Peter Morgan video wins a Curt Jaimungal's Competition for Outstanding Research Explanation

Morgan’s winning submission, at 3h26m20s long, effectively four lectures, is a much extended version of the research seminar that he gave on May 1st, 2025, “A Dataset&Signal Analysis Unification of Classical & Quantum Physics”, which is on YouTube as well.  Part of the attraction of creating this new video is that the invitation was for the explanation to be at a graduate level, which is currently the level at which an explanation of the ideas sits comfortably.

“Explaining Quantum Field Theory as a Dataset & Signal Analysis Formalism” proposes a unifying reframing of classical mechanics (CM), quantum mechanics (QM), and quantum field theory (QFT) by treating all three as formalisms about datasets and statistics rather than about particles and their properties.  This perspective is partly inspired by the way machine learning initially focuses on large-scale datasets rather than on intrinsic system properties — an approach Morgan argues can be carried over into fundamental physics.  A central claim is that CM is demonstrably incomplete (due to Bell inequalities, Gleason’s theorem, and the Kochen-Specker theorem), and that an extended Classical Mechanics (CM+) that incorporates quantum noise and uses the full power of the Poisson bracket to introduce contextuality and generalized probability can bring CM and QM into much closer alignment, dissolving the notion of a “quantum-classical transition.”

Signal analysis serves as a useful conceptual bridge in this framework, positioned as a natural intermediary between CM and QFT precisely because it is about signals, ɴᴏᴛ about particles, and already employs Hilbert spaces through Fourier and time-frequency analysis.  Morgan proposes that experimental devices such as Avalanche PhotoDiodes (APDs) can be understood as signal analyzers operating across multiple frequency scales, reinforcing the view that measurement in physics is fundamentally about recording data, with avalanche transitions being part of an extended thermodynamics that includes both thermal noise and a Lorentz-invariant quantum noise.  Quantum computing, viewed through this lens, discretizes signals unevenly and exploits quantum noise as a computational resource, with QM as an abstract programming language layer for noisy signal-lines and with QFT as a closer-to-the-hardware machine language layer.

The vacuum state of QFT can be thought of as a noisy carrier signal that can be modulated as a preparation for creating permanent data records of measurement results.  A signal analysis mindset encourages us to rethink renormalization as a surreptitious nonlinearity that can be explicitly introduced into the Wightman axioms and, justified by the Reeh-Schlieder theorem, we can construct mathematical models for interacting QFT as patterns of overlapping modulation-transmission-measurement triplets as an alternative to the tradition of loop calculations and renormalization procedures.  Taken together, these ideas give us an empiricist, data-centric methodology for physics that resonates with the dataset-first orientation of modern AI and ML.

Morgan has been a Lab Associate in Yale Physics since 2004, when he followed his wife Christina Kraus from Oxford to be a professor of Latin Literature in Yale Classics. He wishes to acknowledge two particular moments of encouragement by Steve Girvin in 2017 and by Priya Natarajan in 2023.