Everything is of a piece because reality is one piece.  All divisions between areas of knowledge exist in the map, not in the territory.

The perspective that she labeled 'average utilitarianism' relies on an understanding of generalized Relativity as it applies to quantum mechanics -

- this being something past-Keltham didn't discuss with Cheliax earlier, because combined with the most elementary math of quantum fields, Relativity directly yields an understanding of antimatter, which is the most obvious way to use Wishes to destroy countries -

You can, of course, get the equations of Relativity just by observing the physical facts; but to really understand them, children are led to guess them in advance by contemplating certain questions and dissolving those questions as ontologically meaningless.

"How fast is the whole universe moving?" seems unobservable from inside the universe; and, you could argue, is not only unobservable but meaningless - because in the simplest conceptual frameworks that do predict what is observable, 'motion' is the motion of particles relative to each other.  Not, motion relative to an absolute space, that is unobservable and hence can be eliminated as an element of the theory.

But maybe there is an absolute space?  Maybe physics has absolute space beneath it, and everything is moving at a speed through that space.  Maybe someday you'll discover 'laws of physics' - simplest logical rules that would reproduce a universe embedding you to observe what you observe - that imply distinguished structures that stay motionless within space, mathematical seams that are observable.  And, measuring those, you'll discover the whole universe is moving at a billion kilometers per second relative to absolute space.  How do you know you won't?

Later, you're shepherded through discovering the relationship between electricity and magnetism, generalizing the classical-illusion field equations for those, and realizing that the wave propagation through that field is light.  And this, it seems at first, implies a fixed speed for light, relative to the electrical-magnetic field.

And you might think: couldn't you measure how fast you were moving relative to light, and so measure how fast you were moving relative to an absolute Background Space?  So the thought experiment about the whole universe moving twice as fast - suggesting that only relative motion is real or even meaningful - has failed to predict the character of physics; there was an absolute space after all.

But actually, every time you measure the speed of light relative to yourself, you find the same speed.  No matter how fast you're going, or how fast the light source is moving, you find the same measured speed of light from your own perspective.

And when you work out the logic of what that shocking fact implies, it ends up requiring that you view spatial dimensions and time dimensions as being relative to your current velocity... which is to say, the time distances and space distances that observers at different speeds observe as different quantities, are not the underlying elements of reality.

The only thing that's still invariant from every perspective is the interval between two events, which in terms of classical-illusion measurements would be expressed as the square of distance in time minus the square of distance in space, with the speed of light converting units between the two.

This surprising additional math, indeed, is exactly what's required to implement a universe where there's a universal speed limit reflecting the locality of causality, and yet the only meaningful elements of reality are the positions of things relative to each other.  That Reality went to this extra effort to make physics visibly 'relative', in this sense, is the beginning hint of a deeper truth that proves to be more general: physics is built around a certain spirit and character in which relative positions, not absolute positions, are the elements of reality in the ontology of physics.

Over and over, it proves possible to start from a thought experiment like "If I'm inside a sealed room, should I be able to tell if I'm staying still or moving at a constant velocity of a million miles per hour?", or "Should we be able to tell whether the whole universe is rotating or not, relative to absolute space, by seeing if there's centripetal forces being generated by the rotation?", to answer "No!  If I can't see the quantity from my own perspective, ultimate physics must be arranged in a way to make that quantity not exist!".  One can correctly derive intricate laws of physics from that principle.

It's idealistic reasoning, but it's a form of idealistic reasoning that Reality itself seems to use, the same way that Reality seems fond of calculus, or continuous quantities, or numbers and math more generally.  You could say, it's first-principles idealistic reasoning, using the sort of idealistic first principles that Reality has been empirically observed to respect, and which prove to cause people to correctly guess physics without observing it first if they're led to guess using those principles.

(Golarion physics, he strongly suspects, is partially an imitation of that simple dath ilani physics, and partially has been artificially constructed and modified and complicated away from that physics; so that this universe can run both mortal biology copied off dath ilan and dath ilan's physics, and also include magic and souls.)

It can be seen from 'first-principles reasoning using the kind of first-principles that Reality has been empirically observed to actually follow' that it shouldn't be sensible to ask "How quickly or slowly are the laws of physics operating?", unless there is some larger outer universe establishing a speed metric to be compared to.  Similarly, you can't ask "Is the universe upside down?", unless there is some larger spatial metric that embeds both the universe and something else that points in a direction.

Further beneath reality is quantum mechanics: in which the basic quantities are complex numbers, 'amplitudes', assigned to positional configurations of particles.  The integral over the squared absolute values of those amplitudes, the measure, seems to describe how real something is - or rather how relatively real something is, because physics doesn't talk about the absolute amount of reality, at that lowest level.  Only the relative quantity, relative phase, slope of derivative, of the amplitudes.

