In a sign of continued blowback from Google’s controversial ousting of two top artificial intelligence leaders, a researcher just publicly turned down a major grant from the company.
Late last year, Luke Stark, an assistant professor at the University of Western Ontario researching the social and ethical impacts of artificial intelligence, applied for a Google Research Scholar award.
Each year, the company offers grants to early-career professors pursuing topics relevant to Google’s fields of interest. Stark applied with plans to put any funding towards his further research into how technology such as mood-tracking apps and facial recognition are used to monitor human emotions.
“My impression was that Google was really pulling together a top ethical AI team,” he told Insider. “There’s always a question about working with tech firms but I felt relatively comfortable that the company was moving in a good, safe direction.”
However, Stark recalls that about two or three days after he sent in the application, Dr. Timnit Gebru, co-lead of Google’s Ethical AI team, announced she had been fired from Google. He forgot about the grant until last week, he said, when Google emailed him with what should have been good news: He was being offered a $60,000 grant for his work.
He turned it down.
Stark shared the rejection email he sent to Google on Twitter:
“When I applied for this award several months ago, I saw Google’s Ethical AI team as a global leader — it was led by two world-renowned experts in Drs. Timnit Gebru and Margaret Mitchell,” he wrote. “Four months later, Google has fired Drs. Gebru and Mitchell, in the process disrupting the lives and work of these researchers and the members of the Ethical AI team they manage.”
As a result, Stark would not accept the money:
“I’m declining this award in solidarity with Drs. Gebru and Mitchell, their teammates, and all those who’ve been in similar situations,” he wrote, “But I have enormous respect for the great work the many talented and principled folks at Google Research do.”
Stark’s boycott of Google’s grant money is the latest fallout from the ousting of Gebru last December (the firm continues to claim it did not fire her) and, three months later, the firing her co-lead Mitchell for what it claimed were “multiple violations” of its code of ethics.
Two other academics have boycotted a Google research workshop in solidarity with Gebru and Mitchell, Wired reported earlier this week, and a third academic who had previously taken funding from Google also said he would no longer apply for support.
Meanwhile, FAcct, a major AI ethics conference, also suspended its sponsorship relationship with Google after Mitchell’s firing, VentureBeat reported earlier this month.
The money Google offered Stark would have gone towards hiring students as research assistants, supporting events and conferences, and the other usual costs associated with big research projects, he said.
“My first reaction was that I had to turn it down, but I also had to reflect and talk to people and consider it,” he said. “And lots of good can be done with that money to support these questions around the ethical and social consequences of AI.”
He also acknowledges he’s in a privileged position to be able to say no: “This was my decision. I don’t begrudge people who would take it. I think there are a lot of people who can’t afford to turn down money like this.”
A Google spokesperson did not respond to Insider’s request for comment.
Gebru’s departure in December has caused a crisis for Google’s Research group, which is overseen by top executive and AI chief Jeff Dean. Gebru — a rare Black woman in a senior role at Google — said she was fired over a dispute regarding a paper highlighting bias in AI, as well as an email she had sent to an employee resource group airing her frustration with Google’s management.
Following her exit, Margaret Mitchell, her co-lead on the Ethical AI team, searched through corporate correspondence to find evidence that Google had discriminated against Gebru, sources said. Google locked Mitchell out of her corporate account in response. A month later the company fired her.
The events have caused blowback internally, particularly from the Ethical AI team, who sent a list of demands to management. Google recently reorganized its Research org to put employees working on responsible AI under a new lead. The move did little to quell anger among some employees, Insider previously reported.
“I don’t think this is going to blow over,” said Stark.”We engage in good faith with these companies, but increasingly the business models and corporate decisions these companies are pursuing doesn’t allow the good work to take root.”
Top talent in the AI field should scrutinize whether — and how — tech firms are handling issues of social and ethical responsibility, Stark said: “I think there’s a sense across multiple areas of computer science that these social and ethical questions are important and companies want to talk the talk about being responsible, but they need to walk the walk.”