Bust a Move, Disco King

Timnit Gebru was fired from Google in December 2020 for refusing to retract a research paper, and every single warning that paper made about large language models has now happened at a scale the industry spent 4 years trying to make people forget about.

Her name is Timnit Gebru.

She co-led the Ethical AI team at Google. She co-wrote a paper called "On the Dangers of Stochastic Parrots" with Emily Bender at the University of Washington and two other researchers. The paper was 14 pages long. It was submitted to a top AI ethics conference. And it was the reason Google decided that one of the most senior Black women in AI research could no longer work there.

The story Google told publicly was that she resigned. The story she told, confirmed by 2,695 of her colleagues in an open letter, was that she was fired by email while on vacation because she refused to either retract the paper or remove her name from it.

The paper had not even been published yet.

Here is what she actually wrote, and why every prediction inside it has now come true.

The first warning was about scale itself. Bender and Gebru argued that training ever-larger models on ever-larger scrapes of the internet would produce systems that appeared fluent but had no actual understanding of language. They called these systems stochastic parrots because they would repeat patterns from training data with statistical confidence and zero comprehension. The paper predicted that this apparent intelligence would fool both users and developers into trusting outputs that were structurally incapable of being reliable.

This was 2020. GPT-3 had just come out. The paper predicted the hallucination problem before anyone had a word for it.

The second warning was about bias amplification. The paper documented in detail that internet-scale training data contains systematic overrepresentation of dominant viewpoints and underrepresentation of marginalized ones. The models would not just absorb this bias. They would amplify it, because the optimization process rewards confident outputs, and confidence in language patterns tracks frequency in the training set.

The prediction was that hiring tools built on these models would discriminate against women. That healthcare triage tools would underperform on Black patients. That loan approval systems would entrench inequality while presenting their decisions as neutral algorithmic judgment.

Every one of those things has now been documented in deployment.

Amazon's hiring algorithm penalized resumes that contained the word "women" in any context. Healthcare risk scoring algorithms used by major US hospitals were found to systematically underestimate the medical needs of Black patients. Apple Card's credit algorithm gave wives credit lines 10x lower than their husbands for the same financial profile.

The third warning was about environmental cost. The paper calculated that training a single large language model produced emissions equivalent to the lifetime output of 5 cars. The prediction was that the race to scale would create an environmental footprint that would eventually rival entire industries.

In 2024, Google's emissions were up 48% from 2019, and the company explicitly blamed AI infrastructure. Microsoft's were up 29%, same reason. Both companies have now quietly abandoned the climate commitments they were publicly celebrating the year Gebru was fired.

The fourth warning was about documentation. The paper argued that the training datasets being assembled were too large for anyone to actually audit. Nobody at Google, OpenAI, Meta, or any other lab could tell you with confidence what was in the data their models were trained on. This was not a temporary problem to be solved later. It was a permanent feature of the approach.

In 2023, researchers discovered that the LAION-5B dataset, used to train Stable Diffusion and other major image models, contained thousands of images of child sexual abuse material. The companies that had trained on the dataset had no way of knowing. The paper predicted that category of failure 3 years before it was found.

The fifth warning was the one Google cared about most.

Bender and Gebru argued that the deployment of these systems would centralize linguistic and cultural power in the hands of the small number of companies that could afford to train them. The internet would become a place where the dominant voice was a statistical average of dominant voices, presented as a neutral assistant. Languages underrepresented in the training data would degrade over time as more web content was generated by these systems and fed back into the next training run.

This is now happening in real time. A 2024 study found that 57% of new web content in English is AI-generated or AI-assisted. Researchers studying low-resource languages have documented active degradation in translation quality, because the synthetic content fed back into training is itself worse in those languages.

The paper Google fired her for predicted the model collapse problem before model collapse had a name.

The mechanism behind why this all happened is the part of her work that nobody quotes.

