He once "leaked" the idea that Jim Simon's trading success came from his use of ideas called "gauge theory" and "fibre bundles".
I forgot the exact timestamp, but you will have to watch the entire interview to find that segment — https://youtu.be/zVWlapujbfo
https://arxiv.org/abs/hep-th/9710148
https://www.amazon.com/Physics-Finance-Modelling-Non-Equilib...
I don't think it's a crackpot theory. The basic idea is that the gauge group is the group of rescalings of the units of money, and arbitrage appears as curvature in the gauge field, i.e. you end up with a net change when you parallel-transport money around a loop in the (discrete) space of assets and time.
RenTech is quite secretive, but this supports the rumors that simple graphical models for time series were behind some of their trading strategies.
Brown and Mercer (who became ideologically opposed co-CEOs)
https://old.reddit.com/r/algotrading/comments/k299vp/were_br...
"Speech recognition" in those days meant Markov too