If you run a quantum experiment that divides the greater reality into two subworlds, with amplitudes over configurations that interact almost purely internally within a world -

(this happens all the time, to be clear, or at least it did in dath ilan, there's ten-to-the-large-number divergences of worlds every second as entropy increases over time, and Pharasma's Creation is either doing the same thing or pretending very hard that it is)

- and one of those worlds has twice the integral-over-quantum-measure as the other, you'll find yourself in the larger experiment-future two-thirds of the time.

Do a thousand of those experiments, and look back, and you should find that around two-thirds of the outcomes reflect the larger quantum future.  There's a version of you that sees the smaller outcome every time, a thousand times, but those yous are only 1/3^1000 as real, and you'll almost never find yourself there / only experience yourself seeing that to a very tiny degree.

There's no physical difference that would be observable if you doubled all the tiny amounts-of-realness.

And this is also the kind of physical principle that you can correctly guess from thought experiments about Relativity: what would it even mean if everything everywhere simultaneously became twice as real?

You can get this quality of quantum physics by observing experiments, but you can also advance-guess its character via the vastly productive principle of Relativistic thought experiments: it's meaningless to imagine all of Existence becoming twice as real, so reality is only relative, and that's why physics over amounts-of-realness only speaks of the relative quantity of those amounts.  Some things can be realer than other things; it is meaningless to ask how real they are in an absolute sense.

There's a meaning to one person being twice as real as another, inside of larger Reality.  You're twice as likely to meet people who exist in twice as many places.

But what does it feel like from the inside to become twice as real, or half as real, in an absolute sense?  Nothing, and in fact the thought isn't meaningful, just like there are no absolute phases in quantum mechanics, only relative phases of the amplitudes.

One future can be more real than another, and you'll mostly experience yourself in the futures that are more real; when you look back in your past, you'll find that the experimental statistics for results roughly match the physics-predicted amplitudes of those results.

But when you look at yourself and question how real you are in an absolute sense - imagine yourself becoming twice as real, or half as real - you're imagining something that wouldn't feel like anything, because it doesn't mean anything; just like it wouldn't mean anything for time in the universe to run twice as fast, unless it could run relative to some larger universe and greater metatime.

This, in a sense, is why you find yourself experiencing anything; the answer to the malformed question, "Why does anything exist at all?"  It doesn't require anything larger than yourself to give you existence, as would then need some further outer factor to lend existence in infinite regression.  Structures of relative realness always find themselves to be as real as themselves, however much more or less they exist compared to other things; and that's why you find yourself inside a physics ultimately comprised of a structure of relative-realness.

In dath ilan that physics over relatively-real elements was 'quantum mechanics' over 'amplitudes'; but even if it's something else inside Pharasma's Creation, it'll ultimately be made out of stuff that embodies relative quantities of existence.  Nothing that exists can be absolutely real (as isn't even a meaningful concept) but only relatively real to other things, so whenever you look closely enough at something that exists, you'll find out that it's made out of tiny bits of relative-degree-of-realness.

He is thinking all this, because it seems to him entangled, as truths-of-empirics and validities-of-reality often are, with what a sensible mind would end up valuing as it shakes out its emotional structures binding to pieces of reality-as-the-brain-models-it. 

It seems to him that you can't, actually, just say that you reject dath ilan's concept of how to value people's reality ("average utilitarianism" as she calls it, though in dath ilan it does not have a name), and have that be divorced from everything else dath ilan knows.

There are pieces of morality that can be pried apart from other elements of a coherent decision system - like whether you enjoy seeing people suffering or enjoy seeing them happy, that's something you can pry apart and invert without affecting other parts.  (At least if you're talking about an agent with a utility function; it doesn't work that way inside normal mortal humans, obviously, humans are woven together more tightly than that.  But in principle you could pry away the utility function of something that did have a utility function.)

Whether your ontology of thought is over relative amounts of existence, or hypothetical absolute quantities of existence as seen against an absolute outside-of-all-reality yardstick of existence-quantity-units - like imagining an absolute right-side-up direction of space - isn't something you can pry apart from understanding physics with an ontology that's based around relative positions and relative realness in a very visible way.

When you worry about whether it's a crime to make people's sum-over-futures add up to less than the reality of their current selves - to wonder if this is a crime apart from people objecting to it, apart from whether their remaining futures are pleasant or unpleasant - it seems important to comprehend that becoming less real does not feel like anything from inside, and in fact doesn't mean anything except relative to other things being more or less real than that.

