Melanie Mitchell

Melanie Mitchell is the Davis Professor of Complexity at the Santa Fe Institute and one of the field’s most trusted explainers. She earned her PhD at Michigan in 1990 under Douglas Hofstadter and John Holland, building the Copycat model of analogy-making, and went on to write standard works on genetic algorithms, the award-winning “Complexity: A Guided Tour” (2009), and “Artificial Intelligence: A Guide for Thinking Humans” (2019). Her research centers on the hard question beneath the hype: what would it take for machines to genuinely understand?

Against the doom motion

At the June 2023 Munk Debate in Toronto, Mitchell and Yann LeCun argued against the motion that AI research and development poses an existential threat, opposite Yoshua Bengio and Max Tegmark. She reminded the audience that AI’s history is a history of failed predictions — superintelligence has been declared imminent since the 1950s — and argued that apocalyptic framing distracts from tractable, present-tense problems like disinformation and bias. Her research papers, including “Why AI is Harder Than We Think” (2021), make the same case with data: intelligence is not a single dial that scaling inevitably turns to superhuman.

Mitchell is no booster — she punctures inflated capability claims as readily as extinction fears, and her Substack dissects each new “sparks of AGI” announcement with a skeptic’s scalpel. But she remains an active researcher who believes the science is fascinating, valuable and nowhere near the cliff its doomsayers describe.