Genady

There are many phenomena moving with any arbitrary speed. They all can't carry energy or information.

One needs to take in account a teacher/student relative weight.

In principle, anything that has energy-momentum affects the curvature. Practically, a few million particles will not have a measurable effect. They are all fields, i.e., machines that get spacetime coordinates in and produce a scalar / spinor / vector / tensor out. They all have Lagrangians and obey stationary action principle. But gravity does not appear in interaction terms of Lagrangians the way other fields do. Thus, it is entirely distinct. (These are short preliminary answers. Markus will reply with much more depth, I expect.) No. All charged particles interact with EM (photon) field, but neutral particles don't. Quarks and gluons interact with gluons. All particles interact weakly. All massive particles interact with Higgs (not sure about neutrinos.) Seems, there are more interactions than fields.

4:1

Their interactions are described by interaction terms in the Lagrangians. They operate in the curved spacetime. A particle and its anti-particle are excitations of the same field. No.

I am not sure about the particles. Take electron neutrino and electron anti-neutrino, for example. They have the same 1), 2), and 3), but they are different particles.

I don't know.

My context is my life experience. AI's context is human library of texts and visual and auditory images.

Concepts are building blocks of thoughts. In fMRI studies, the difference between concepts and their visual representations manifests itself, for example, in areas of brain which are engaged when a task requires thinking of something vs. when a task requires imagining that something. In the latter, the same areas are engaged as in the former PLUS primary sensory or motor areas. Presumably, engaging these primary areas is what makes the images specific.

In addition to the above, there is usually logarithmic relation between a stimulus and our perception of it, aka Weber–Fechner law (Weber–Fechner law - Wikipedia). See example here: Logarithmic Time Perception - Exponential/Logarithmic Functions (weebly.com)

I'm uncertain about why. Because you do not comment on how AI works anymore. No, it's about how context limits understanding. Context limits understanding. What is there to discuss?

No, we are not. It has nothing to do with the topic of what context AI has.

I guess you are not interested anymore in figuring out how AI works. This is OK.

Yes, interpretation of QM is about how its math relates to reality. The cat example shows that that specific interpretation does not relate very well. So, use another interpretation which does a better job. They exist. In any case, it is not about reality but only about how to interpret QM without getting absurd results.

No, the cat example is about interpretation of QM.

Typo. Fixed. HUP, Heisenberg uncertainty principle. We are not talking about lions now. But about how AI is different from human brain.

The cat example is not about HUP. It is about quantum superposition of states, which is a separate QM principle.

You mean, HUP? It works in computers exactly the same way it works in neurons. There is no way around it. It is an underlying physical principle of everything.

In what way? Is this difference what makes human brain different from AI? What uncertainty? What it has to do with anything?

If you look inside the human brain, you find neurons firing pulses which are not very different from 01 etc. This by itself does not indicate if there is or there is not understanding. I have no doubt that AI does not understand anything in the human sense of the word, but this is not because of 01 etc.

Maybe in Sci-fi, but not in the current AI.

Yes, perhaps it does not understand any of that library. However, it behaves as if it does. People are different in that they do understand the meaning. They are not different in that they also behave as if they do (although they usually in fact do).

My understanding is that the context an AI has is the (almost) entire library of texts and images accumulated by humans.

Yes, this might be a problem. I don't know what we can do in this case. OTOH, may be the problem is more specific, i.e., a different understanding of how the current AI works. In this case, the problem could be cleared out.

What do you mean here?