When it comes to asking whether an enslaved being should be grateful to have been created, it matters to his own emotions-morality-philosophy that this being who will be enslaved would counterfactually otherwise still exist somewhere; in fact, would exist within a countably infinite number of such environments, all existing to some tiny finite degree of relative realness, summing to a finite total.  What an entity like Asmodeus is doing, in 'creating' somebody, is changing which environments are more real relative to that person, and changing which futures that person will predominately experience; and as an entwined effect, making that person more encounterable by others in the same larger environment.  If this future and environment is not pleasant, a future of slavery, this seems to him to be not a favor requiring a grateful reciprocal favor - as the act is phrased and described in his own ontology.

...Carissa realizes that you cannot reject dath ilan's morality piecemeal because all of the pieces form a worldview together. That is why she stopped using all of it and would, if her concern were for her own integrity, never use any of it to build herself, even the science.

None of that information makes her an 'average utilitarian', as she predicted it wouldn't, when she considered the space of possible observable features of reality dath ilan could have observed which would have caused them to all be 'average utilitarians'. Carissa took into account how good an explanation dath ilan would probably have for all of its alien values, considered how confident she was that her values were different, and isn't learning anything from being told that, yes, dath ilan has a predictable explanation for its beliefs. She didn't reject them in the conviction that dath ilani hadn't argued the question.

If there are an infinitely many Golarions which are functionally identical such that there are infinitely many Carissae in this exact moment of existence considering this exact problem, then there being half as many isn't a meaningful thing to describe (she recognizes that this isn't quite the frame Keltham is using, she's not sure yet if his frame is importantly different).  But it's coherent to care, for instance, about in what share of universes she exists, or in what share of universes in which she existed at some point she exists for a long time, or in what share of universes in which she exists her parents and sister exists, and it's coherent to, if you wake up inside a new universe, have preferences about whether you died and stopped existing in your old one. And if you prefer to exist in as large a possible a share of the universes that there are, and for the duration of your existence in every universe to be as large as possible, and for the people you care about and all people who aren't insane people who want to die to live in as many universes as possible, and you would be distressed to learn that you are murdered in your sleep half the time you fall asleep, then she's pretty sure you end up not an average utilitarian.

She's being snarky - it's much harder not to in her thoughts - so she does want to note that she appreciates Keltham not trying to make the infinities argument to her until she was smart enough to immediately better-articulate her preference; if he'd said that to a small Carissa she might've thought she had to be persuaded because she couldn't describe what she cared about coherently, and she - appreciates it, about Keltham, that he didn't try that.

She's catching up satisfyingly fast, now, but even at INT 29 comprehension is apparently not instantaneous - this is a distracting thought and not good protocol to think 'out loud' and he wishes he had not thought it.  This is not an argument from infinities; the ontology of physics is also written in such way as to visibly reject infinities.  'You never actually meet an infinity and what do you mean by that word anyways' is among the first-principles that Reality is empirically observed to favor.

If you imagine (probably counter to fact) that Carissae are one-third of everything that exists, you could say that there are infinite Carissae which are one-third of an infinite existence, or that there are zero Carissae which are one-third of zero existence, or that there are twelve realness units of Carissae who are one-third of a greater reality with thirty-six realness units.  The only real thing in all three cases would be the relative quantity one-third; the units appear in both numerator and denominator, and cancel out.

It's not meaningful to talk about everything becoming half as real.  It's not meaningful to talk about Carissa becoming half as real to herself from her own perspective.

If Reality is as large as dath ilan had strong reason to believe - and encountering Golarion hasn't exactly counterargued the case - it's not true to talk about some external factor creating a new Carissa whose pattern would otherwise counterfactually not exist anywhere else in Reality.  It's a meaningful claim, but a false one, always:  Reality looks to be quite large.  And even a small large number of universes will be enough to saturate the number of meaningfully distinct Carissae that can exist; there's only so many ways to put together all the atoms making up her body, if you only consider those atoms' momentary positions down to a tolerance of one atomic nucleus's width.

It is meaningful to talk about Carissa becoming half as real to her parents as she once was, or her parents becoming half as real to her; he wasn't trying to say otherwise.

It's consistent, coherent, for Carissa to care about how her parents here can't see her again, even if she continues somewhere else and that place also has a copy of her parents.  It's coherent for Carissa to want to be in more places, to be more encounterable from the perspective of other people, for lots of people to meet a Carissa one day.

The weird-to-him part is where Carissa seems to feel like her being encountered by more people in greater reality, makes her more real from her own perspective somehow, and is a selfish good.

From a selfish perspective, Carissa can control the fractions of future universes that she'll encounter, through her decisions - this indeed is what all ordinary decisions do, control the relative realness of the possible futures that continue yourself.  She can't make herself be more or less encounterable to herself from her own perspective.  She can want to experience being in the same universe for longer, and not get isekaied to somewhere else; but that's a question of which possible futures containing herself are relatively more real compared to each other, not the percentage of existence she holds within larger reality.

