Timnit Gebru is a computer scientist and AI ethics researcher known for co-authoring the widely cited paper "On the Dangers of Stochastic Parrots: Can Language Models Be Too Big?" and for her high-profile departure from Google in December 2020.
Gebru earned a PhD in computer vision at Stanford University under Fei-Fei Li, and her early research included studies of demographic bias in commercial facial recognition and analysis systems within Computer vision, work that fed into a broader body of research on Algorithmic bias in AI systems trained on unrepresentative data. She went on to co-lead Google's Ethical AI team, one of the most prominent corporate AI ethics research groups at the time.
Stochastic Parrots and departure from Google
In 2021, Gebru, Emily Bender, Angelina McMillan-Major and, under a pseudonym due to Google's objections, Margaret Mitchell published "On the Dangers of Stochastic Parrots," a paper arguing that ever-larger language models trained on uncurated web text posed risks including environmental costs, encoded social biases, opacity about training data, and the risk that fluent output would be mistaken for genuine understanding, the Stochastic parrot critique. Internal disagreement over the paper at Google, including a request that Gebru and her co-authors retract it or remove their names, led to a public dispute; Google stated Gebru resigned, while Gebru and many observers characterized her exit as a firing. The episode, along with the subsequent departure of her Ethical AI co-lead Margaret Mitchell, drew widespread attention in and outside the AI research community and became a touchstone case in debates over AI ethics research inside large technology companies.
DAIR
In 2021, Gebru founded the Distributed AI Research Institute (DAIR), an independent, non-corporate research institute focused on studying AI's harms, particularly on marginalized communities, and intentionally structured outside the funding and incentive pressures of large tech companies. Through DAIR, Gebru has continued to publish and speak on AI ethics, AI governance and the risks of scaling large language models and generative AI systems without adequate attention to their social costs.
Influence
Gebru's case is frequently cited in discussions of whether internal corporate AI ethics teams can meaningfully constrain the products of the companies that employ them, and the Stochastic Parrots paper remains one of the most cited critical papers of the generative AI era, continuing to shape debate over the costs and limits of scale as the primary strategy in AI development.