Electrons in solids can do two remarkable things: they can pair up and flow without resistance, and they can collectively behave as fractionally charged particles. Predicting these phenomena remains a major challenge because the physics spans enormous spatial scales and exponentially large quantum state spaces. I will show how my group uses AI to understand electron pairing in realistic materials: machine-learned Hamiltonians reproduce moiré electronic structures with meV accuracy for systems of up to millions of atoms, enabling studies of superconductivity from weak to strong coupling in twisted semiconductors.

To understand fractionalization, we turn to quantum processors. Constant-depth circuits prepare 18 fractional quantum Hall states on up to 156 superconducting qubits; remarkably, some of the most exotic states are also the easiest to prepare. Together, these efforts show how AI can extend quantum-materials modeling to realistic scales, while quantum processors open access to highly entangled states beyond classical reach.

No colloquium: Labor Day Holiday.

Our understanding of the Red Planet is evolving as we continue to receive data from the ongoing activities of NASA’s Curiosity and Perseverance rovers. In this talk, we will review the questions driving our exploration of the martian surface, take a look at the some of the basic data derived from rover science, and explore some of the fundamental difficulties of interpreting data from the surface of another planet. We will particularly focus on issues familiar to all of science: our current state of knowledge, intrinsic bias and the assumptions we make and the questions we ask, and the necessity of integrating information across both discipline and scale.