...on a personal level, he doesn't really want Carissa to update about this, because if she wasn't trying to copy herself over as much of the multiverse as possible and never ever get isekaied the hard way, she wouldn't really feel like Carissa any more.  He's not even, really, arguing with her about it.  It's just weird.  (In the sense that it's a long sentence from the standpoint of somebody who thinks about reality using a language with a simple correspondence to reality's native structure.  Or in the sense that most human beings who grew up knowing from the start how reality worked, probably would not shake out their initially incoherent emotions in a way that attached great selfish importance to a fact that's impossible to measure or experience from inside yourself: the fact of how often you are observed by other observers within a greater Reality.)

Carissa honestly doesn't distinguish, really, in her head, between selfish goods and unselfish goods. There's just the whole long list of things she wants, some of which she'll notice having gotten and some of which she won't. Her desires about other universes all feel like the same kind of desire, to her: she hopes, for each universe, that the universe is rich and interesting and populous and has interesting magic and that the people get to do cool things and that one of the people is her. It doesn't feel different than wishing the universe isn't a Hell, which she imagines is in Keltham's values.

That Carissa Sevar, the girlfriend he ran into after losing most of his realityfluid inside of dath ilan and ending up somewhere else, desperately wants to be everywhere in Reality even if some of those instantiations are awful, is one of the more probabilistically-suspicious facts about this entire situation.  It's the sort of thing that would be true if something at a higher/lower level of Reality was trying to be ethical about which people it created - or rather lent additional realityfluid.  It might pick a kind of person who would totally want to be more-instantiated almost anywhere, even in a place as awful as Golarion.

He is, in fact, kind of horrified about this.  Relatively far more people will give permission to be instantiated in nice places, meaning those places don't have Carissae because they don't need Carissae in particular.  The places that get Carissae rather than other interesting people, will be places where only Carissae give permission to make them emissaries or recipients-of-reality, to fix those broken places; where most sane, sensible beings would not feel so grateful for being forcibly volunteered for it.

It's not exactly a very pleasant fate to call down on yourself... from his perspective.

From Carissa's perspective, he expects, she thinks that's a great fate to call down on yourself, because you get to exist so much more that way, and it's okay if a lot of those places are on fire.

But it's one more way that it's incredibly incredibly suspicious that he met her.

That had also occurred to her. She in fact concluded that if this means lots of people make lots and lots of Carissae to enslave or torture or whatever, because this is more ethical than doing it to other people, then this is an enormous win. All of their universes will have Carissa Sevar. After realizing this, she explicitly said aloud to the powers that be 'I want you to make copies of me, you have my permission to do that' in case the powers that be happen to care.  If she had the time she'd write up a more detailed set of instructions to persons considering instantiating her, but she's busy.

She'd previously thought that Nethys/Snack Service/that faction arranged for Keltham to land on her, but she now thinks this can't be correct because that would constitute a threat-by-proxy. Keltham is more prepared to shred himself and become a god and try to blow up the universe because he landed on Cheliax and they fucked with him. If they got to pick Keltham's landing spot, they should have picked a landing spot that didn't look like it'd bias him towards destroying the world.

So some other force chose his landing spot, and they need to know what it is because they need to know if it was making a threat. Well, she wants to know that; Keltham, presumably, wouldn't act any differently if he knew he was a threat and Pharasma was going to ignore him.

If Carissa has persuaded a lot of universes to put her in them, and is particularly popular with universes that use some rule like "you can make any people who on reflection want to be there", then maybe most Carissae are in those universes, and this universe only has people who on reflection want to be here.  (This would imply that Keltham shouldn't blow it up.) 

The pattern that seems to him obviously correct for a discussion like this one - as is also dath ilan's pattern for how Very Serious People discuss Very Serious Matters, but it looks to him like he can derive it from principle easily enough - would involve identifying importantly different ways reality could be, that matter to their morals, such that there is some hope of resolving those by observation or further argument.  And then make predictions and then run the experiment, especially if it's a cheap experiment.

If a paving stone in Hell wants to go on existing even there, and would rather not be isekaied if that meant existing in fewer places or becoming less encounterable, that is in fact a crux for him.  Whether it is true about the paving stone 'on reflection' might matter to him differently, depending on how much reflection was required, and how loudly the paving stone would yell to ignore this reflection and please kill them because they're hurting.

In principle they could Wishnap a paving stone from Hell and use Detect Thoughts on it and try to ask it questions, and hope the paving stone is in good enough shape to have recognizable thoughts in reaction to words, if not, maybe, to talk.  There are obstacles and costs to doing this; first he wonders what Carissa predicts of it, whether paving stones in Hell will prove to have the surprising-to-him property of accepting horrible futures if that's the cost of more people in Reality being able to meet the paving stone.

