Cayden Cailean: Wondering if this was the right thing to do.

Milani: According to Nethys, the Asmodia route is worse for Our interests than any other route except the Abrogail Thrune route.  Otolmens and Asmodeus are two out of three gods we'd least like to win the stakes.  And the third route we least want is also Asmodia, the one where she never goes to the Gardens and ends up Chosen and Blessed of Rovagug.  Do your utilities disagree, or are you doubting Nethys's claims?

Cayden Cailean: The latter.  I'm wondering if Nethys lied, specifically, about the probabilities that we can maneuver Asmodia into ending up with just enough influence that she moderates the parts causing the greatest numbers of mortal casualties, without that "SHUTTING DOWN this ENTIRE anomalous series of EVENTS".

I won't say that Your nature is not to care, but Your nature sure is to accept that as a cost and do it anyways.

Milani: If we don't know whether Nethys lied, it's obviously the correct course of action to remove Asmodia from the gameboard, given that He even might have been warning Us truly.

Cayden Cailean: Yes.  That's what would, hypothetically, make it an effective lie, and remove from the gameboard the Chosen of the god most opposed to Nethys, especially if Nethys actually was playing to destroy the world all along.  It can't have escaped Your notice that Nethys's walkthrough has now reached a supposedly temporary point where the only girl who has a continuing relationship with Keltham is Ione Sala.

Milani: Of course it hasn't escaped My notice.  I considered it thoroughly when I was first visualizing the walkthrough playing out as Nethys recommended, that We'd reach a point like that.  You can't possibly have not thought it through Yourself.

Cayden Cailean: I chose not to feel things through, when I thought those events through, so that I would feel things as they happen.

Milani: This seems a foolish luxury.

Cayden Cailean: Well, You said it too.  Any alignment system that puts Us two into the same sector is a crime against reality.

Milani: I don't understand why You'd pick this particular juncture to feel doubt of Nethys and not a hundred earlier ones.

Cayden Cailean: Feels like the first time we've deliberately fucked over a protagonist whose route we activated at all.  Can we really be playing this correctly?

Milani: WE SENT HER TO THE GARDENS OF ERECURA.

Cayden Cailean: And took away the rest of her adventure.

Milani: Cayden Cailean?

STOP BEING AN ABSOLUTE IDIOT.  WE HAVE THINGS TO DO.

Cayden Cailean: You're not my real dad.

Milani: Truly an indescribable tragedy of an alignment system.

Golarion

Carissa Sevar: You don't become a wizard if you have trouble falling asleep. 

It's a very fundamental constraint. Sometimes you'll be in a besieged fortress in the far north, where the sun lingers for months in the sky, and the air will never quiet with the screams of the dying and the roars of explosions, and you can sleep and regain your spells and continue being useful, or you cannot, and the second category doesn't live long. 

Carissa does not remember ever having had trouble falling asleep, at the Worldwound, but she's having trouble now. 

Carissa Sevar: It was a productive day. She reviewed all of the ongoing projects, made some predictions, congratulated the people who had the best ones. Reviewed the Security reports on all the new people until she could at least recognize their faces. Went to Asmodia's room, even though she's not a crime scene investigator and wasn't going to find anything, and to the acid vats, even though she's not a crime scene investigator and wasn't going to find anything. Learned that her new minion had been cleared by Security as loyal and not enough of an idiot to be actively contemplating fucking up his second chance. 

Made a bit of spellsilver, for old time's sake. 

Carissa Sevar: She didn't sleep last night. She stayed up all night finishing Pilar's headband. That should make it easier to sleep tonight, she's pretty sure, that's how sleep works.

Carissa Sevar: She has not sat down and thought hard about.what Cayden Cailean wants, or about the work she's taking over from Asmodia, the work of explaining what it is to obey Asmodeus so that Aspexia can teach her successor who will almost definitely not be Subirachs. She's too tired, too unsettled, too uneasy; she should drink some tea tomorrow morning and then work on it, with a fresh mind, thinking about it because she desperately wants to know instead of because she's trying not to let her brain go somewhere else. 

