The rapid spread of generative AI tools has sparked urgent debates about ethics, governance, and even existential risk. These concerns are real, but they often miss a prior and constitutive dimension: the aesthetic. In this talk, I argue that no adequate understanding of artificial intelligence—and no robust AI ethics—can be developed without sustained attention to the aesthetic forms through which AI enters human experience.

Today, many critical responses to AI focus on transparency, bias, or political economy. Yet when machine learning systems generate images, sounds, and texts, or when they infiltrate experience in subtler ways, they reshape foundational lived relations to the sensible world. Aesthetics is not merely a matter of artistic style but of the mediation of experience itself—a matter of the ways we sense, interpret, and imagine. Accordingly, to speak of “AI aesthetics” is to invoke both aesthesis—the broad field of perception and sensation—and aesthetics in the narrower sense of artistic form. Both are crucially at stake in today’s machine-learning algorithms. AI systems like Midjourney, DALL-E, or GPT-5 not only generate potential artworks but also make otherwise invisible computational processes indirectly perceptible and actionable; in so doing they insinuate themselves into the fabric of experience and reshape the very conditions of perception. In this sense, aesthetic forms are not secondary embellishments but essential mediators of how AI becomes intelligible to us—as well as crucial vectors with respect to who “we,” as perceiving, deliberating, and agential subjects, are. By analyzing artworks that grapple with these new technologies, I show that AI aesthetics is foundational to the cultural, political, and ethical challenges now unfolding.