Our mission to improve business is fueled by readers like you. To enjoy unlimited access to our journalism, subscribe today.
Google faced mounting criticism on Friday for its treatment of a prominent black ethics researcher on AI who says she was fired after questioning the company’s commitment to diversity in an email to colleagues.
Friday mid-morning Pacific Time over 1,400 people, including nearly 600 Google, had signed an open letter expressing his solidarity with researcher Timnit Gebru. The letter also called on Google to strengthen its commitment to research integrity and support ethical research in line with the company’s AI principles.
Gebru, who was co-lead of Google’s AI ethics team, is well known to machine learning researchers for his exploration of the racial and gender biases inherent in many AI systems. She was also one of Google AI Research’s most prominent black employees and is a co-founder of Black in AI, an organization that has sought to highlight the work of black researchers in machine learning and provide mentorship and support to black academics. in the field.
She says she was fired after writing a message to members of a “women and allies” mailing list within Google AI Research, in which she accused the company of making statements in favor of diversity and ‘inclusion.
She said her managers informed her that the message was “incompatible with the duties of a director” of the company and immediately terminated her employment.
Jeff Dean, Google’s senior vice president of AI research, told staff in an email that Gebru had offered to step down and the company had accepted his resignation.
Google declined to comment on the open letter.
Gebru and the company say her departure followed an incident in which the company asked Gebru to withdraw a research paper that she and several other Google researchers had co-authored and presented to a conference on the IA. In it, Gebru and his fellow researchers highlighted the racial prejudices inherent in a kind of State-of-the-art natural language processing AI., known as great language models, which Google and several other companies have developed. Google now uses one of these great language models to fuel parts of its search results.
In his email to staff, Dean says the article “did not meet our publication target” because “it ignored too much relevant research,” including information on how some of the issues raised by Gebru and its co-authors could be mitigated.
Gebru said the process used to examine the paper was opaque and irregular. The open letter calls on Dean and others involved in the decision to reject the document to meet with the Google AI ethics team to explain the process they used. He also calls on the company to explain to the general public what happened. And that requires Google Research “to make an unequivocal commitment to research integrity and academic freedom.”
Google said it was “determined to embed diversity, equity and inclusion in everything we do,” but the company has struggled for years to hold that position. According to the company’s 2020 Diversity Report, black employees make up only 3.4% of its U.S. workforce, up from 2.4% in 2014. The company also noted that its female attrition rate black women, who make up 1.6% of its US workforce, had increased over the past year.
Gebru’s departure is almost certain to be a blow to the company’s efforts to reverse this trend. Many, including Gebru, have emphasized Twitter that Google had in the past often viewed her and her work as examples of the progress the company is making in addressing diversity and inclusion issues. Gebru had said in his email to his colleagues that his treatment had convinced that “there is no responsibility” at Google for the real improvement of its culture.
Numerous machine learning researchers have posted messages on Twitter saying that Gebru’s exit has dealt a serious blow to Google’s reputation on issues of ethics and diversity in AI.
Joy Buolamwini, a computer scientist who founded the Algorithmic Justice League, a group dedicated to combating algorithmic bias, and who co-authored with Gebru an article influencing racial bias in facial recognition systems, said the Gebru’s dismissal “seriously undermines the credibility of Google for its support. rigorous research on the ethics of AI and algorithmic auditing. “
More to read absolutely technological coverage of Fortune:
- Robinhood’s Next Adventure: Stealing market share from the rich
- Why the power to change the founder’s double standard based on VC
- Quantum computing is enter a new dimension
- How Chinese phonemaker Xiaomi conquered India—And surpassed Apple
- Google ethics researcher leaves renews its concerns, the company silences whistleblowers