Google's Habit of Firing AI Researchers
Google fires yet another AI researcher who reportedly challenged findings
MAY 3RD, 2022 TUESDAY. 6:30 AM.
This is AiSupremacy Premium.
Special Issue: Breaking News
Something weird is going on at Google, again.
It wasn’t so long ago that Margaret Mitchell and Timnit Gebru, co-leads of the ethical AI team, were both terminated at Google. The withering email that got an ethical AI researcher fired at Google, is a most peculiar story.
Timnit Gebru really lit up Twitter when she dared to question Google last year. This time let’s try to unpiece the breaking story (in the last 24 hours).
Another AI Ethics Researcher has Been Fired
Google terminated Satrajit Chatterjee, a member of its Brain team, after he questioned the validity of a recent paper. Most recently Sat was an Engineering Leader and Machine Learning Researcher at Google AI where his work focused on fundamental questions in deep learning (such as understanding why neural networks generalize at all) and on various applications of machine learning ranging from Chip Design to new forms of Interactive Entertainment.
It’s not entirely surprising that Satrajit, it not a white male. The NYT broke the story.
The New York Times has learned Google fired machine learning scientist Satrajit Chatterjee in March, soon after it refused to publish a paper Chatterjee and others wrote challenging earlier findings that computers could design some chip components more effectively than humans.
Sat has a PhD in Computer Science from UC Berkeley, and has published in the top machine learning, design automation, and automated reasoning conferences.
The scientist was reportedly allowed to collaborate on a paper disputing those claims after he and fellow authors expressed reservations, but was dismissed after a resolution committee rejected the paper and the researchers hoped to bring the issue to CEO Sundar Pichai and Alphabet's board of directors. Sorry Sat, no can do! That’s not how Alphabet works any longer.
Sat was at Google nearly four years. Google said and maintained that the original paper had been "thoroughly vetted" and peer-reviewed, and that the study challenging the claims "did not meet our standards."
Still whatever the situation we cannot escape the reality at Google.
Another staffer who questioned research ethics is out
Google is the global leader in artificial intelligence. If its top Ethics people are forced out for speaking up, what does that say about the future of A.I. regulation at BigTech? It’s not as if this is the first time.
It’s really bizarre, Google fights with its own employees in several ways and in many questionable business decisions of the last few years. There’s a major disconnect. The exit underscores the ongoing clash between Google's management and the AI teams that drive many of its projects. It speaks to management issues and internal rebellion and discontent with Google’s own standards on A.I. and the future of ethics related to its products.
Less than two years after Google dismissed two researchers who criticized the biases built into artificial intelligence systems, the company has fired a researcher who questioned a paper it published on the abilities of a specialized type of artificial intelligence used in making computer chips. What does it all mean? In an era where OpenAI, DeepMind and others fight PR battles on the hype behind A.I., it’s clear these companies are cutting corners for glory. The headline and clickbait SEO culture Google itself is most responsible for, it’s that ironic. In short, the internet is not okay.
If there are issues inside of the most powerful A.I. company in the world, we are all in serious trouble guys!
The researcher, Satrajit Chatterjee, led a team of scientists in challenging the celebrated research paper, which appeared last year in the scientific journal Nature and said computers were able to design certain parts of a computer chip faster and better than human beings. Don’t question the dogma! The A.I. is God thesis is really important to companies like Google, Microsoft and Facebook. Recently an OpenAI employee made a comment about AGI that seemed equally outrageous:
Sorry but it’s hard to respect some of the OpenAI crew that sold out at the earliest opportunity while they were supposed to be a non-profit AI lab.
In an internal email, Chatterjee, 43, challenged some of the paper's assertions. Sorry Sat, I guess 4 years at Google and a Ph.D. just arne’t good enough to challenge much of anything. Do you think Sundar Pichai has a mind of his own, what planet are you living on? Academics is political, especially when you work a a company that violates most ethics for profit on a regular and outrageous basis.
The problems began in earnest when Google fired ethicist Timnit Gebru in 2020 following a dispute over a paper, with two others following suit. The matter escalated when the internet giant terminated Margaret Mitchell in early 2021. This isn’t a new story at Google.