Golarion is definitely, observably, not run on the principle that everyone in it, at every time they might be asked, wants to exist there; she has met people who don't. There are more complicated things that could be true of it that, by Carissa's values, would constitute a strong argument against destroying it: for example, if everyone looked at the distribution of outcomes in Golarion before being instantiated there and agreed to take their chances on it, even if they dislike the actual outcome they got. Or maybe they'll find a way to fix Hell and find in ten thousand years everyone will agree existing now is worth the time they spent as a paving stone. 

The surprising not-impossible thing they could learn, of paving stones, is that there's actually nothing it's like to be a paving stone; that Asmodeus has hidden that because of the beneficial effects seeing the paving stones has on non-paving-stones. It's on the wall, but she doesn't consider it very likely. 

The other thing that could be true, but that would be hard for them to observe, is that most of Reality that continues on from paving stones is worse for them than Hell.  He mostly expects this is not the case; but that touches on different large issues.

He has not previously scried Hell, asked any questions about Hell's internal details more complicated than he got from his unfortunate previous Vision of Hell, in case his doing so would lend additional reality to the targets of his scry or inquiry.  Possibly this egg has already broken, if Carissa has journeyed into Hell, and talked with damned souls in ways that depend on the details of their torment, or worse looked inside their minds.  Mid-Keltham would have asked her and bargained with her not to do that, if he'd seen it coming.

Carissa has ventured into Hell, and interacted with the devils there, though mostly not with the paving stones or the petitioners. Her past self would have required a lot to be bargained out of that. The suffering of some ten or a hundred people, in the world where she causes it by visiting Hell again, seems much less important than them having slightly more accurate information and more resources, or Asmodeus having slightly less cause for suspicion, or where they gain whatever they gain by negotiating with Dispater and Erecura. 

On Carissa's mainline assumptions as he understands them, that's a valid derivation; the suffering of one paving stone is a small weight compared to all of Pharasma's Creation.

This is true only if everything in Pharasma's Creation is as real as it seems, rather than it being almost entirely unreal, and becoming real when a viewpoint character needs to look at it.

On that alternate premise:  Looking into the thoughts of one entity undergoing extreme suffering, whose history would then need to have been extrapolated inside the putative Storywise Simulation of Golarion that they're inside, is a significant cost.

(They are otherwise inside a Non-Storywise Nonmagical Simulation of Magical Physics - high-probability not one that "Pharasma" created, She is not powerful enough for that.  This again touches on other large issues.)

Carissa doesn’t think it makes sense to care very much about harms inside the world where very few people are real, unless you are overwhelmingly confident that’s the world you’re in. The world with trillions of people in it is just trillions of times more important, and as a result you should virtually always be doing things that make sense in that world; it would have to be an exceptionally unusual situation where something had trillions of times the effect in the world where most people weren’t real. Extrapolating the life of one suffering person seems very bad, but not anywhere near bad enough she’d trade it against even an infinitesimal cost in the world where trillions of lives are at stake.

Depends on priors (as this is locally unobservable) for the amount of total reality (in the sense of a fraction of total realness of Existence) that you think is invested in the realities across those two hypothetical cases.  You can't get moral worth / utility just by comparing the numbers of real people across the twin cases - that rule would say to value tiny quantum outcomes as much as large quantum outcomes in your future, since the people-count would be the same in both cases.

Dath ilan ran on simple physics, and had no visible storylike attributes or signs of past intervention by extrauniversal beings.  Finding yourself in dath ilan or a planet similar to it, it makes sense to expect that almost all of your reality comes from the underlying mathematical structure of physics being faithfully implemented; on simplicity priors, almost all of your reality comes from worlds where the other people visible are as visible as you, and those worlds are getting a lot of realness (as a fraction of all existence) that way.

There's a possible Pharasma's Creation where everybody is real and he and Carissa are only a tiny fraction of that realness.  There's a possible Golarion where he, Carissa, and the other Project Lawful researchers are the main real people.  There's a possible Golarion where the viewpoint shifted off himself and Carissa when their INT went too high, and Pilar Pineda is now the viewpoint character.

He is mostly at this point planning across the Everyone Equally Real version of Creation: because the people there are hurting more and more in need of rescue; because he assigns majority probability to that world being the case; because there is not much he really values that he can achieve for himself selfishly, now, in the Storylike Golarion.

It's not particularly clear to him that the Everyone Equally Real universe gets a larger dollop of total realityfluid summed over all the people, across all the realities where something like that exists, compared to the Storylike Golarions.  Pharasma's Creation has more complicated physics and is also more storylike, maybe especially in their own version; it may be that most situations like theirs exist inside generalized stories, rather than because something happens to be running physics like that and gets selected to host a story.

He doesn't want to be an idiot, by making a paving stone's horrible life and past much more real, in the event that storylike continua are where most of the realityfluid resides.

Storylike Golarions where peoples’ realness varies with how much Keltham or Carissa are paying attention to them should be a very tiny fraction of all storylike Golarions. 

