Google is still firing AI researchers amid criticism of its work. The New York Times learned that Google fired Satrajit Chatterjee machine learning exercises in March shortly after refusing to publish an article, and Chatterjee and others wrote disputing earlier findings that computers could design some components of the chip more efficiently than humans. The scientist was reportedly allowed to collaborate on a document challenging the allegations after he and fellow authors expressed reservations, but was fired after a restructuring committee rejected the document and researchers hoped to refer the matter to the CEO. Sundar Pichai and the board of directors of Alphabet.
The company did not specify why it fired Chatterjee, but told the Times it was “terminated for a reason”. He also claims that the original document was “thoroughly reviewed” and reviewed by partners, and that the study challenging the allegations “does not meet our standards.”
Whether science lasts or not, the outcome highlights the ongoing clash between Google’s management and the AI teams that run many of its projects. The problems began in earnest when Google fired ethicist Timnit Gebru in 2020 after a document dispute, with two others following suit. The issue escalated when the internet giant fired Margaret Mitchell in early 2021. While Google claims Mitchell violated data privacy rules, it also publicly criticized executives and the decision to fire Gebru. These layoffs have forced veteran engineers to leave the company and demand reforms, and this does not include company-wide protests against military contracts. Google is investigating Gebru’s dismissal and has promised changes, but Chatterjee’s dismissal may not ease tensions.
It’s no secret that Google is proud of its AI research. He wants to promote algorithms that can outperform people in areas such as chip design and cancer detection, and his flagship Pixel 6 phones are built around a custom processor focused on AI. Any flaws could significantly hamper Google’s business, not to mention tarnish its reputation as a leader in AI development.
All products recommended by Engadget are selected by our editorial team, regardless of our parent company. Some of our stories involve partnerships. If you buy something through one of these links, we can earn an affiliate commission.
Add Comment