Tensions in Google's ethical AI group increase as it sends demands to CEO

Re:Tensions in Google's ethical AI group increase as it send... by fatbird on Thursday December 17, 2020 @10:41AM (#25456540)

The only evidence of toxicity was an email she sent after being forced to retract a paper that she wouldn't have to retract under proper academic circumstances, based on feedback they weren't even going to provide to her.

That's not toxic, that's anger. One angry incident, justified or not, does not make the case for firing a "toxic" employee.

[ Parent | on HN ]

Re:Tensions in Google's ethical AI group increase as it send... by room500 on Thursday December 17, 2020 @11:14AM (#25456909)

I disagree. There were several incidents where she appears toxic (and we only know about the ones made public).

1. Absolutely refusing to update her paper with feedback from a colleague - the whole purpose of the internal review

2. Demanding that Google reveal who gave her critical feedback - presumably to attack them in some way

3. Sending an e-mail to her superiors/colleagues, making ultimatums and threatening to leave if her demands were not met

4. Sending a message to a group at Google, telling them to abandon any attempts to help Google with diversity initiatives and instead start attacking their co-workers

5. Publicly blaming Google and saying that she was "fired" for an e-mail. She was relying on information assymetry, knowing that Google would be hesitant to respond fully to all allegations

[ Parent | on HN ]

Re:Tensions in Google's ethical AI group increase as it send... by thu2111 on Thursday December 17, 2020 @11:39AM (#25457177)

You forgot publicly calling both of the best known researchers in the AI revolution racist: Yann LeCunn and Jeff Dean (her manager's manager). And suing her own employer for not being feminist enough, or something.

[ Parent | on HN ]

Re:Tensions in Google's ethical AI group increase as it send... by Aunche on Thursday December 17, 2020 @11:07AM (#25456825)

She also doxxed an HRBP despite having no evidence she was involved during Gebru's firing. I do think she comes off as more passionate than "toxic." a lot of otherwise good people say regrettable things. My issue is that activists will take her accusations of racism at face value, but will dismiss anything outright ridiculous as irrelevant. You can't have it both ways.

[ Parent | on HN ]

Re:Tensions in Google's ethical AI group increase as it send... by xenihn on Thursday December 17, 2020 @10:49AM (#25456607)

>The only evidence of toxicity was an email she sent after being forced to retract a paper that she wouldn't have to retract under proper academic circumstances, based on feedback they weren't even going to provide to her.

What about the paper trail she left on Twitter?

[ Parent | on HN ]

Re:Tensions in Google's ethical AI group increase as it send... by fatbird on Thursday December 17, 2020 @10:53AM (#25456655)

That happened after she was fired. You can't cite that behaviour as a reason for firing her. Violates causality.

[ Parent | on HN ]

Re:Tensions in Google's ethical AI group increase as it send... by dekhn on Thursday December 17, 2020 @11:27AM (#25457038)

I went back and looked at Timnit's twitter history. She has a history of calling out her bosses and saying outrageous things, from well before this incident. People have linked, for example, a tweet where she called out Jeff Dean based on things he said in an internal meeting, and he had to do a bunch of apologizing and backtracking. I can't see how an employee can remain viable after doing that.

[ Parent | on HN ]

Re:Tensions in Google's ethical AI group increase as it send... by xenihn on Thursday December 17, 2020 @10:56AM (#25456687)

Her spat with the Facebook guy was before that.

[ Parent | on HN ]

Re:Tensions in Google's ethical AI group increase as it send... by fatbird on Thursday December 17, 2020 @11:07AM (#25456823)

That was a public argument on her core area of research. Google literally hired her to do that work. Being professionally assertive isn't toxic.

[ Parent | on HN ]

Re:Tensions in Google's ethical AI group increase as it send... by thu2111 on Thursday December 17, 2020 @11:43AM (#25457227)

"I’m sick of this framing. Tired of it. Many people have tried to explain, many scholars. Listen to us. You can’t just reduce harms caused by ML to dataset bias"

That's not being professionally assertive.

[ Parent | on HN ]