Timnit Gebru: Tensions in Google’s ethical AI group are mounting as it sends demands to Sundar Pichai

The letter, sent to CEO Sundar Pichai, is entitled “The Future of Ethical AI at Google”. It describes the steps the researchers want the company to take to “rebuild trust” with the team and create an environment in which to continue their work. The team researches the ethical implications of artificial intelligence and advises the company on AI policies and products.

It comes in the aftermath of the sudden departure of Timnit Gebru, who was the co-leader of Google’s Ethical AI team until early December. She was also one of the few black employees at the company overall (3.7% of Google employees are black, according to the company’s 2020 annual diversity report) – let alone in the AI ​​division. The research scientist is also a co-founder of the group Black in AI, which aims to increase the representation of black people in the field.
Google to investigate the departure of a leading AI ethics researcher

The letter’s demands include the removal of a Google vice president, Megan Kacholia, from the team’s management chain; for transparency around Gebru’s departure; and to Google’s head of AI, Jeff Dean and Kacholia, to apologize for the way Gebru was treated. It also calls on the company to offer Gebru a new higher-level position, to publicly commit to the integrity of the research being conducted at Google, and to ensure that it does not harm employees working for Gebru plead.

In a message to Pichai and others related to the letter, Alex Hanna, a senior research scientist with the Ethical AI team, said it was sent on behalf of her team. A source familiar with the team said it has about a dozen members.

First reported by Bloomberg, CNN Business also received a copy of the letter and confirmed its authenticity. Google did not immediately respond to a request for comment.
On Wednesday December 2, Gebru tweeted that she was “fired immediately” for an email she recently sent to Google Brain Women and Allies internal mailing list. In the email, she expressed dismay at the ongoing lack of diversity within the company and her frustration with an internal process related to the assessment of an unpublished research paper that she was co-authored.
In later tweetsGebru explained that no one at Google had explicitly told her she was fired. She previously said Google would not meet some of her conditions to return and immediately accepted her resignation because she felt her email reflected “ behavior inconsistent with a Google manager’s expectations. ”

Gebru’s sudden departure soon angered thousands of company employees and others in the engineering and academic world.

Google said on Wednesday December 9 that it would investigate Gebru’s departure. In a memo sent to Google (GOOG)employees, Pichai wrote that the company should “assess the circumstances” that led Gebru to leave Google last week, and investigate “where we could have improved and conducted a more respectful process.”

“We will begin with a review of what happened to identify all the points where we can learn – taking into account everything from de-escalation strategies to new processes we can put in place,” he wrote in a memo released by CNN Business as authentically confirmed. .

Gebru is known for her research on bias and inequality in AI, and in particular for a 2018 paper she co-authored, highlighting how poorly commercial facial recognition software fared at classifying women and people of color. Their work sparked a widespread awareness of problems common to AI today, especially when technology is tasked with identifying something about humans.