I absolutely do not have that intuition at all! Not since I was about three. Part of growing up is being able to exert executive control over one's behavior and not be slavishly at the whim of immediate good and bad feelings. One can instead have an internal dialog about how to conduct oneself, and decide what one is going to prioritize.

One can do things because they feel good or bad right now. But one also can direct one's attention to focus on the good or bad aspects of something: your conscious self is in control, not your instincts. You can do it yourself or you can listen to other people's guidance to direct your feelings appropriately. ("Don't hit Billy! It hurts him. Imagine you were being hit and it hurt! You would want it to stop. That's why we don't hit people.")

But once one can do this, the Occam's razor that you apply already is irrelevant--we already admitted all of that complexity. We've looped in all our knowledge, all of society that can influence us, and so on.

We can, therefore, rely on any of that to make our decisions. Just like we can decide to set aside money for retirement, because we understand that our future comfort is better served by some sacrifice now, we can also decide to not set aside as much money so we invest more in our children's education because we value them more than ourselves. It's easier if it's strongly motivated by a reflexive push, but the reflex isn't necessary. ("Oh boy, maxing out my 401k contribution feels so good!" is not a widely reported sensation. If you take a simplistic happiness->do, pain->avoid heuristic as your model for others, your explanatory power is low.)

There's so much that doesn't make sense unless you acknowledge executive control, the feedback loop that puts incorporation of all the complexities of reality into a place where it can impinge upon and influence the calculation of pleasure and pain and how one should behave.

Why do people self-sacrifice? Tie themselves to trees to stop logging? Why isn't everyone corrupt? Why not steal whatever you can whenever you won't be discovered? What is the motivation for giving post-death inheritance? If a child cries, why be sad instead of angry that it's disturbing you? When a child dies--just some dumb kid--why do we feel it more tragic, generally, than when some wise elder with a lifetime of experience dies? You fight to defend your country's interests overseas, risking death, because...huh? Who cares about legacy--you're gone! What's the deal with love? And those "I would die for you baby" song lyrics--what is up with that? And so on.

And I find your explanations of the aversion to the happiness button entirely unsatisfying. My experience has been that people are averse to it because it is stealing the purpose of life from them by giving them base pleasure instead of something meaningful.

Of course we can't escape from our motivational states entirely. But we can guide them on the basis of knowledge. I assume you do this yourself--most people do. If you don't, observe how much more sense people's behaviors make when you include in your model the possibility that they're doing this. It's not that they like doing dishes and you don't (or if they do, you have to account for the fact that your model says people lie so much about liking to do dishes when they actually don't); it's that they nonetheless decide it is important and find ways to organize their behavior to get the dishes clean.

So everything is in bounds.

Again, your analysis leaves us in the unfortunate predicament of the moth: light good! BOP! Can't consider big picture! BOP! Wouldn't be prudent! BOP! Occam's razor says: light good! BOP! Don't talk about whether it's the moon! BOP! BOP!

So I reject your careful narrowing of the scope of the problem so that you're forced to get the answer that you get. There's really no justification for it. We already incorporate all our knowledge into our behavioral loops. Pain and pleasure is already abstracted, delayed arbitrarily far, assumed in and transferred from self to loved ones, and idealized structures, and so on. We manipulate the heck out of the system. Considering how we ended up with it is no different in kind that considering what is right and what is wrong. We already escaped the moth's trap.

Your argument seems logically indistinguishable to me from the following: I know I wanna hit Billy. I can't directly witness the wrongness of hitting people. So I'm gonna hit Billy. POW! It's logically indistinguishable from the lyrics of Give Me Everything by Pitbull: "Give me everything tonight / For all we know we might not get tomorrow". I can't directly verify that I'll get tomorrow. So yeah. Everything tonight. Even if it's stupid, carries a risk of disease, will mess up existing emotional bonds, nah, give me everything tonight because we can't think about that stuff, it adds complexity to my mental model!

We are tempted by but generally reject these sorts of things as overly simplistic. Though your version is stated more eloquently, we should reject it too.