Prominent AI researcher Dr Timnit Gebru claims she was fired over email to colleagues, but Google insists she didn’t follow procedure
Google’s human resources (HR) decisions are once again the spotlight after a prominent AI researcher claimed she was fired over a disagreement about a research paper.
Some will not know Dr Timnit Gebru, but she achieved fame for her work on algorithmic bias, particularly with facial recognition technology. In 2018, she co-authored a paper that highlighted the higher error rates for identifying darker-skinned people, partly because facial recognition AI algorithms were mostly trained with datasets that were overwhelmingly white.
Google’s HR decisions are under intense scrutiny at the moment. This week the National Labor Relations Board (NLRB) filed an official complaint against the firm, allegedly because it spied on workers who were organising staff protests, and then fired two of them.
Now Gebru, who until recently was one of Google’s top artificial intelligence researchers, alleged in a tweet that she had essentially been sacked by the Alphabet unit.
“Apparently my manager’s manager sent an email my direct reports saying she accepted my resignation,” she tweeted. “I hadn’t resigned – I had asked for simple conditions first and said I would respond when I’m back from vacation. But I guess she decided for me :) that’s the lawyer speak.”
“I said here are the conditions,” she added in a subsequent tweet. “If you can meet them great I’ll take my name off this paper, if not then I can work on a last date. Then she sent an email to my direct reports saying she has accepted my resignation. So that is google for you folks. You saw it happen right here.”
The issue began when Gebru alleged that Google was trying to get her to retract a research paper, because she had co-authored a paper with four other Googlers as well as some external collaborators that needed to go through Google’s review process.
Google however alleges that this particular paper was only shared with a day’s notice before its deadline, and that it normally requires two weeks to conduct a review. But instead of awaiting reviewer feedback, the paper was approved for submission and submitted.
Gebru then emailed a Google unit (the Brain Women and Allies listserv), in which she voiced frustration that managers were trying to get her to retract the research paper.
The full text of that email was first published in Platformer.
“A week before you go out on vacation, you see a meeting pop up at 4:30pm PST on your calendar,” it reads. “Then in that meeting your manager’s manager tells you ‘it has been decided’ that you need to retract this paper by next week… You are not worth having any conversations about this, since you are not someone whose humanity (let alone expertise recognized by journalists, governments, scientists, civic organizations such as the electronic frontiers foundation etc) is acknowledged or valued in this company.”
After the email went out, Gebru reportedly told managers that certain conditions had to be met in order for her to stay at the company.
Otherwise, she would have to work on a transition plan.
But it seems that Google called her bluff, and according to the Verge, Megan Kacholia, vice president of engineering at Google Research, reportedly said she could not meet Gebru’s conditions and accepted her resignation as a result.
“However, we believe the end of your employment should happen faster than your email reflects because certain aspects of the email you sent last night to non-management employees in the brain group reflect behavior that is inconsistent with the expectations of a Google manager,” Kacholia reportedly said.
Gebru however claimed on Twitter, that the decision to fire her ultimately came from Google’s head of AI, Jeff Dean.
Dean sent an email to Google staff on Thursday morning explaining his view Gebru’s departure, which was also published by Platformer.
“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),” the email reads. “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.”
“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,” he added.
“Timnit wrote that if we didn’t meet these demands, she would leave Google and work on an end date,” he added. “We accept and respect her decision to resign from Google.”
But it seems that Gebru’s position has been supported by the Google Walkout Twitter account, which is run by current and former employees.
“We stand with @timnitGebru,” they tweeted. “We call on Google Research to strengthen its commitment to research integrity and to unequivocally commit to supporting research that honors the commitments made in Google’s AI Principles. Sign the letter…