Gebru's argument was not that AI is dangerous in some abstract sci-fi sense. Her argument was that AI is dangerous in a very specific structural sense. The technology was being built by a small group of researchers who shared similar backgrounds, worked at similar companies, and were rewarded for shipping products faster than competitors. The incentive structure made it impossible for safety, ethics, and bias concerns to slow anything down. Anyone inside the system who raised those concerns was either ignored, sidelined, or removed.

She was making that argument from inside Google.

Then Google proved her right by removing her.

The team Google had built to make sure their AI was safe was dismantled in 90 days because they did the job they had been hired to do. Margaret Mitchell, the other co-lead of the Ethical AI team, was fired two months after Gebru for searching through her own emails for evidence of how Gebru had been treated.

Gebru did not stop. She founded DAIR, the Distributed AI Research Institute, in 2021. The mission is to do AI research outside the control of the companies that have a financial interest in not hearing the answers.

Every prediction in the Stochastic Parrots paper has now been validated by deployment. Hallucinations are an industry-wide problem the largest labs cannot solve. Bias amplification has been documented in hiring, healthcare, lending, and criminal justice. Environmental costs are larger than entire small countries. Training data audits remain impossible. Model collapse is an active research crisis at every major lab.

The question worth sitting with is the one almost no one in the industry will say out loud.

Every researcher with the technical credibility to call out these problems watched what happened to her in December 2020 and made a calculation about their own career. The number of people willing to speak publicly about safety and ethics issues inside the major AI labs collapsed after that firing and has not recovered.

The researcher Google fired for warning about exactly what is now happening was right.

The company that fired her is now the second-largest deployer of the technology she warned about.

Sadly there is no body of recent feminist writing addressing men that is accessible, clear, and concise. There is little work done from a feminist standpoint concentrating on boyhood. No significant body of feminist writing addresses boys directly, letting them know how they can construct an identity that is not rooted in sexism. There is no body of feminist children’s literature that can serve as an alternative to patriarchal perspectives, which abound in the world of children’s books. The gender equality that many of us take for granted in our adult lives, particularly those of us who have class privilege and elite education, is simply not present in the world of children’s books or in the world of public and private education. Teachers of children see gender equality mostly in terms of ensuring that girls get to have the same privileges and rights as boys within the existing social structure; they do not see it in terms of granting boys the same rights as girls—for instance, the right to choose not to engage in aggressive or violent play, the right to play with dolls, to play dress up, to wear costumes of either gender, the right to choose. Just as it was misguided for reformist feminist thinkers to see freedom as simply women having the right to be like powerful patriarchal men (feminist women with class privilege never suggested that they wanted their lot to be like that of poor and working-class men), so was it simplistic to imagine that the liberated man would simply become a woman in drag. Yet this was the model of freedom offered men by mainstream feminist thought. Men were expected to hold on to the ideas about strength and providing for others that were a part of patriarchal thought, while dropping their investment in domination and adding an investment in emotional growth. This vision of feminist masculinity was so fraught with contradictions, it was impossible to realize. No wonder then that men who cared, who were open to change, often just gave up, falling back on the patriarchal masculinity they found so problematic. The individual men who did take on the mantle of a feminist notion of male liberation did so only to find that few women respected this shift. Once the “new man” that is the man changed by feminism was represented as a wimp, as overcooked broccoli dominated by powerful females who were secretly longing for his macho counterpart, masses of men lost interest. Reacting to this inversion of gender roles, men who were sympathetic chose to stop trying to play a role in female-led feminist movement and became involved with the men’s movement. Positively, the men’s movement emphasized the need for men to get in touch with their feelings, to talk with other men. Negatively, the men’s movement continued to promote patriarchy by a tacit insistence that in order to be fully self-actualized, men needed to separate from women. The idea that men needed to separate from women to find their true selves just seemed like the old patriarchal message dressed up in a new package.

