Artificial intelligence researcher Timnit Gebru set off a firestorm a year ago when she announced she was fired from tech giant Google for questioning the potential dangers around the emerging technology. Now, she's taking AI ethics a step further with a new research institute, according to WIRED.
She has recently founded Distributed Artificial Intelligence Research (DAIR), an organization dedicated to analyzing the impacts of AI technology while having a more inclusive workforce. Gebru, who also serves as DAIR's executive director, claims the industry is dominated by Western white males and needs to make room for other minorities and nationalities.
Gebru reportedly butted heads with Google higher-ups over a research paper focusing on bias in algorithms and text-processing technology. The tech giant, along with other major companies, quickly adopted the technology despite the risks, WIRED says.
Google claims Gebru resigned and wasn't fired. When reporters asked the company about the incident, a spokesperson declined to comment but pointed to Google's work on AI governance and "responsible innovation."
"Instead of fighting from the inside, I want to show a model for an independent institution with a different set of incentive structures," Gebru says. Her ousting has since sparked controversy around tech companies using researchers to study the implications of the technology they could profit from. Gebru claims DAIR will allow experts and researchers to question the potential threats and downsides to AI more.
Another initiative of DAIR is showing how artificial intelligence can be used in other ways. The non-profit is poised to enter the academic world with its research later this month. Researchers plan on showcasing how apartheid has affected land use in South Africa using a public data set of aerial imagery at NeurIPS, the world’s most prominent AI conference, according to reporters.
You can read more about the organization here.