← Back to Home
Risto Miikkulainen
Professor of Computer Science, University of Texas at Austin. VP of AI Research, Cognizant AI Labs. AAAI Fellow. IEEE Fellow. INNS Fellow.
Professor of Computer Science, University of Texas at Austin. VP of AI Research, Cognizant AI Labs. AAAI Fellow. IEEE Fellow. INNS Fellow.
Professor Risto Miikkulainen teaches Computer Science at the University of Texas at Austin and serves as VP of AI Research at Cognizant AI Labs. With an M.S. in Engineering from Helsinki University of Technology and a Ph.D. in Computer Science from UCLA, he is a global leader in AI innovation. His research centers on neuroevolution and generative AI, tackling decision-making challenges in natural language processing, vision, and societal applications.
He has authored or co-authored over 500 articles, including the first-ever book on neuroevolution, published in 2025 from MIT Press. At Cognizant, he scales these solutions for real-world impact. Recognized as an AAAI, IEEE, and INNS Fellow, Risto's recent accolades include the IEEE CIS Evolutionary Computation Pioneer Award, the Gabor Award from the International Neural Network Society, and the Outstanding Paper of the Decade Award from the International Society for Artificial Life.
Miikkulainen opened with a deceptively simple question: what if AI could invent solutions humans cannot dream up? That is the promise of neuroevolution, the technique of using evolutionary algorithms to optimize neural networks rather than relying on gradient descent alone. Where conventional training nudges a network toward known solutions, neuroevolution explores the space of possible network architectures and weights in ways that routinely produce surprising, high-performing strategies that human designers would not have considered.
The masterclass traced the practical application of these ideas at Cognizant through Project Resilience, an open-source platform that applies neuroevolution to complex societal challenges. He walked through a case study in optimizing land-use policies, showing how the platform generates and tests thousands of candidate strategies across competing objectives, sustainability, equity, and economic output, to surface a Pareto frontier of real-world options for policymakers.
Three lessons anchored the session: first, how neuroevolution works mechanically and why it escapes local optima that trap gradient-based methods; second, why this approach is particularly valuable for AI safety, since it can discover failure modes and robust solutions that deterministic training misses; and third, how founders, policymakers, and researchers can access and extend the open-source tooling today. The session closed with an invitation to collaborate on applying these techniques to participants' own challenges, from climate mitigation to equitable urban planning.
Connect with AI safety researchers, policymakers, and practitioners shaping the future of responsible AI.