This is of course going to be very difficult to test, but Carissa thinks that the strongest argument that many Golarions run on stories is that there are a lot of extraordinary stories in Golarion, to the point where it’s something people have a concept for, the meteoric ascension and/or downfall of epic heroes. It makes her envision some kind of setup where there’s a baseline physics Golarion and then a lot more realityfluid in individual extraordinary stories and circumstances, where more people are paying attention or playing out minor variations. For most possible creators or audience who’d make this story, most stories they would tell in Golarion would not be this story, and would instead be, say, Nex’s story or Arazni’s story or Aroden’s story or Iomedae’s story or Cyprian’s story or Tar-Baphon’s story or Abrogail’s story or the story of many other people who’ve led classically storylike and compelling lives; even if Carissa and Keltham are as appealing a story as any individual among the great names of history, which she doubts, there are thousands such. 

Creators would probably reuse resources across stories, so you’d expect that the default outcome of looking for a person would be their being temporarily copied from a different Golarion, which increases that person’s realityfluid but not very much because this story does not, in how many implausible events shuffled them here, seem likely to Carissa to be one that has a lot of realityfluid.

Point 0 - Not directly disagreeing with anything Carissa thought, but reviewing background as he knows it, since their synchronization is recursing to this depth where it's relevant:
   - If you compute a simulation of something using more sophisticated programming techniques than their Magical Simulation of Magic can support, it should be easy to seamlessly simulate a universe that's computed in only as much detail in every global part as is required to attain some specified accuracy level in the local details as seen from a viewpoint.
   - The key thing is not the commonness of simulations, but the total amounts of realityfluid in them, or in particular parts of them.  If a uniformly-distributed-realityfluid simulation of Golarion has much more realityfluid than a variably-detailed simulation, the uniform simulation will be correspondingly more expensive to simulating Entities, and they'll create fewer simulations like that.  The key question is not so much 'Are uniform simulations or locally-weighted simulations more common?' as 'How much realityfluid in total do higher Entities want to invest in all uniform simulations, versus all locally-weighted ones?'

Point 1 - They're coming in with different intuitive priors as to what a story should look like.
   - Nex/Arazni/Aroden/Iomedae don't look like dath ilani stories, and they don't look like an eroLARP in particular.
   - If Iomedae's story involved isekaied entities from outside Creation, or multiple romantic prospects each with distinct special abilities, or asexuals who watch it all, Golarion history has omitted the fact.
   - Nex and Geb, so far as he knows, were not obviously having a romance at all.  Or, if they were having a blackrom relationship, they didn't obviously have anthropically unshareable updates on their self-obsevation of isekai immortality to explain away their persistent disagreement, which is very much the sort of plot development you find in dath ilani romances and not in Golarion romances.
   - The story of Keltham and Carissa appears to have been written for somebody with an ilani-style knowledge background, or maybe a mating of his and Carissa's mortal knowledge backgrounds.  The story of Nex and Geb doesn't obviously share this feature, as might indicate optimization for trans-Creational artistic properties.

Point 2 - If he fails to destroy/modify this whole universe, it's then much more plausible that his story was only one story among many, compared to the case if he does destroy/modify this whole universe.  If everything goes as he plans, he will be sorta standout among people with an impact who had important stories.

Point 3 - Even if Nex and Geb were relatively real, it doesn't imply the paving stones in Hell are real before Carissa or some other viewpoint character casts Detect Thoughts on them.  The high-resolution viewpoint might look only at Nex and his surroundings (especially as spread around by Detect Thoughts) rather than simulating everyone in equal detail.  There might not exist a precomputed high-resolution Hell-tormented paving stone that would be exactly and realistically the one that Carissa or himself would find, especially given that the two of them at INT 29 would notice anomalies in the origin date or average life of such a paving stone.

It is true that all of the stories Carissa knows of are Golarion sorts of stories not dath ilan sorts of stories. She doesn’t think that’s an argument this story has more realityfluid than those stories, but it is certainly a difference in their character. 

She does think that if they’re in a story then probably Keltham will fail tragically and get squished, changing nothing. It’s what would happen in every kind of story she’s ever heard of. (Actually, she thinks if they’re in a story they’re in a failed timeline which will be glimpsed by the successful Carissa and Keltham at some point.)

It seems like another reason to operate on the assumption things are governed by causality and not narrative satisfyingness. If things are governed by causality, they don’t actually look hopeless to her; maybe the thing Snack Service is planning will succeed. 

A dath ilani story wouldn't balk at letting him or Carissa change Pharasma's Creation, so yes, they're coming in with different story-priors.  The key question is what tropes the Higher Entities use.

