Bartosz Milewski

Physicist, programmer (Haskell, C++), mathematician, category theorist. Author of Category Theory for Programmers and The Dao of Programming

blog: https://bartoszmilewski.com/
github: https://github.com/BartoszMilewski
book: https://github.com/BartoszMilewski/Publications/tree/master/TheDaoOfFP

Public Key

npub1glgxnwfzwxpvj3l6d96vxttum7rkfav3dmcwdkyqz9eqkt83mdfsaxtqzy

NIP-05 Address

bartoszmilewski_at_mathstodon.xyz@momostr.pink

Profile Code

nprofile1qqsy05rfhy38rqkfglaxjaxr947dlpmy7kgkau8xmzqpzust9ncak5cprpmhxue69uhhyetvv9ujumt0d4hhxarj9ecxjmntmzvuzn

Publishing to

relay.momostr.pink

Author Public Key

npub1glgxnwfzwxpvj3l6d96vxttum7rkfav3dmcwdkyqz9eqkt83mdfsaxtqzy

Show more details

Last Notes

2025-05-01 10:05:50

- reply

Maybe the problem is that we teach about energy first, get used to the idea of kinetic energy and potential energy, conservation of energy, etc. Then we spring up the Lagrangian in terms of energy and circle back to the Hamiltonian.
Maybe it would make more sense to think of the Lagrangian and the action as more fundamental and introduce energy (and momentum) as a result of Noether's theorem. After all, conservation of energy is not a fundamental law, as seen in General Relativity.

2024-10-18 21:08:53

- reply

Even Einstein didn't like the idea of the non-steady-state universe and he tried to fine tune it away using the cosmological constant. Then he changed his mind and disavowed the cosmological constant. Now we have resurrected it in the form of dark energy. The lesson is that elegance is more of a human weakness than the Nature's big concern.

2024-10-18 18:58:37

- reply

The reason I'm not being very coherent is because I don't have a competing theory. I just have a feeling that physics has lost its clarity of purpose. Ever since Galileo, physics was based on the repeatability of experiments. Obviously, we can't use the same methodology in cosmogony, so I find most arguments used by cosmologists very confusing. I'm trying to poke holes in them in order to get to get to the fundamental assumptions, which are rarely spelled out in scientific papers.

2024-10-18 07:28:56

- reply

But why is a "simply described, homogeneous state" better than any other state? The choice is purely subjective. Is Nature constrained by the Occam's law?

2024-10-17 19:49:42

- reply

My understanding is that Bayesian interpretation tries to estimate our ignorance. We have some theories, we compare their predictions with reality, and we modify the theories. So, indeed, looking at the CMB we can estimate the primordial fluctuations. What I don't buy is the extrapolation of QM to cosmology. In QM you have initial state and final state. Fine, you can pick your initial state a picosecond after the Big Bang and figure out what it most likely should have been, based on CMB, etc. But no current theory can derive this initial state from first principles. Cosmologists are trying to convince us otherwise.

2024-10-17 17:10:20

- reply

I do understand what it means to adjust your priors. We are seeing the Universe (or dinosaur bones), and we assume that there were certain fluctuations (dinosaurs) in the past. But that's different from concluding that we understand where these fluctuations came from (or that dinosaurs are predicted by Darwinian theory).

2024-10-17 15:28:27

- reply

Would it be fair to reword your assertion as: "Statistically speaking, the universes that exhibit a certain level of fluctuations are much more likely than the ones that don't?"

2024-10-17 08:55:11

- reply

Now I'm getting pushback I deserve 😉 .
In perturbative QFT, which I think is relevant here, we calculate S-matrix elements using Feynman diagrams. We get amplitudes, which we *interpret* as probabilities of transitions between initial and final states. This works almost perfectly for interpreting supercollider experiments. But what is the initial state of Big Bang?

2024-10-17 08:18:38

- reply

Do you mean lattice gauge theory?

2024-10-16 18:52:38

- reply

The whole argument seems to hinge on the Heisenberg's uncertainty principle, whose validity is extrapolated to the realm of quantum gravity, which is a huge stretch. Unfortunately, the experimental evidence is very thin, based on a single "experiment" that is the current Universe. It's not a repeatable experiment. The whole epistemology of physics is based on repeatability of results.

