
Timnit Gebru, the Ethiopian-born computer scientist and one of the most consequential voices in the fight for ethical AI, has announced that her debut book, Deep Unlearning: The Rise of AI and the Radicalization of a Tech Idealist, is now open for preorder. Sharing the news on Instagram, Gebru admitted the process took more out of her than she expected: “I don’t know how people write so many books, because this one took a lot out of me. It turns out, I had a lot to say, with editors cutting out 50,000 of my words.” She used the announcement to ask fans directly for help: naming local bookstores that should stock the book, and cities she should visit on tour once it’s out.
Gebru’s name carries real weight in tech circles, and for good reason. She is the founder and executive director of the Distributed AI Research (DAIR) Institute, and before that, she was the most senior Black woman on Google’s Ethical AI team, a position she was forced out of after raising the alarm about bias embedded in large language models, a departure that sent shockwaves through the industry and made her one of the most recognizable critics of unchecked AI development. Her influence has been formally recognized well beyond tech: she’s been named to TIME‘s 100 Most Influential People list, the BBC’s 100 Women of the Year, and Elle‘s 100, among other honors.
Deep Unlearning traces the arc of that career from its earliest roots; Gebru’s childhood fleeing the Ethiopian-Eritrean war in 1998, the long road to refugee status, and a formative belief that science and math were reliably objective, safe spaces, insulated from the injustices she witnessed elsewhere in life. Published as a memoir woven with manifesto, the book follows her through high school in Greater Boston, her time at Stanford, and her work at Apple, Microsoft, and Google, tracing how she came to recognize that the same biases she’d encountered outside of academia were very much alive inside the technology being built to shape the future. It’s a book about the choices the tech world has made and the ones Gebru insists are still available to it.
Deep Unlearning is out February 16, 2027, via Atria/One Signal Publishers, and is now available for preorder.







COMMENTS -
Reader Interactions