Main image of article Another Google AI Researcher Quits, Deepening Unit's Talent Crisis

The shakeup within Google’s artificial intelligence (A.I.) division continues, with researcher (and renowned computer scientist) Samy Bengio announcing his resignation. His decision to step down is widely viewed as part of the continuing fallout over the departure of Google A.I. ethics researcher Timnit Gebru, who left over controversial circumstances.

In December 2020, Gebru, who was co-lead of Google’s Ethical Artificial Intelligence Team, claimed that Google’s management wanted her to take her name off a paper that pointed out how a large-scale A.I. language model (LLM) can generate biased results. When she refused, she claimed, management fired her. (Google executives subsequently insisted that she resigned, which she denied.)

Gebru is also the co-founder of Black in A.I., which aims to diversify the A.I. field, and her departure ignited an emotional conversation over bias, discrimination, and diversity within tech. In a bid to cool the situation, Google CEO Sundar Pichai issued a public apology, while sidestepping whether she was actually fired or not. “I’ve heard the reaction to Dr. Gebru’s departure loud and clear: it seeded doubts and led some in our community to question their place at Google,” he wrote in an internal memo. “I want to say how sorry I am for that, and I accept the responsibility of working to restore your trust.”

However, the anger continued, with a petition of support for Gebru on the Google Walkout for Real Change’s Medium page drawing thousands of signatures. In addition, prominent Google engineers quit in response to the controversy. A few weeks later, Google announced plans to restructure its ethical artificial intelligence (A.I.) teams with an aim toward stability. Those teams will report to “Marian Croak, a Black Google executive who currently serves as a vice president of engineering focused on site-reliability matters,” according to Bloomberg, with Croak reporting to Jeff Dean, who heads up Google AI. 

The Verge describes Bengio as a “strong advocate” of Gebru and Margaret Mitchell, another Google researcher who was terminated for allegedly scanning emails to find evidence of discrimination against Gebru. Although his farewell note to his fellow Google staffers reportedly made no mention of Mitchell or Gebru, he had expressed solidarity with the latter in past messages.

Bengio’s last day in the office is April 28. In the meantime, Google faces a severe issue with talent retention on its A.I. team—just when having a strong A.I. team is more important than ever