Just to understand the baseline you're working form, what data are you drawing on when stating what's "normal"? It has appeared to me throughout this conversation that you are starting from the assumption that your internal experience (at least, with regard to mental imagery) is consistent with the majority and making assertions about the majority based on that.
If you are actually drawing on a larger amount of data, specifically pertaining to the mechanisms of people's internal experiences, that'd be useful to know. If you are not, I'd gently suggest that you're working from a flawed statistical assumption, and that looking at the numerous interactions in this thread where people describe materially different internal experiences ("I can only recall tastes in terms of descriptors" - "I can experience a taste by imagining it"[0]) may be more informative without presuming universal consistency.
Both my wife and one of my co-workers experience hi-res images superimposed over their actual vision when prompted (and occasionally involuntarily). This opposite end of the spectrum is hyperphantasia[0]. There are accounts in this thread of people having trouble reading books because they're too caught up in the visuals the story creates in their head (something that same co-worker has also mentioned happening).
I just responded to you elsewhere assuming that you're operating from a baseline experience of depending on visualization, but this comment has made me think you might actually also lack it, and that your assertion that aphantasia must be debilitating is from an assumption that the lack we're describing is something beyond your experience. Since it's all a matter of your internal experience, though, it's impossible for me to know.
I can only tell you that I do not have mental imagery and am not constantly surprised by my surroundings. Object permanence has absolutely no dependency on visualization; it is completely unsurprising to me that the stop sign near my house looks the same each time I encounter it.
I totally get that you, having lived a life where mental imagery is such an integral part of your baseline experience, assume that many of the things you rely on it for require it. However, the human brain is impressively adaptable, and it turns out many, many aspects that people assume are linked (and may well be for them as individuals) are not globally so.
This thread is full of such realizations - people assuming ability at chess, art, tetris, spatial reasoning, abstract reasoning, architectural design, conceptualizing of DB schemas, etc. must be correlated with facets of thought like strength of mental imagery, presence of an inner monologue, ability to dream, etc. In all of those cases, people have chimed in with (anecdotal, to be sure) counterexamples. It turns out brains are generally capable of doing a lot of different things in a lot of different ways.
This is actually one of the things about aphantasia that has already had a (small) study[0], and the results are actually the opposite! The participants with aphantasia were more likely to both (a) take longer to answer and (b) be _correct_.
I personally have aphantasia and have quite good spatial reasoning, including with mentally rotating objects. I think it's fair to assume that visualization and spatial reasoning are mostly (but not entirely) orthogonal spectra, and it's completely possible for people to fall anywhere on either, just based on how their brain developed its internal strategies throughout normal life.
Similar questions throughout this thread make it clear that people assume recognition and/or knowledge of what something looks like is tightly coupled to internal visualization, but (writing as someone with aphantasia) they're not. I can draw both of the examples you gave, thinking about them (including thinking about how I would go about drawing them) just doesn't include an image in my head.
Ed Catmull, former head of Pixar, also has aphantasia[0] and had the staff take the only real diagnostic that exists (the Vividness of Visual Imagery Questionnaire[1]) and found that, on average, the animators had less ability to visualize than the production managers. Obviously they varied significantly - the article mentions one that could play the entire movie in their mind - but lack of mental imagery clearly does not preclude professional level artistry.
> it sounds like the way you remember things must be missing a lot of details.
> Do you think that's accurate?
I don't think someone with aphantasia (of which I am one!) is really capable of assessing this - which details our memories have routinely omitted is an unknown unknown until we're explicitly made conscious of the gap.
With regard to your example of a movie, I think you may be overweighting the importance of visualization in recalling narratives (and details within them). Fiction books (for adults, anyways) generally lack visuals, yet readers across the entire visualization spectrum[0] can engage with them and recall/discuss scenes, plot points, etc. Absorbing the narrative of a movie isn't so far off from that. I just took my kid to see Inside Out 2 earlier this week, and have a pretty clear recollection (sans whatever gaps I'm incapable of being aware of) of all of it.
[0] I think, but am not sure, that it strengthens my point here to note that people who visualize the story as they read it are nearly certainly visualizing it _differently_, but that almost never poses a problem for engaging with others about it.
"think of" is far too ambiguous here to really meaningfully drill into the differences of _how_ people think about things. I know that a stop sign is a red octagon (with a white boundary) with the word "STOP" in the center. I could draw you a plausible picture of one without issue. Thinking about it just involves no imagery.
Information you might rely on visualization for, I don't (because I can't). I couldn't tell you with any confidence what my family that just visited this afternoon was wearing (something I assume is easier for people who would remember it visually), but I can recall what happened - in what order, where, who said what, etc. - throughout their visit.
Constructively using LLMs tends to require validating the quality of their output; even when not hallucinating content, they do hallucinate confidence. Increasing the size of the output dramatically increases the effort needed to validate it.
Both examples have sections where the model simply left in placeholders for concepts - in example 1, "Summary of Key Takeaways" repeatedly references `[Book Topic]`, while example 2 starts doing so much earlier in "Overview of the Book".
What is the goal of this project? If it is meant to be more than a learning exercise, I'd hope for a lot more investment on quality control for the final output (but then I'd perhaps be even more worried that people would trust that output uncritically).
That's exactly the model used by the Kakoune editor[0]. It definitely feels more intuitive to me, but I personally didn't stick with it due to vim's ubiquity.
But Michael Jackson was born on April 19th[0]! As well as March 27th[1], and many other days[2].
"Michael Jackson was born on August 29th" is the most likely answer to contextless queries like "When was Michael Jackson born?", but that does not make structurally identical sentences with different information false, merely less probable to be contextually correct.
If you are actually drawing on a larger amount of data, specifically pertaining to the mechanisms of people's internal experiences, that'd be useful to know. If you are not, I'd gently suggest that you're working from a flawed statistical assumption, and that looking at the numerous interactions in this thread where people describe materially different internal experiences ("I can only recall tastes in terms of descriptors" - "I can experience a taste by imagining it"[0]) may be more informative without presuming universal consistency.
[0] https://news.ycombinator.com/item?id=40764813