peshkov - stock.adobe.com
Prominent Google Brain research manager Samy Bengio resigned from Google after months of internal rifts between Google executives and AI researchers.
His resignation comes after the terminations of Google Ethical AI co-leads Timnit Gebru and Margaret Mitchell.
An AI reckoning
"The AI industry is facing a reckoning," said Alan Pelz-Sharpe, founder and principal analyst at Deep Analysis.
Artificial intelligence has made unchecked advances in an unregulated environment, and while technically impressive, many of these advances impinge upon the public's privacy and raise ethical concerns, he continued.
Governments and the public are realizing that needs to change, however, Pelz-Sharpe said.
"Questions such as whether an AI system is biased or explainable, for example, and who to hold responsible if they are biased and discriminate or simply get things wrong are being asked," Pelz-Sharpe said.
Even so, he added that AI technology companies are motivated by growth, which can be at odds with designing AI systems that respect human rights, fairness and diversity.
Changes at Google
Google fired Gebru, a well-respected AI ethics researcher, in December. The firing sparked questions about how Google treats employees who are women and people of color, and drew condemnation from hundreds of academics, AI researchers and Google employees.
Gebru alleged that her termination, in part, was due to her criticism of the way Google uses AI language tools. Google rejected a paper Gebru had written about the tool's dangers, including bias in the system and its ability to help spread hate speech, before her termination.
Google fired Mitchell in February after locking her accounts down soon after Gebru's ousting. Mitchell was supposedly looking for correspondence concerning Gebru's firing before she was terminated.
Alan Pelz-SharpeFounder and principal analyst, Deep Analysis
Bengio, who oversaw the AI ethics group Gebru and Mitchell co-led, conveyed his support for Gebru on Facebook days after she left Google, and noted that also he stands by the rest of his research team.
"When we see firms like Google that are undoubtedly a major player and influencer in the sector seemingly in a state of disconnect with their own AI ethics team, we may well be seeing signs of a reckoning and even a potential standoff coming," Pelz-Sharpe said.
Bengio's last day at Google will be April 28, according to an internal email obtained by Bloomberg. He joined the company in 2007 and is a co-founder of Google Brain, a large group of machine learning and deep learning researchers at Google.
Bengio's responsibilities within Google's AI ethics group were cut in February after Google reorganized the group's structure and placed it under vice president Marian Croak's authority. In a YouTube video announcing her appointment in February, Croak did not address any of the controversies.
Google confirmed Bengio's resignation, which was first reported by Bloomberg on April 6, but declined to comment further.