Sat was humiliated by a company that doesn’t value the truth. Dr. Chatterjee, 43, was fired in March, shortly after Google told his team that it would not publish a paper that rebutted some of the claims made in Nature, said four people familiar with the situation who were not permitted to speak openly on the matter.
Google has announced a reorganization to its AI teams working on ethics and fairness, on regular intervals. In 2019, Google even canceled its A.I. ethics board. I’ve been following this story since 2018, and it’s just really bizarre. Google’s newly created AI ethics board was falling apart amid controversy over several of the board members. There seems to be a crisis of leadership and trust at Google. A lack of accountability. Good people are being thrown under the bus.
Mitchell joined Google in 2016 as a senior research scientist, according to her LinkedIn. Two years later, she helped start the ethical AI team alongside Gebru, a renowned researcher known for her work on bias in facial recognition technology. All did not end well for them, sadly. Business cuts both ways, don’t criticize the dogma. Don’t question the Public Relations.
Google confirmed in a written statement that Dr. Chatterjee had been “terminated with cause.” Like Amazon terminating people trying to make a union, Google has gotten a bit high on its horse in recent years. Google protests haven’t humanized the culture, it does appear. Which leads to even more discord.
Dr. Chatterjee’s dismissal was the latest example of discord in and around Google Brain, an A.I. research group considered to be a key to the company’s future. My last article actually featured a startup of people who ditched Google Brain to do their own thing. I guess there’s a certain sense in the air at Google Brain. Not everyone is comfortable going along with the dogma at Google. Does Google Brain have a brain drain problem in 2022?
In a statement to the Times, Google defended its research paper and practices. Don’t rock the boat. It's no secret that Google is proud of its AI research. It's keen to tout algorithms that can outperform humans in fields like chip design and cancer detection, and its flagship Pixel 6 phones are built around a custom AI-focused processor, according to Engadget’s reporting. But there is pride and there is arrogance.
It is a messy situation and another challenge to Jeff Dean’s handling of researchers who question the efficacy or pitfalls of AI. Jeff doesn’t suffer fools and outsiders lightly. When you are a heavyweight Manager for 20 years at one of the most dominant companies in the world, maybe you get that way. Or Maybe Jeff Dean just does his job, and what is right for Google’s own self-interests (fuck the truth).
You could make the argument that any flaws could significantly hinder Google's business, not to mention tarnish its reputation as a leader in AI development. Not to mention it would look bad on Jeff Dean himself. Long-time Googlers likely surgically removed ethics from profiteering a long time ago. That maybe had to do this as some Silicon Valley imperative of how early Silicon Valley consolidated itself.
A lawyer for Chatterjee, however, said his client was only trying to maintain "scientific integrity." If you are a Researcher at an A.I. lab owned by a corporation, do you think your scientific integrity will be upheld? Not apparently, in all cases. There are pros and cons for standing up for what you believe in.
In December 2020, Mitchell and Gebru were working on a paper about the dangers of large language processing models when Megan Kacholia, vice president of Google Brain, asked that the article be retracted. Gebru pushed back, saying the company needed to be more open about why the research wasn’t acceptable. Shortly afterwards, she was fired, though Google characterized her departure as a resignation, according to the Verge’s coverage.
A Pattern of Behavior at Google
With a few firings and departures at Google, there is definately a pattern here. Timnit has turned it into a career pivot. She launched DAIR, her new A.I. ethics institute.
DAIR was founded in December, 2021. Distributed Artificial Intelligence Research institute (DAIR) — an independent, community-rooted institute set to counter Big Tech’s pervasive influence on the research, development and deployment of AI.
Hold your head up high Sat, not everyone has the guts to call out Google. The NYT is really kind to Alphabet, when it says:
After spending billions of dollars to hire top researchers and create new kinds of computer automation, Google has struggled with a wide variety of complaints about how it builds, uses and portrays those technologies.
It is somehow totally ironic. Dissidents won’t be tolerated! Thanks for reading guys, till next time! Long live the dogma of A.I. hype that matters more than the integrity of an individual or the scientific gaps in a paper.