Algorithms anticipate our tastes, our decisions, and our futures, and in doing so, they influence them. Optimized for immediate performance, they tend to trap our societies in futures we do not choose: recommendations that homogenize imaginaries, predictions that fulfill themselves, categories that freeze identities. What if queer thought offered an alternative?
In Queer Algorithms (Éditions B42, 2026), Guillaume Chevillon shows that the characteristics long cultivated by minority experiences – tolerance for error, right to be forgotten, taste for chance, rejection of fixed categories, claimed opacity – paradoxically constitute what makes algorithmic systems more robust, more sustainable, and more respectful of the diversity of human experiences. In these systems, error is no longer a defect to be eliminated but a potential driver of prediction.
For there is no need to design or possess the algorithms in order to act on them. Refusing to cooperate or respond to their stimuli, playing with memory and favoring forgetting to break away from the determinism of recommendation, cultivating chance and experimentation, building communities and coalitions that strengthen our capacity to act, testing the limits of systems to escape the categories they prescribe for us: these are all practices, arising from minority experiences, that allow everyone to stand up to algorithms.