And you think deep learning isn't physics? How much RG did you study in grad school?
More importantly, many "fundamental" questions about mathematical physics will soon get to be resolved through tabletop type experiments. AdS/CFT type methods led to the first analytical proof of Bethe ansatz results I've ever seen that made me sit up and take notice of that field (BA, not AdS/CFT).
It disappoints me that on pressing issues of our time like climate change, APS can't seem to get its shit together and take a stand. This may be a symptom of what you are saying, the best and the brightest are going to the fields you describe and taking the power and the klieg lights with them.
PS: And where the dickens do you think belief propagation came from - I certainly first heard about it from statistical physicists. To treat deep learning in a divorced manner as you are is like saying oh, I'm too lazy to learn how the foundations of my house are mortared, I just want to rearrange the furniture in the solarium.
To establish in a semi-rigorous way the (btw obvious) relationship between coarsening in physics and in stacked neural networks, as was done long after the latter became a thing [1][2], is at the same level as observing the analogy between Black-Scholes and diffusion. It's cute, but it does not turn the subject into physics.
It should go without saying that you can not, by definition, resolve fundamental questions about mathematical physics by experiment.
The APS took a stand on climate change years ago [3]. It was clear enough to make Harold Lewis [4] and Ivar Glaever [5] resign in protest.
P.S. Belief propagation was first proposed by Judea Pearl in 1982 [6]. The closest connection to physics one could realistically claim is that Pearl got a M.Sc. in physics before turning (back) to electrical engineering and then AI.
and in any case, I recall Schwinger resigning from the APS over cold fusion. Doesn't make cold fusion right.
Also I didn't realize the ground state of the Hubbard model wasn't mathematical physics? Spin liquids aren't mathematical physics? This is the same old story where people like des Cloizeaux and Pearson discovered bosonization and helped establish the long tradition of condensed matter people studying the Heisenberg model and making useful contributions to "high energy" physics and basically get ignored
In the end my point is that you loved the idea of changing the world more than you loved physics. There's nothing wrong with that.
More importantly, many "fundamental" questions about mathematical physics will soon get to be resolved through tabletop type experiments. AdS/CFT type methods led to the first analytical proof of Bethe ansatz results I've ever seen that made me sit up and take notice of that field (BA, not AdS/CFT).
It disappoints me that on pressing issues of our time like climate change, APS can't seem to get its shit together and take a stand. This may be a symptom of what you are saying, the best and the brightest are going to the fields you describe and taking the power and the klieg lights with them.
PS: And where the dickens do you think belief propagation came from - I certainly first heard about it from statistical physicists. To treat deep learning in a divorced manner as you are is like saying oh, I'm too lazy to learn how the foundations of my house are mortared, I just want to rearrange the furniture in the solarium.