more than that. I think that she was trying to say that ML systems have an inherent bias that is independent of their data. In general I take these types of tropes to be cloaked (politically correct) way of saying "fuck x group".

I'm not sure I understood your comment correctly, so I'll apologize in advance if I misunderstood.

That statement is contained in a paper I linked that dates from 1980. I'm pretty sure it's factually correct, because I'm trying to make nothing but a demonstration that an idée reçue, that ML systems are only based due to their data, is incorrect.

There's really no cloaked message behind it, and I'm not sure why one would think there is (hence the disclaimer). I think it's a pretty uncharitable interpretation.

Re: AI researcher Timnit Gebru resigns from Google

#417
post #389

Earlier quoted context omitted.

> so... I guess you're tired of it? I'm not. Many other rhetorical moves have a lot of politics to muddle through / implicit assumptions that can be disavowed in bad faith, etc. I don't like to engage in that kind of discussion at all. But the "I'm so tired" bit is both condescending (and it takes... something... to condescend to Yan LeCunn) and evasive. The end result is authoritarian -- look man, this is decided an…

Yann was (actively or unintentionally) refusing to participate in the same conversation. His response was the "facts don't care about feelings" one. It's not wrong but it is a thought terminating cliche.

It is. Yann also seems blissfully unaware of how sampling bias issues have always been dealt with (eg the Heckman correction model) in traditional statistics, where “getting more data” is not an option.

Re: AI researcher Timnit Gebru resigns from Google

#418
I remember Timnit from that Twitter thread a while back with LeCun on the subject of bias in a deep learning system.

In that thread Timnit established that she was an "authority" and then instead of engaging on the merits in the conversation, simply said basically, "You have to submit to me because I'm the authority" and refused to offer any information of any substance ... as though she was some kind of wizard and what she knew was just way too complicated for us mortals to understand.

Meanwhile, LeCun just very calmly engaged on the specific technical details about bias in Deeplearning systems.

After that, Twitter just piled on LeCun and claimed he was being a "mansplainer" ... as though saying anything technical to a black woman is "mansplaining".

As an engineer what would you seek to research if you wanted to reduce the bias in specific AI systems? You'd figure out how to minimize bias in the algorithms and in the data, and identify any other sources that could introduce bias, and figure out how to put controls on developers and data collectors to prevent bias ... right?

Ok, now go look at what is actually discussed in AI bias papers.

Sometimes I feel the emperor has no clothes and everyone just pretends like vacuous bullshit is something profound because if you criticize the papers you'll be fired or blacklisted in Silicon Valley.

Re: AI researcher Timnit Gebru resigns from Google

#419

I'm very confused by all of this. I've read the tweets and this article and can't make heads or tails of what is going on. What am I missing?

Gebru wrote a research paper and circulated it for feedback. Google managers said "retract the paper because we say so." Gebru responded with "I'd rather resign (and also you suck)." Google "accepted" the resignation (i.e. fired her).

I was able to discern this, but it doesn't explain why they wanted her to retract the paper. The contents of the paper and why they want her to retract it are key to the story, otherwise it's he-said-she-said.

Re: AI researcher Timnit Gebru resigns from Google

#420

Earlier quoted context omitted.

> so... I guess you're tired of it? I'm not. Many other rhetorical moves have a lot of politics to muddle through / implicit assumptions that can be disavowed in bad faith, etc. I don't like to engage in that kind of discussion at all. But the "I'm so tired" bit is both condescending (and it takes... something... to condescend to Yan LeCunn) and evasive. The end result is authoritarian -- look man, this is decided an…

> Many other rhetorical moves have a lot of politics to muddle through / implicit assumptions that can be disavowed in bad faith, etc. I don't like to engage in that kind of discussion at all. But isn’t what you described so vague and subjective, that you can apply it to pretty much anything? IMHO it’s more honest to say “I sit out of conversations I don’t like”, which is an entirely fair and understandable decision…

It’s hard to like something that makes you tired when it’s not one of your kids.
Post reply on HN