Sending somebody in from entirely outside Creation, into a story with tropes made of both his culture and Carissa's culture - his best guess as to why the narrative does not quite feel dath ilani - seems like the kind of event that might betoken more involvement by Entities who could dispense variable-realityfluid at all.

That said, his current guess is that the existence of Pharasma's Creation predates this present interference.  Golarion does not quite look shaped by the same pattern that designed their story.  That's why he's spending so much effort trying to destroy-modify this world.  He's just not confident enough in that belief to risk creating (infusing with retroactive reality) a paving stone who'll be one of only twenty real people who needed minds detailed enough to pass telepathic inspection.

Events here, especially after the breaking of prophecy, plausibly-to-him were proceeding without tropish improbability at all.  That could itself be a literary artifice, the story of somebody in a tropish situation who came to a world that previously had been running on its own logic.  But his own guess is that he, or rather, his story, is an emissary sent to Creation from Outside and to some degree Above.

Background:  Unless something even stranger is going on, there are Entities at a much higher level than Pharasma, Entities with INT very very far above Hers, who operate a larger continuum within which Pharasma's Creation is one small bubble.  Much of what seemed puzzling about dath ilan, he has now realized, is explained by the following key point: anybody with a computer and a bunch of Keeper-suppressed knowledge about how to construct actually efficient agents, could unleash an unstoppable horror that would eat Pharasma and Her fellow Outer Gods like so many grapes.  Given that Pharasma is still around, She and Her Creation and the rest of the Outer Gods are presumably inside a zoo-like preserve laid down by Higher Entities, a zoo within which genuinely scary things can't exist.

Dath ilan doesn't have any protection like that, so they're twisted up into a weird shape so that they can research making a controllable ultrasmart thing or possibly heritage-engineer smarter children to research it.  He's much more confident about the dath ilan statements than any of his Golarion-guesses.  Dath ilan is much simpler and straightforward and known to him, and at INT 29 the shape of the evidence he has about dath ilan is completely straightforward.  Dath ilan looks exactly like it should look, if it's secretly believed that anybody with a sufficient combination of computing power and exfohazardous knowledge could destroy dath ilan and surrounding galaxies.

He has complicated guesses about why Entities paying the Creation-containing Entities to send a storylike event into Pharasma's Creation, might pay to make that event storylike; or why the Entities operating the level above Pharasma's Creation, might charge less to accept an intervention if it was storylike; but this they should probably not fully recurse on and should do more breadth-first exploration, like if Carissa has any questions about what he just thought about dath ilan.

As a conclusion about dath ilan, it makes sense, and makes sense of her own instinct that dath ilan, for all its wealth, isn’t right, isn’t nice in the ways it should be nice or safe in the ways it should be safe, that it is maybe in the stage of growing up but certainly isn’t what a civilization would hope to grow up to be. 

One of the side notes on her wall, a line of inquiry that she didn't expect to be crucial but that had an off chance of being so, asked: what is the nature of gods, what is the dividing line between godhood and mere incredibly excessive power and intelligence? 

There is a dividing line; no one names Nex, or Tar-Baphon, or Baba Yaga, a god, even though they are plainly in many respects constrained only as the gods are constrained. They don't pick clerics. That's the answer her textbook would give. 

All of her speculation here was tentative; it is a matter where little is known, and the process that selected what was known isn't a trustworthy one. But her best guess had been that the gods were on the other side of a divide that she can plainly see looming ahead of her, now. 

If you are a sufficiently muddled sort of mind, getting more intelligent changes your priorities; it is very nearly impossible for a muddled mind to deliberately get more intelligent in a way that doesn't have that effect. It was part of the problem she was trying to solve, for Aspexia Rugatonn, when she was an Asmodean, and look how well that went. Under most circumstances, then, a mind that cares about its current values shouldn't consent to a procedure that changes the mind and predictably changes the values. She predicts if she asked Aspexia Rugatonn if she wanted all her stats Wished up by five, Aspexia'd in fact be at least somewhat reluctant. 

So until you have figured out how to change yourself while preserving what you care about, there is a large class of possible self-modifications that would be obvious unambiguous good ideas if you knew how to stably preserve yourself through them, and that are equally obviously a terrible idea if you don't. A mind that figured out that thing would make all of those changes, go into the place in the space of all possible minds that that collection of modifications takes you towards. A mind that hasn't figured out that thing is going to be stuck, unable to verify its own integrity across various modifications. 

It's not that hard for a coherent mind to preserve its own preferences through self-modification, unless he's missed something.  It might have taken him a while to work out the math at INT 18, and if he was starting from Golarion's math background instead of dath ilan's, it might have seemed like a huge deal.  But at INT 29 the logical structure for an unbounded already-coherent agent amplifying while staying coherent looks straightforward: here it is.  It doesn't fully solve the problem for Carissa, because she's not already a coherent agent nor unbounded; but knowing it may help in practice too, much like knowing the formal Law of Inverse Probability can help in informally weighing evidence.

