A leading Google artificial intelligence ethics scientist said she was fired for criticizing Alphabet Inc. controlled company diversity initiatives. Google on Thursday responded to a recent disagreement between internet giant and employee activists.
Timnit, a black man, announced on Twitter on Twitter that she was fired on Wednesday after emailing colleagues expressing dissatisfaction with gender diversity policies in Google’s AI department. She also questioned whether company executives reviewed their work more severely than people from other backgrounds. February was a co-founder of the non-profit organization Black in AI [Negros em IA], which aims to increase the representation of blacks in the field of artificial intelligence, and co – authored an important article on bias in facial analysis technology.
Google AI unit manager Jeff Dean told staff in an email to Reuters that Gebu was threatening to resign unless he knew which colleagues thought her draft work should not be published. Dean rejected the request.
“We accept and respect his decision to abandon Google,” Dean wrote in an email, adding, “We all truly share Timnita’s passion for making AI fairer and more inclusive.”
In a series of Twitter posts in February, Google said it cut it out of its systems without warning or talking to him about concerns.
The sudden departure of February adds to years of suffering, including several resignations and layoffs in the AI department and other Google organizations related to the company’s diversity, and efforts to reduce potential harm to its services.
More than 150 employees expressed support for Gebra, demanding that Google reaffirm its commitment to academic freedom and explain why it decided to “censor” her article, according to an online petition. Sherlin Ifila, president of the National Association for the Advancement of Color People’s Progress (NAACP), wrote on Twitter that Gebu’s resignation was “utterly outraged” and a “disaster.”
Also on Wednesday (2), the National Industrial Relations Council issued a complaint accusing Google of illegal surveillance and interrogation of several employees who were later fired for protesting against company policies and attempts to organize a union.
The February article argued that technology companies could do more to ensure that AI systems designed to imitate writing and human speech do not exacerbate historical gender bias and the use of offensive languages, the project sees in the project. Reuters news agency.
In an email to employees, Dean said the article was not submitted to the company for review in time and was presented at a conference without Google’s permission.
He also questioned some of his findings, which he said were based on outdated concerns, including the environmental impact of a large number of computing computers.
In response to the company’s rejection of his work, Gebra wrote on Twitter last week: “There is nothing like a bunch of privileged white men trying to squeeze marginalized research into marginalized communities with zero conversation, ordering them to stop. The level of disrespect is incredible.”
Google declined to comment on Gebra’s dismissal in addition to Dean’s email, first reported by technology news site Platformer.
February previously worked for Microsoft Research and co-authored a widely cited article in 2018 that found higher error rates in facial analysis technology for women with darker skin tones.
His new article, co-authored by people who don’t work for Google, is yet to be presented at a computer science conference in March, according to someone familiar with the issue.
Translation by Luiz Roberto Mendes Gonçalves