10% of the year is not a sensible way to measure your error. If I ask "When did X happen?" and you answer 2000 CE when the real answer was 2020 CE, there's a sense in which you are more wrong than if you answered 300 BCE while the real answer was 500 BCE. Even if you don't think this, you probably don't think that the sensible target should be narrower closer to 0 BCE.

Whereas you are in a meaningful way about as incorrect if you say 10 km when the real answer is 11 km as if you say 1 m when the real answer is 1.1 m

"How long have you had with the current biggest issue in your life?"

What does this mean? Is this "how long have I had the issue?" or "how long have I tried solving?"?

"How many X would you trade for a Y?" You should be more specific. Am I to imagine that you are generously offering me either n extra units of X or one extra unit of Y, and I need to figure out how big n needs to be for me to be indifferent? (This is how I assumed you meant it).

I think I know of the trick you are talking about, in that there does seem to be an obvious pseudoprediction place in my mind that interfaces with motor output, and it's obviously different from actually believing, or trying to believe. However I mostly can't manage more than twitches or smaller motor movements, and it gets harder the more resistant I am to doing it (thus, less useful the more I would need use of it). If I'm thinking of the right thing, then the failure of me to sometimes send the pseudoprediction to my muscles seems to be the cause of some various stuff I experience when I essentially can't get myself to do certain things (e.g. get out of bed) (going by how people react to my more detailed descriptions this phenomena appears to be something very unusual about me).

It feels to me like the same sort of "prediction" as my Inner Sim that visualizes what happens when I throw a ball at the wall - it's clearly distinct from what "I" believe.

I separately also have experienced the thing where I think the script says I ought to feel X and so I feel X, but that feels totally different to me. Possible exception: I recently (for completely unrelated reasons) had a panic attack (which are very rare for current me), and for a while after the big spike I would get close to having it again partially due to what might have been having that sort of pseudo expectation of the hyperventilating and then accidentally causing it to actually happen, which would then threaten to launch me back into the panic attack. This might secretly be how the script thing works, though it doesn't feel like it to me.

This is where I disagree the most. I have not particularly noticed cognitive reflectivity decreasing my passions, though it's possible that most of the reflection I do is either when I have caught myself in the grips of some feeling that will likely be bad (e.g. getting angry at someone over a triviality, or having a panic attack), or when I am already in a pretty neutral state. Thus to me it mostly feels like the only passions made more distant are the ones that I wanted distanced.

I try to be followed by my positive passions when they occur, with some triggers for when seriousness or distance is necessary.

You seem to also observe the thing Eliezer does here. What's it like for you, when reflection makes you less happy?

The way I think about it is it's based on what I care about. I am in fact unwilling to do certain things to save the life of someone who is threatening suicide and blaming me, because I care more about myself, and I am fundamentally okay with caring about myself in that way. If, say, my best friend made some stupid mistake that put her at risk of great harm, I would be doing the heroic responsibility thing because I care a lot about the outcome.

It's fine to care about yourself! The principle of "I am obligated not to harm you, but not obligated to help you" is a fine one. The point of heroic responsibility is to see what I could do in cases where I do want to go all out to achieve some outcome.

You can know why the analogy holds without having enough detail to compute with it. Likewise you could have a picture that is truly part of you, but where not all of it is, so that you only get halfway there.

It is pretty easy to understand why the Doppler effect happens (at least, if you happen to be exactly my past self). You can easily have the ability to independently come up with the explanation without being able to derive the formula.

Printing with my printers sucks - but e.g. my university has a nice upload service where you upload documents to a website and scan a card at one of multiple printers and can print; and those have always worked like a charm for me. So why is printing sometimes... easy?

He's going for in the sexy way, not not in the sexy way (perhaps he had a typo when you read it).

A well known example is the idea of using geostationary orbits for communication satellites popularized by Arthur C. Clarke (which he said was possibly subconsciously influenced by George O. Smith's story in the first Venus Equilateral)

Interesting. To me frosting feels almost physically painful to eat more than a small amount of, and I have no memories of any consequences from eating frosting (besides the immediate "ow")