Describing the men’s movement spearheaded by Robert Bly in her essay “Feminism and Masculinity,” Christine A. James explains: Bly claims that women, primarily since feminism, have created a situation in which men, especially young men, feel weak, emasculated, and unsure of themselves, and that older men must lead the way back…. Bly holds up the myth of the Wild Man as an exemplar of the direction men must take and never challenges the hierarchical dualisms that are so integrally linked to the tension he perceives between men and women. Arguably, the notion of the Wild Man merely reinforces clichés about “real masculinity” instead of trying to foster a new relationship between men and women, as well as the masculine and feminine. The men’s movement was often critical of women and feminism while making no sustained critique of patriarchy. Ultimately it did not consistently demand that men challenge patriarchy or envision liberating models of masculinity. Many of the New Age models created by men reconfigure old sexist paradigms while making it seem as though they are offering a different script for gender relations. Often the men’s movement resisted macho patriarchal models while upholding a vision of a benevolent patriarchy, one in which the father is the ruler who rules with tenderness and kindness, but he is still in control. In the wake of feminist movement and the diverse men’s liberation movements that did not bring women and men closer together, the question of what the alternative to patriarchal masculinity might be must still be answered.

Clearly, men need new models for self-assertion that do not require the construction of an enemy “other,” be it a woman or the symbolic feminine, for them to define themselves against. Starting in early childhood, males need models of men with integrity, that is, men who are whole, who are not divided against themselves. While individual women acting as single mothers have shown that they can raise healthy, loving boys who become responsible, loving men, in every case where this model of parenting has been successful, women have chosen adult males—fathers, grandfathers, uncles, friends, and comrades—to exemplify for their sons the adult manhood they should strive to achieve. Undoubtedly, one of the first revolutionary acts of visionary feminism must be to restore maleness and masculinity as an ethical biological category divorced from the dominator model. This is why the term patriarchal masculinity is so important, for it identifies male difference as being always and only about the superior rights of males to dominate, be their subordinates females or any group deemed weaker, by any means necessary. Rejecting this model for a feminist masculinity means that we must define maleness as a state of being rather than as performance. Male being, maleness, masculinity must stand for the essential core goodness of the self, of the human body that has a penis [note: obviously this is very cis-perisex language]. Many of the critics who have written about masculinity suggest that we need to do away with the term, that we need “an end to manhood.” Yet such a stance furthers the notion that there is something inherently evil, bad, or unworthy about maleness. It is a stance that seems to be more a reaction to patriarchal masculinity than a creative loving response that can separate maleness and manhood from all the identifying traits patriarchy has imposed on the self that has a penis. Our work of love should be to reclaim masculinity and not allow it to be held hostage to patriarchal domination. There is a creative, life-sustaining, life-enhancing place for the masculine in a nondominator culture. And those of us committed to ending patriarchy can touch the hearts of real men where they live, not by demanding that they give up manhood or maleness, but by asking that they allow its meaning to be transformed, that they become disloyal to patriarchal masculinity in order to find a place for the masculine that does not make it synonymous with domination or the will to do violence. Patriarchal culture continues to control the hearts of men precisely because it socializes males to believe that without their role as patriarchs they will have no reason for being. Dominator culture teaches all of us that the core of our identity is defined by the will to dominate and control others. We are taught that this will to dominate is more biologically hardwired in males than in females. In actuality, dominator culture teaches us that we are all natural-born killers but that males are more able to realize the predator role. In the dominator model the pursuit of external power, the ability to manipulate and control others, is what matters most. When culture is based on a dominator model, not only will it be violent but it will frame all relationships as power struggles.

btw i do in fact have to add at the end of this, that Abdullah Öcalan's paradigm of democratic modernity vs capitalist modernity not only centers women's liberation as a core element of true socialist and democratic liberation, but also i think Rojava and the Zapatistas in Chiapas are great examples of how actually intersectional revolutionary feminism can in fact lead to not just better conditions for women but true connection and solidarity between men and women.