AI, creativity and interesting ideas
For more than thirty years, I’ve explored creativity through research, music, invention, and writing. With collaborators, I’ve developed interactive musical systems such as the Continuator and Flow Machines, studied how constraints guide generation, and investigated synthetic populations.
I’ve led research teams at Sony CSL and Spotify. Across fundamental research and practical applications, two questions guide my work: what makes music, texts, ideas and cultural artefacts in general interesting, and how do such objects emerge?

What I’m working on
Generative AI research
How can a generative model meet explicit constraints while staying faithful to what it learned? Pierre Roy and I revisit the research that introduced Markov constraints.
NeurIPS 2026 · Exact Regular-Constrained Sampling for Variable-Order Markov Generation
Constraints Journal 2026 · Markov constraints revisited: commentary on “Markov constraints: steerable generation of Markov sequences”
IJCAI 2026 · Constraining Generative Models: A Survey from the Constraint Programming Perspective
Imagine All The People
Play animation
Stop animation
I co-founded Imagine All The People to help teams test ideas, messages, and decisions with simulated audiences.
Forthcoming book
La virtuosité à la portée des caniches
Du plaisir de dire n’importe quoi
IJCAI 2027 · AI, the Arts and Creativity — call for papers. I’m co-chairing this edition. Submissions due January 11, 2027.
About
I’m an associate researcher at LIP6 in Paris.