Nobel Prize in Physics awarded to John Hopfield and Geoffrey Hinton [pdf]
Overall reaction
- Thread is dominated by surprise and skepticism that the Physics Nobel went to work on neural networks rather than to “core” physics.
- Some commenters are pleased that foundational ML work is honored and call the award “awesome” and overdue, but they are a minority tone-wise.
Is this really physics?
- Many argue the work is computer science / applied math / neuroscience, not physics, and say the committee is “jumping on the AI bandwagon.”
- The official rationale (Hopfield nets as spin-glass energy landscapes, Boltzmann machines from statistical physics) is widely seen as a stretch: “inspired by physics” vs “advancing physics.”
- Counterpoint: physics today includes complex systems, information theory, and statistical mechanics; ANN theory can be viewed as “physics of computation.”
Importance and influence of Hopfield & Boltzmann models
- Several note that Hopfield networks and Boltzmann machines were historically important and rooted in condensed-matter/statistical physics.
- Others say these models have had limited practical use compared to backprop-based deep nets, CNNs, transformers, etc., and that the committee cherry‑picked physics-flavored models to justify an AI prize.
- Some stress their role in showing deep networks could be trained at all (via RBMs/Deep Belief Nets), calling them foundational despite later obsolescence.
Impact on physics practice
- Supporters point out ANNs are now standard tools in particle physics, astrophysics, materials science, medical physics, etc., especially for classification in massive datasets (e.g., Higgs searches).
- Critics respond that tools enabling experiments (NNs, spreadsheets, OSes) are not themselves physics discoveries and shouldn’t displace unrecognized work in e.g. quantum information, condensed matter, cosmology.
State of physics and Nobel credibility
- Some read the award as evidence that fundamental/high‑energy physics is “stuck” or lacking clear Nobel‑worthy advances.
- Others say physics beyond HEP is thriving and the choice reflects hype rather than stagnation.
- Frequent comparisons to controversial Peace and Economics prizes; some feel this “devalues” the Physics Nobel.
Nobel categories and calls for change
- Recurrent complaint: there is no Nobel in mathematics or computer science; this misfit drives category‑stretching.
- Suggestions include a dedicated “Nobel Memorial Prize in Computer Science” or an applied math/CS category to avoid forcing ML into physics.