The controversy behind a star Google AI researcher’s departure

Google’s corporate culture is once again embroiled in controversy.

AI ethics researcher Timnit Gebru – a well-respected pioneer in her field and one of the few black women to be found in the industry – said earlier this week that Google fired her after blocking publication of his research on biases in AI systems. Days before Gebru left, she sent a scathing internal memo to her colleagues detailing how senior Google executives tried to crush her searches. She also criticized her department for what she described as a continuing lack of diversity among her staff.

In his widely read internal email, which was published by Platformer, Gebru said the company was “silencing in the most fundamental way possible” and claimed that “your life gets worse when you start advocating for the underrepresented” at Google.

After Gebru left, Google’s head of artificial intelligence research Jeff Dean sent a note to Grebu’s department on Thursday morning stating that his research paper did not meet the company’s publication standards after an internal review. According to Gebru, the company also told him that his critical note to his colleagues was “incompatible with the expectations of a Google official”.

A Google representative declined to comment. Gebru did not respond to a request for comment.

Gebru’s allegation of being kicked out of the powerful tech company under questionable circumstances is causing a stir in the tech and academic communities, many prominent researchers, civil rights leaders and Gebru colleagues on Google’s AI speaking publicly on Twitter in his defense. A petition to support her has already received the signatures of more than 740 Google employees and more than 1,000 academics, leaders of nonprofits and industry peers. His departure is significant as it touches on broader tensions around racial diversity in Silicon Valley as well as whether or not academics have enough freedom to publish research, even if it is controversial, while working in large companies that control the development of powerful technologies and have their own interests to consider.

What happened

People are still trying to find out exactly what led to the departure of Gebru from Google.

What we do know is that Gebru and several of his colleagues were planning to present a research paper at an upcoming academic conference on unintended consequences in natural language processing systems, which are the tools used in the field. computer science to understand and automate the creation of entries. words and audio. Research by Gebru and his colleagues, according to the New York Times, “flaws identified in a new generation of language technology, including a system developed by Google that underpins the company’s search engine. “He would also discussed environmental consequences large-scale computer systems used to power natural language processing programs.

As part of Google’s process, Gebru submitted the article to Google for internal review before it was published more widely. Google determined the document fell short of its standards because it “ignored too much relevant research,” according to Dean’s memo sent Thursday.

Dean also said in his memo that Google rejected Gebru’s article for publication because it submitted it one day before the publication deadline instead of the required two weeks.

Gebru requested further discussion with Google before removing the article, according to the Times. If Google couldn’t address her concerns, Gebru said she would resign from the company.

Google told Gebru that he could not meet his conditions and that the company was accepting his resignation immediately.

It’s a standard process for a company like Google to review employee searches before they are published outside of them. But former colleagues and outside industry researchers defending Gebru have questioned whether Google is arbitrarily enforcing its rules more strictly in this scenario.

“It just seems strange that someone who has had books written about her, who is quoted and cited daily, would be fired because an article was not properly edited,” said Rumman Chowdhury, a data scientist. head of responsible AI at Accenture Applied Intelligence and has now started his own company called Parity. Chowdhury has no affiliation with Google.

The conflict and Grebu’s dismissal reflect a growing tension between researchers studying the ethics of AI and the big tech companies that employ them.

It’s also another example of deep and enduring issues dividing part of Google’s workforce – on Wednesday alone, the National Labor Relations Board filed a complaint that Google had spied on its employees and probably raped the right of work. when he fired two activists employed Last year.

After several years of internal unrest at Google among its staff over issues ranging from its controversial plans to work with the U.S. military to sexual harassment of its employees, the past few months have been relatively calm. Instead, the company’s biggest public pressure has come from an antitrust legal review and unproven accusations by Republican lawmakers that Google’s products display anti-conservative bias. But Gebru’s case and the recent NLRB complaint show that the company is still waging internal battles.

“What Timnit has done is present rigorous but important assessments of how the company’s efforts are going with diversity and inclusion initiatives and how to fix that,” said Laurence Berland, a former engineer. from Google who was fired after organizing her colleagues around a worker. problems and is one of the employees contesting his termination with the NLRB. “It was passionate, but it wasn’t just non-constructive,” he said.

Why Gebru’s departure is important

In the relatively new and developing field of ethical AI, Gebru is not only a foundational researcher, but a role model for many young academics. She is also the head of key groups such as Black in AI, which promote greater diversity in the largely white field dominated by males of AI in the United States.

(Although Google does not list its demographics specifically for its artificial intelligence research department, it does share its diversity figures every year. Only 24.7% of its technical workforce are women and 2.4% are of Blacks, according to its 2020 Diversity and Inclusion Report.)

“Timnit is a pioneer. She is one of the founders of responsible and ethical artificial intelligence, ”said Chowdhury. “Computer scientists and engineers are entering the field thanks to her.”

In 2018, Gebru and fellow researcher Joy Buolamwini published groundbreaking research showing facial recognition software identified people and women with darker skin, at much higher rates than people and men with lighter skin.

Her work has contributed to a wider consideration in the tech industry of the unintended consequences of AI that is formed on datasets that can marginalize minorities and women, reinforcing existing societal inequalities.

AI academics outside of Google fear that Gebru’s dismissal could scare other researchers away from publishing important research that could work on their employers’ toes.

“Researchers don’t know how they’re going to continue doing this work in the industry,” said Moritz Hardt, professor of computer science at UC Berkeley, who specializes in machine learning and has studied equity in IA. “It’s a scary time I would say.