Google fired an AI researcher who known as into query a well known paper the corporate printed in 2020, a 3rd high-profile termination in lower than two years, though the corporate says his removing was for different causes.
Satrajit Chatterjee was fired in March after he and his crew challenged a Google analysis paper about using synthetic intelligence to make laptop chips, in response to The New York Occasions. The analysis paper, which was printed in Nature, a prestigious science journal, mentioned computer systems might design elements of chips sooner than people.
In an inside electronic mail, Chatterjee, 43, challenged a number of the paper’s assertions, in response to the Occasions. He additionally questioned whether or not the expertise had been absolutely examined, the paper reported.
In an announcement to the Occasions, Google defended its analysis paper and practices.
“We totally vetted the unique Nature paper and stand by the peer-reviewed outcomes,” mentioned Zoubin Ghahramani, a vp at Google Analysis within the assertion to the Occasions. “We additionally rigorously investigated the technical claims of a subsequent submission, and it didn’t meet our requirements for publication.”
A tweet by Occasions reporter Daisuke Wakayabashi made point out of one of many paper’s writer’s being harassed by Chatterjee.
The superhuman claims had been made in a Nature paper final yr. Google mentioned rebuttal paper didn’t meet its requirements. One in all Nature paper’s authors mentioned fired researcher harassed her and impugned the work. The fired researcher’s lawyer mentioned he was sustaining scientific integrity.
— Daisuke Wakabayashi (@daiwaka) May 2, 2022
Google didn’t reveal any further particulars or context relating to the character of this harassment, however mentioned he was fired for trigger.
“We’re dedicated to creating an inclusive analysis surroundings, so we firmly uphold our commonplace for respectful discourse amongst our researchers,” a Google consultant mentioned in an announcement.
Chatterjee’s dismissal follows two high-profile firings inside Google AI. In December 2020, the Google unit fired Timnit Gebru, one of many leaders of its Moral AI crew after she printed a paper figuring out the risks of bias in AI language fashions. Two months later, Margaret Mitchell, who co-led the Moral AI crew, was fired for violating the corporate’s code of conduct for reportedly utilizing automated software program to scan her messages to search out discriminatory remedy of Gebru.
Google has mentioned Gebru printed the paper with out going via the correct channels.
The search big has made its AI analysis central to the corporate’s popularity for innovation. Alphabet CEO Sundar Pichai has mentioned AI shall be extra vital than electrical energy or hearth. The corporate has built-in AI fashions in its core search product as a solution to give customers higher outcomes whereas decreasing spam.
Google rejected a paper Chatterjee and his crew wrote that rebuts the unique paper, in response to the Occasions. The researchers tried to escalate the difficulty to Pichai and Alphabet’s board, the paper reported.
“In keeping with our AI Rules, we additionally rigorously investigated the technical claims of a subsequent submission, and it didn’t meet our requirements for publication,” mentioned Ghahramani. He went on to say that Google Analysis is rising and plans to rent extra this yr than final.
