Notes on “Lost in Math”
As part of my mental and emotional well-being program, I’ve decided to read more good books and spend (a lot) less time consuming the nooz. This week I read “Lost in Math: How Beauty Leads Physics Astray” by Sabine Hossenfelder. Recommended, and here are some notes (not really a full review).
The central thesis of the book is that theoretical physics of the last couple decades or so has become unmoored from experimental validation, and much of what is actually produced is more “beliefs” than scientific method.
A lot of current-day theoretical physics is actually more at the intersection of philosophy and physics, but that intersection is basically unfunded. Trust between physicists and the humanities was burned in the Sokal hoax, and never re-mended. Very few philosophers have enough credibility in even baby physics to be taken seriously. Richard Dawid is one of the few exceptions (I’m personally also reminded of Sherry Turkle, who was similarly one of the few people with credibility in both the computer and sociology worlds).
The crisis in theoretical physics is related to deeper dysfunction throughout academia, including a move away from tenure to a gig economy, the never-ending chase for funding being more and more driven by marketing, and many of the same forces that are behind the replication crisis in the social sciences. Academics also tend to be prone to a wide variety of cognitive biases, and in denial about the fact.
I enjoyed reading about the details of recent progress in physics, as someone who was very interested in the area when I was younger. There used to be a strong interplay between experimental results (particle accelerators) and theory, but we’re now in something of a desert, because the “standard model” basically predicts not only pretty much every experiment that’s been done, but also the experiments which are feasible. Quantum mechanics and general relativity aren’t even consistent with each other at massive energy scales, but on the other hand experiments at those energy scales aren’t practical; at lower energy the approximation of assuming (for the sake of gravity) that mass is at a single point is fine.
I’m not going to try to summarize the narrative about truth and beauty in physical theories, though it’s the core of the book, as I wouldn’t be able to do it justice. If you’re interested, read the book.
Some personal thoughts. Looking back, I’m glad I chose computers rather than physics as a path, because I think I’ve done more interesting things than I would have (to say nothing of impact).
A lot of what Dr. Hossenfelder says in the book is relevant to computer science as well, I think. There’s a massive amount of belief in the field, masquerading as something more intellectually rigorous. I think perhaps the closest parallel is the almost-complete takeover of academic programming language theory by type theory, though I think the interplay between theory and practice is healthier than in string theory (Rust in particular has benefitted from type theory, while other popular languages such as Go represent almost a rejection of modern academic trends). Looking back, object oriented programming was a belief-based trend which is now fading. AI is another area where intellectual rigor is lacking, and the explosion of papers there reminds me of the 500 papers written to explain the “diphoton anomaly,” which turned out to be statistical noise.
Many of Dr. Hossenfelder’s suggestions on how to improve physics research seem well argued to me, though I suspect we’ll see incremental improvement at best. I’ve turned away from academia because of a general sense that it is not well, but would love to have a place to just do research. But these days it seems we as a society can’t get the basics right. Maybe the best model to follow is Garrett Lisi (one of many colorful and well written interviews in the book), spending half time thinking deeply and half time surfing.