When M.R. Hasan talks about the future of artificial intelligence, he doesn’t begin with algorithms or benchmarks.
Instead, he speaks of inspiration not from science or engineering, but from the humanities.
Specifically, Paul Cézanne’s still lifes in which reality becomes a mosaic of shifting angles and perspectives — apples tipping into impossible balance, tabletops bending toward the viewer, shadows drifting where they shouldn’t.
Cezanne painted not what the eye captures in a moment, but what the mind assembles over time: a world built from movement, memory and imagination.
Hasan sees in that vision a guide for AI’s future. Just as Cézanne broke from convention to reveal deeper truths about perception, Hasan argues that AI must learn to look beyond its fixed vantage points. Intelligence, he believes, is not a single frame but a constellation of perspectives—and machines must be taught to embrace that complexity.
This idea has resonated widely.
“I’m so obsessed with Cézanne,” said Hasan, assistant professor of electrical and computer engineering. “He broke with convention to reveal a deeper truth about perception.”
That realization, Hasan said, became the seed of a new approach to the future of AI. And Hasan’s unconventional framework, once a curiosity, is now gaining traction across the global computing community.
A series of papers grounded in Cézanne’s work earned Hasan three major honors in the past year, including Blue Sky Awards in 2025 and 2026 from the Computing Community Consortium and the ACM International Conference on Multimodal Interaction (ICMI), and a recent Spotlight designation at the International Conference on Machine Learning (ICML).
His book – “Reconceiving AI: The World as an Apple or a Blue Orange,” arriving in August – expands that vision. Its central figure is not a computer scientist but Cézanne himself, Hasan said.
The book examines five dimensions of intelligence: visual, linguistic, diverse human intelligence, moral intelligence, and a proposal for a new human-AI hybrid system.
Threaded through these parts is the metaphor of the “Blue Orange Mind,” drawn from Paul Éluard’s surreal line, “The earth is blue like an orange.” Hasan uses it to describe human creativity: the ability to merge categories that shouldn’t fit, to generate meaning from contradiction, to imagine realities not yet formed. For him, it becomes a model for a future in which AI complements human intelligence rather than imitates it.
The book’s first part critiques how AI systems process vision. Cézanne’s paintings depict objects from multiple angles at once, mirroring how humans build visual understanding over time. AI systems, Hasan contends, learn from static images taken from a single moment, leaving them unable to construct reality the way humans do.
The second part turns to language, drawing on the psychoanalyst Jacques Lacan to explain why language models hallucinate and how human consciousness is shaped by symbolic gaps AI cannot inhabit. “Language models can mimic humans,” Hasan said, “but behind the scenes there is a strange gap.”
Challenging the assumption that AI should be modeled on neurotypical “smart humans,” Hasan in the book’s third part argues instead for an understanding of intelligence that includes the full diversity of human cognition — synesthetic, contemplative, post-emotional, and other modes.
The fourth part explores moral intelligence and its three dimensions AI lacks — the experience of lived time, the capacity for self-reflection, and the ethical imagination to distinguish what is from what ought to be.
Finally, Hasan imagines humans and AI evolving together into a collaborative hybrid form of intelligence — what Hasan calls the Blue Orange Mind. He believes this evolution is already underway.
His journey to Cézanne began with philosophy. While reading Maurice Merleau Ponty’s writings on phenomenology, Hasan encountered the philosopher’s analysis of Cezanne’s paintings. That led him to the Art Institute of Chicago’s 2022 Cézanne retrospective, where he spent hours studying the works with special access granted for research.
“I started seeing things I didn’t see before, such as strange angles and wrong compositions. And then I thought, ‘why would he do that?’”
Those questions became the foundation for a new way of critiquing AI systems – and a new way of thinking about intelligence itself.
Hasan worries that engineering often sidelines the humanities, even though understanding intelligence requires perspectives from art, philosophy, psychology and ethics. His book maintains that building better AI will demand a broader conversation about what intelligence is and what it should become.
"If we want better AI, we must stop building machines in our image. The path is to understand human intelligence in its full diversity and welcome machine intelligence as different from us in kind," Hasan said. “The future won’t be human alone or machine alone. It will be a Blue Orange Mind.”