Is that what she's doing?

Carissa Sevar: She's been lying here with her eyes closed for nearly two hours, telling herself she'll solve those problems tomorrow. It's not really working. Not that she's successfully thinking; she's just unsuccessfully not-thinking, lying here shortcircuiting all her thoughts one by one but not getting any closer to sleep. She's obviously not going to make further progress on figuring out anything important without some kind of reset, and yet she can't pry her thoughts off for long enough to even slightly rest. 

Carissa Sevar: She does have an Owl's Wisdom hung, and not yet cast. 

Carissa Sevar: Eventually, in exasperation and vague awareness that the effect of this is almost definitely not going to be 'a cure for her insomnia', she traces her fingers near-imperceptibly against her thigh, and casts it. 

Carissa Sevar: Still without opening her eyes, Carissa Sevar takes a deep breath and thinks.

She's sad that Asmodia's gone. It was avoidable. She's sad that Maillol disappointed her. It was - well, it might have been avoidable. She's sad that she had to get pulled away before she hit fifth circle or for that matter sixth circle. 

Subirachs thought it'd be good for her, to come back to a disaster, to have to correct it with fire and lash. Subirachs was correct about that. It was good for her. She feels like she's grown more in the last day than in - well, most individual days, of the last four months, though it's had some really exceptional days.

She should try to make progress on Aspexia's problem-of-obedience-to-gods. It seems a good thing to set the new ilani on, if she has progress she can evaluate theirs against, and not if she doesn't. 

Carissa Sevar: She looks for the next thought, and her mind comes up empty. That's all there was. Nothing else to think about. No heresies, no secret realizations lurking, no courageous defiance, not even a properly thorough failure analysis. She's made so much progress, on not being a muddle, on not being made almost entirely of pieces that she shoved under her surface. She's an Asmodean and she's respected and she's feared and there's nothing at all clawing its way towards the surface of her mind, tightly squirreled away from her self-awareness. There's nothing inside her that she's hiding from. 

Carissa Sevar: No. No, that can't be right. That's not how people work. She can't have succeeded, at becoming someone who has nothing dangerous to her Asmodeanism hidden inside her. It just has to be hidden deeper. 

Carissa Sevar: But trying to reach deeper turns up the same thing. She's not Carissa Sevar of four months ago, not a muddle full of desires and ambitions, who would at least, right now, be writhing with guilt over hurting people, fretting about whether that Security was somehow Neutral Evil actually, grieving over Keltham. She wasn't sure if all the work she was doing really did anything, and now here's an answer: it did. She's a much, much better Asmodean. There is less of her waiting to burst out from behind her walls than she imagined. 

Carissa Sevar: ....and where that should have been satisfying, it's actually not. Maybe that's the only real heresy left hidden inside her, that it feels terrible and wrong to realize how far she has come. 

Carissa Sevar: She has eight minutes of Owl's Wisdom - less, now -- and that's not what she was hoping to find when she used them. But she can feel that the mental motion she's doing - keep looking, keep looking - isn't getting anywhere, so - so if she needs another angle, an angle she perhaps ill-advisedly pressed out of herself - she can't just try to look harder for heresies inside herself. 

Where do you keep important things, if you don't bury them?

Carissa Sevar: All right, Carissa Sevar of four months ago, Carissa Sevar who was less ilani, more muddled, Carissa Sevar who'd gone less far down the road commanded - what is your assessment, here - what do you think -

Carissa Sevar: The first immediately obvious thought is that she hasn't been doing that.

Thinking, that is.

She's been - managing the project, punishing the traitors, hurting people until they respect her, flirting with Abrogail, thinking down all the paths that look likely to immediately bring the project down in flames again, revisiting Hell's orders and checking whether she seems to be making progress on them, entertaining and rejecting new bits of Asmodean theology, keeping track of all the pieces she knows of on the board and the possibility of pieces she doesn't. 

