Speaker
Description
For fifty years, lattice gauge theory has rested on the Euclidean path integral and importance sampling — a combined paradigm whose successes are matched only by the sharpness of its failures. These failures are exponential and few: the sign problem at finite density and non-zero θ, real-time dynamics, signal-to-noise ratios that degrade exponentially in Euclidean time, small overlaps between trial and target states, and tunneling times at first-order transitions and in topologically frozen simulations. In this talk, I revisit these five challenges from the perspective of a lattice practitioner and ask what quantum computation changes. Two of them — the sign problem and real-time evolution — are solved structurally by unitarity. The remaining three are not removed but rephrased: they resurface in state preparation and measurement, where the exponentials may or may not return depending on the algorithm. I argue that string breaking, combining statics and dynamics in a single process, is the natural community benchmark, and that quantum advantage must be measured against the best classical methods, with the error-bar culture the lattice community has built.