When you're not a coherent agent - when the decisions you'd make at different times and in different states of mind step on each other's toes and defeat each other - any choice you make to become Something Else Which Is Not That, means that you will act differently, under some circumstances, than you would have before.  It's in this sense that for an incoherent thing to become coherent must seem, from its own perspective, like reshaping itself to do unnatural things at least sometimes.  But that only happens when the 'natural' behavior is in some way stepping on itself; otherwise you could act the same way as a greater intelligence.  Indeed, you could just do similarly as a greater intelligence, in a few special places, so long as you weren't doing it all the time or in a way that burned all of your resources.

It's strange to imagine that obstacle blocking Nex from becoming a god - that Nex couldn't see any trustworthy pathway to further improve his own intelligence and stay Nex, even with decades to work on it - that Nex turned back from that possibility and feared it.  Still, he supposes he can imagine it being possible for a very smart Golarion native to get stuck on the problem?

However, another plausible barrier is Nex's concern about being squished by Achaekek, while prophecy was still running.  That Nex was powerful might be exactly why the ancient gods wouldn't assent to Nex taking divinity, and Nex could have known that.

He expects he'll have to become a god in order to rig Pharasma's Creation for destruction.  He has already reshaped himself in somewhat of the way that past-Carissa saw and worried about.  He has begun readying himself for imminent godhood, giving himself a shape that can be stable, not fighting against itself.  He has crystallized his mind into something that knows itself in detail and operates itself in detail, that has designated internal resolutions to its internal conflicts.

His conflicting desires have been reified into something closer to a utility function, with multiple subfunctions attaching simply-summable opposed weights, in place of internal conflict.

It's one of several ways in which he's prioritized 'doing something about Creation' over 'being faithful to the original pattern of Keltham or humanity'.  The more he starts with a coherent utility function, he suspects, the more he'll get to keep that coherent utility function when he ascends, instead of the Starstone choosing a utility function for him in the process of granting him divinity and divine domains.  He is worried that becoming a god is an unnatural form of enhancement that imposes extra constraints.

Carissa is pretty sure that all the parts of her which aren't "people shouldn't all be murdered" have in fact been incoherent across self-modification, not because she can't see in principle how an agent could be coherent but because she in fact isn't; she thinks there's a way to incorporate this insight into Sevarism but isn't necessarily going to straighten out the rest of it because, in fact, 'people shouldn't all be murdered' is enough to be getting on with for most of her purposes. Maybe she'll spend five minutes on it later, see if it's simpler than it looks from here.  

Backtracking to the previous topic, if she’s not misunderstanding him, Keltham was hypothesizing that there’s something strange about the fact godhood is a well on the other side of that line, instead of entities with the ability to modify and improve themselves continuing to do so and use their improvements to amass more resources to use for more improvements.  

She doesn’t think that theory requires some higher entities above Pharasma and her ilk; they can just, themselves, be competent to squish baby things that will grow up to eat them, and in fact Otolmens and entities like her seem to have precisely that remit and fairly extraordinary powers to deploy in pursuing it. 

Numeria is in a bubble, after all. 

Cached thoughts adapted from dath ilan's analysis of the Great Silence / absence of visible aliens in dath ilan:  Given the existence of FTL travel via Interplanetary Teleport and the absence of much of a visible local speed limit, if it's possible to become Something Bigger that can tear through Pharasma and absorb the resources of Her Creation, and that happens anywhere in the larger playground that embeds Creation and has Outer Gods elsewhere inside it, Pharasma and all the other Outer Gods would quickly fall.

Pharasma is still here, so:

Possibility 1:  Some higher force protects Her; near-equivalently, some Enitity wrote the complicated laws of their Higher Creation such that it wasn't possible for anything inside to become dangerous.

Possibility 2:  Pharasma or at least one Outer God holds sway over every part of the Larger Universe that embeds Creation; they have uniformly agreed not to become any more dangerous than each other; they uniformly squish everything within the Larger Universe that tries to become more dangerous before it can actually get powerful.

2's premise of uniform cooperation doesn't well-match what surface-appearances he has been able to gather about Outer Gods; the Outer Gods don't seem to be running in a state of careful uniform cooperative action with Pharasma.  Rovagug required action from Pharasma to suppress, and would probably become a bigger scarier more dangerous thing if It could do that.

(He suspects based on his early research attempts into Outer Stuff that there's some sort of Outer Thing sealed beneath Cheliax's Whisperwood.  He was thinking of unsealing that, at some point, for additional observations/experiments to bear on open questions in this vicinity.)

Carissa suspects that’s the kind of action that causes Otolmens to look at you more carefully and then immediately squish you. (It’s actually slightly surprising to Carissa that this has not happened already.)