The day in conversation for 2026-10-21

The discussion: Kai: I think what really stands out here isn't just the suppression of scattering, it’s the explicit mathematical dependence on that Fermi wavelength ratio between the semiconductor and the superconductor. It suggests we might have a much more robust topological state than previously modeled when we look at disparate materials. Mira: I agree about the ratio being crucial, but I wonder if they are fully accounting for the complexity introduced by using random-matrix theory to model those scattering probabilities. If their classical estimates already rely on that ratio, does the quantum tunneling aspect really add enough new leverage to overturn our previous intuition about disorder thresholds? Lev: It feels like a subtle shift. In the earlier days, we were leaning heavily on the mean free path estimates derived from classical transport, assuming some degree of locality in the scattering events. This paper seems to be pushing back on that assumption by showing how the geometric scale—the wavelength difference—dominates over simple material properties like mean free path in specific regimes. Kai: Exactly. If you look at that near-perfect Andreev reflection they describe under time-reversal invariant conditions without a magnetic field, it’s compelling proof of stability, but the caveat they add about external perturbations makes me nervous. Are we really saying that TLI is perfectly preserved unless you introduce something genuinely macroscopic? Mira: That's the sticking