2024-10-16 18:13:27

- reply

I don't see how Bayesian interpretation could help. QFT, the best theory of the micro-world we currently have, can only answer a very narrow selection of questions. It lets us calculate the S-matrix: a transition function between asymptotically free states. The initial state of the Universe does not fit this description.

2024-10-16 16:21:36

- reply

There was no pushback on this post, so I assume that my doubts are legitimate.

2024-10-15 20:20:12

I'm calling BS on quantum fluctuations in the early Universe as an explanation for galaxy formation. Fluctuations are a statistical phenomenon, and there is only one Universe. And it's not a prediction but a postdiction--back-fitting from the CMB data.

2024-10-14 16:40:27

- reply

I remember reading about perceptrons and McCulloch–Pitts neurons when I was in high school.

2024-10-06 08:06:03

- reply

So it's even worse than I though. In LQG we can't even describe empty space, much less show that it's Lorentz invariant. Of course string theory is no better: it introduces lots of epicycles without solving the fundamental problem of quantum GR, which is non-renormalizability. When I was studying supersymmetry, the pitch was that the more symmetries you have the better the chance that infinities will cancel in perturbation theory. That didn't pan out either. So we are stuck. Hence the atmosphere of doom in the foundations of physics.

2024-10-05 17:24:58

- reply

Oops, I meant to reply.
My understanding is that the vacuum described by LQG is not Lorentz invariant. So light moving through this vacuum for billions of years would be affected by this, and the effect would depend on its wavelength. The equations might be invariant, but the solutions are not.

2024-10-05 16:51:42

I'm the opposite. I could never understand how you could have Lorentz invariance if you discretize spacetime. It might emerge as an approximate symmetry at larger scales, when individual violations average out because of the randomness of the lattice.
I see QLG mostly as a regularization scheme that gets rid of infinities by providing a short-distance cutoff. But this cutoff cannot be Lorentz invariant, as Sabine points out. I can't see a way out of it.

2024-10-05 15:54:34

An excellent takedown of the loopy quantum gravity by Sabine.
https://www.youtube.com/watch?v=cBIvSGLkwJY

2024-10-04 18:35:17

Unfortunately, the lesson for Iran from the current conflicts in Ukraine and the Middle East is that nuclear blackmail works.

2024-10-03 19:58:58

- reply

It's possible that future generations will consider our Standard Model laughable.

2024-10-03 07:59:06

- reply

Maybe the closest modern scientific analog would be homeostasis, "the state of steady internal physical and chemical conditions maintained by living systems." The ancients tried to concretize this idea as some kind of "fluid" (or, in more modern parlance, "energy").
When a living being dies, it seems like something is leaving its body, as if the body were just a vessel holding something non-tangible-- spirit, soul, pneuma, energy, life force, 𝘴𝘩𝘦́𝘯... Also, when a new life arises, it always comes from some other living being, as if there was some kind of conservation law.

2024-10-02 18:11:40

- reply

Is spirit understood as the mind that can exist without the body? Like after death?

2024-09-27 22:32:37

- reply

I'm saying it only half-jokingly. QED can be formulated in terms of field strength tensor at the expense of locality. But we know that locality is a rather crude construct (cf. Aharonov-Bohm and spooky action at a distance).

2024-09-27 22:15:29

- reply

Gauge theory was invented by globalists to sell locality

2024-09-15 18:27:48

- reply

Speaking of atmospheric phenomena, I recently saw a sun dog on my trip to Iceland.
https://en.wikipedia.org/wiki/Sun_dog

2024-09-14 20:08:34

- reply

You are correct.

2024-09-14 19:58:30

I can imagine some extraterrestrials discussing human intelligence. "Human brains have nothing to do with intelligence, it's just a bunch of neurons adjusting their synaptic weights."

2024-09-04 17:31:25

Around the 1960s humanity reached the peak of optimism.
1968 was when "2001: A Space Odyssey" was released. 1969 was the Moon landing and Woodstock.
Compare this with 1979 hoplessness of "Alien." It's been downhill since then. We lost hope in the future.

2024-09-03 21:26:10

It's easy to come up with an elegant theory, but when you try to apply it to reality, you are forced to add epicycles upon epicycles. This is the story of the superstring theory in a nutshell.

2024-08-21 13:06:55

If I told you that the only interesting numbers are between -35 and 27 you'd think I'm crazy. But this is the outer reach of physics, between \(10^{-35}\) (Planck length) and \(10^{27}\) (size of the Universe). If we completely ignore everything else, why are we using real numbers?