If you look at the rescorla wagner model of classical conditioning, you see that there's this (lambda-Vtot) term in the deltaVx equation, where lambda is the strongest possible association for the unconditioned stimulus and Vtot is the sum of the association strengths for all conditioned stimulus. When the two terms are equal, the learning rate drops to zero. This keeps us from "learning" extraneous associations for a stimulus that has already been explained.
And from what I have seen with hypnosis, it looks like Vtot ~= lambda in general. In other words, I think we walk around thinking that we already know everything and that there is nothing left that needs to be explained, so there's nothing to learn. It fits with how people can just brush off unexpected events as "god dun it" or "it's just random" - and then go on with their day as if there's nothing strange happening. Hypnosis and mind hackery are all about creating a gap where you think you have something to learn - whether it's explaining the uncharted territory of "hypnosis" from a position of authority, or driving in a wedge with a pattern interrupt or confusion pattern.
So here's why I think this:
The scientific literature keeps saying things like "hypnosis doesn't improve much over normal suggestibility". As a hypnotist, a common first response is "well, the scientists are shitty hypnotists" - which is absolutely true. However, the interesting part is that you can totally get crazy things by just suggesting them - like, with no "induction" or anything. My never-been-hypnotized friend told me about the time his friend was being a goof and said "Look! Your hand is orange!" and by golly it was orange! Just like that. And people that have ever been hypnotized before are often capable of re-experiencing phenomena by simple direct suggestion with no special drawn out state-inducing process. Anthony Jacquin dubs this "permanosis".
Even further, the science says that "hypnosis" doesn't really make the responses more intense either. Which seems weird, given some of the responses I've gotten with hypnosis. There's stuff I've done with hypnosis that sounds totally insane if you didn't know about what hypnosis can do. I can get people in a state where I can tell them to really focus on what I'm typing, and then everything else shuts out and they can't even hear. If I say they feel nauseous and will puke, they feel nauseous and will puke. Then again, I've gotten some intense responses without hypnosis as well. And there doesn't need to be a difference between Vtot and lambda to get an intense response if the response has already been conditioned.
There are a lot of people who sit for extended periods of time to learn how to control their attention, but hypnosis can give them the same response much faster. One might think "Oh, okay, these hypnotic subjects are inherently special and they have massive top-down connections or something" - yet it's fairly reliable. As long as the subject hasn't thought about how to resist and is cooperative enough to let me lead their direction for the first minute until I get name amnesia, then I'll get everything. If not the first try, then the second. I can tell who it's going go work with not by innate features, but by their behavior and cooperation. It sure as hell looks like it's just expectation and if you can get expectation, you win - screw extended conditioning trials.
All of this tells me that hypnosis is about removing blocks, not about building strength over repetitions - even with the intense stuff. It's just that sometimes blocks are hard to remove and are better removed through other intense forces or "cheats" than by reasoning things through.
It's not about intense salience, attention, and repetition - those are easy. It's about making sure that Vtot does not come in and squeeze out room to learn.