John D. Norton - Claims that information is fundamental are mistaken: The thermodynamic myth at the heart of information theory
See his article published in IAI News on 10th July 2026. "Considerations of information and computing complicate thermodynamic analysis and contribute precisely nothing."
CEO and co-founder of Google DeepMind, Demis Hassabis, has recently speculated that information might be as fundamental as mass and energy. As AI hype rolls on, many like him are increasingly tempted to apply computing principles to reality as a whole. Those who support this view rely on the idea that information has a thermodynamic cost, as proposed by physicist Leo Szilard.
See Happy Birthday Raymond Solomonoff (in particular 22:33 on how to get information from experiments) and Mike McCulloch on Why Everyone Should Care About Quantized Intertia.
I recently heard that the formalisation of the simplified Taylor/Wiles proof of Fermat's Last Theorem had been completed by Anthropic AI agents. I asked Google's AI how much energy this took and the estimate was of the order of one million MegaJoules which it says is equivalent to the annual power consumption of thirty US homes, so about the amount of electrical power I will have used in my lifetime. What is the information content of the Lean proof? It was "13 million lines of Lean and proved 29,500 intermediate theorems" apparently it also proved some 800 or so other theorems which were not necessary for the proof of FLT, ...
"The finished proof was checked by Lean; it uses just Lean’s three standard axioms, and a comparator confirmed that the theorem’s statement matches Mathlib’s own statement of FLT." What is a comparator?
"Comparator [apparently there is only one!] is a trustworthy judge for Lean proofs. It relies on having an existing Lean installation as well as:
- landrun, compiled from the main branch's source, present in PATH
- lean4export, at a version that is compatible with whatever Lean version your project is targeting, present in PATH
- (optional) nanoda, compiled with a recent version of Rust. This is only necessary if you want to check with the nanoda kernel in addition to the builtin one. cargo build --release will place nanoda_bin in the target/release directory of the checked-out directory, this directory must be present in PATH."
Oh it's trustworthy, so that's OK then. I would hate to think that Million MegaJoules was all for nothing, ...
See Fermat’s Last Theorem: The Curious Incident of the Boasting Frenchman and A “proof” of Fermat’s Last Theorem that fits the margin.
Comments
Post a Comment