The outstanding artificial intelligence scholar Timnit Gebru helped raise Google’s public image because the company promoted the status of black computer scientists and questioned the harmful use of AI technology.
But internally, Gebru, a leader in the field of artificial intelligence ethics, is not afraid to express doubts about these promises-until this week she was expelled in a dispute about a research paper on the social dangers of the emerging sector. Out of the company. AI.
Gebru announced on Twitter that she was fired. Google told employees that she had resigned. More than 1,200 Google employees have signed an open letter, calling the incident an “unprecedented research review” and accusing the company of racism and defensiveness.
I got fired @JeffDean Send me an email to Brain Woman and Allies. My company account has been cut off. So I was fired immediately 🙂
-Timnit Gebru (@timnitGebru) December 3, 2020
The anger over Gebru’s sudden departure is a recent incident that has raised questions about whether Google has deviated from its original “Don’t Be Evil” motto, which is that the company now routinely removes employees who dare to challenge management. . The departure of the black Gebru (Gebru) also raises further doubts about diversity and inclusion. The company’s black women account for only 1.6% of the workforce.
When Google’s conclusions may threaten profits or national interests, the usefulness of ethical efforts (from this week’s White House executive order to the ethics review team established by the entire technology industry) is beyond Google’s focus.
Gebru has always been a star in the AI ethics community. She engaged in Apple product research during her early technical career and obtained a PhD in computer vision from the Stanford University Artificial Intelligence Laboratory.
She is the co-founder of the Black in AI group, an organization that promotes Black’s employment and leadership in the field. She is known for a landmark study in 2018 that found racial and gender biases in facial recognition software.
Gebru is currently working on a paper on the risks of developing a computer system that can analyze a huge database of human language and use the database to create its own human-like text. A copy of the paper has been shown to the Associated Press (Associated Press), which mentions Google’s use of new technologies in the search business and other companies developed.
In addition to pointing out the potential dangers of deviations, the paper also enumerates the environmental costs of investing a lot of energy to run the model – this is an important issue for the company’s commitment to carbon neutrality since 2007, because Yu becomes greener.
The Google manager was worried about the omissions and timing of this work, and wanted to cancel the names of the Google employees for the study, but Gebru objected based on an email exchange shared with AP and first published by Platformer.
Jeff Dean, Google’s head of artificial intelligence research, reiterated Google’s position on the research in a statement on Friday.
Dean writes that the paper presents valid points, but “there are some important gaps that prevent us from putting Google affiliation on it with peace of mind.”
“For example, it does not contain important findings on how to make the model more efficient and actually reduce the overall environmental impact, and it does not take into account some of the recent work done by Google and elsewhere in mitigating deviations,” Dean added.
On Tuesday, Gebru expressed his frustration with this process to a diverse and inclusive email group within Google. The subject was: “Do everything possible to silence the voices of marginalization.” Gebru said on Twitter. It was the e-mail that she was fired.
Dean said in an email to employees that the company accepted “her decision to resign from Google” because she told managers that if her requirements for the study were not met, she would leave.
Joy Buolamwini, a graduate researcher at Massachusetts Institute of Technology, said: “Because Timnit dared to demand research integrity, this severely damaged Google’s credibility in supporting rigorous research on AI ethics and algorithm audits. Brother
Buolamwini said in an email on Friday: “She deserves what Google doesn’t know about. Now, she has become an All-Star free agent and will continue to change the technology industry.”
How Google handles its AI ethics initiative and internal dissent caused by Gebru’s exit is one of the many issues the company faces in the new year.
While she was out, the National Labour Relations Board (National Labour Relations Board) brought attention again to Google’s workplace on Wednesday. In the complaint, the NRLB accused the company of monitoring employees during its labor union efforts in 2019, and then the company fired two human rights workers to engage in activities permitted by U.S. law. Google denied the allegations in this case, which is scheduled to open in April.
In the antitrust lawsuit, Google was also convicted by the US Justice Department as the overlord for profit, accusing the company of illegally abusing the power of its dominant search engine and other popular digital services to stifle competition. The company also denied any wrongdoing in this legal battle, which may continue for many years.
Will Apple Silicon launch an affordable MacBook in India? We discussed on the weekly technical podcast Orbital, you can subscribe via Apple Podcast, Google Podcast or RSS, download the episode, or click the play button below.