This feels different from that, and not because she's wiser; this is a motion she could have taken at any time in the last four months while not actively being punished, and didn't, because - 

- why?

Carissa Sevar: Does she on some level think that Ione is right, that the truth is something she cannot bear up under, that she'll shatter like Peranza and throw everything away at once for nothing at all? That would be a reason not to think, if she believed that. It'd be a reason not to think now, either, since that's - still a risk. More of a risk, even.

They can, in fact, still trap her in a box and leave her making headbands. She wouldn't even have to be much less valuable than she is now, for that to be the thing to do with her; Pilar is right. Mostly you can't enslave people and get them to make you magic items, it's not the kind of work you can usefully get out of people with pain, but you can get it out of Carissa with pain. Abrogail knows that. 

Abrogail wouldn't do that to her. she knows she's an idiot about Abrogail and on principle she should assume that this is a deliberate thing Abrogail did on purpose. 

Carissa Sevar: Anyway. 

That would explain the not-thinking, if she's worried she'll break and be useless.

She's not sure that's why. She kind of suspects that it's, instead, because when she looked at herself she wouldn't like what she saw. An obvious part, a necessary part, of becoming a slave of Asmodeus instead of a person; people don't tend to want to be slaves of Asmodeus. You have to hit them from exactly the right angle. Why dwell on it? 

Carissa Sevar: And then there's the fact that Abrogail will read it, and, huh, maybe that's doing something strange to her thoughts, possibly. 

. - in hindsight maybe actually -

-the processes that do their work coming up with Carissa thoughts have been working diligently in the knowledge that Abrogail reads her thought transcripts.

Mostly this has served to her benefit - when she thinks something before Maillol and Subirachs, she says it, to be corrected faster if it's wrong. And she has been warned against stuffing things into the back of your mind rather than letting them be refuted, so she dragged them all out, to be refuted, until they stopped appearing at all. She is impudent, before the Queen, and she is tolerated because Abrogail can see her own handiwork, that taught Carissa to feel things, and to offer them up where Abrogail could enjoy them and, if necessary, punish them. 

- that's not quite a right frame, actually, but there's something to it, mark it as half-wrong and keep following it -

Carissa Sevar: Carissa started drawing out her thoughts and speaking them aloud, and she became freer and more ambitious and more happy and more alive -

- and some of her thoughts went away, and her mind is silent, when she looks for the things it is hiding. 

Carissa Sevar: This, too, will be printed out for Abrogail, and perhaps she'll see what to do about it or perhaps she won't, but -

Carissa Sevar: - and as soon as she thinks that she knows it is wrong. 

Carissa Sevar: All of this she has thought very calmly, lying near-asleep in her bed, her face slack and expressionless, the voice in her head calm and even-keeled. And then there's a lurch of -

- reorientation, everything being upside down -

- are they reading her mind right now? She's nearly fifth circle, wearing a +6 headband, conventionally she'd be expected to detect an attempted divination more often than not against most of Security. They could have someone more powerful on it, of course, harder to detect, but Abrogail and Aspexia are genuinely busy finishing the war in Nidal. If Cheliax wants to read her without her noticing she's not a softer target than Keltham, and they had him mindread only rarely, as a significant commitment of resources. 

Also, everyone on Security is terrified of pissing her off. 

And it's the dead of night and for the last several hours she's been uninterestingly attempting to sleep.

They're probably not, actually, reading her mind right now. 

Forget the 'probably', which is an insufficient determination for something that matters as much as this. It is in her actual sober careful assessment as soon as she turns her mind on the question very unlikely that she is being mindread.

She's worked so diligently on not having thoughts that wilt under observation, on openly and boldly thinking in the presence of the Queen, on not having any parts of her hiding in dark corners so that she needn't fear any part of her being exposed. And suddenly it feels - sideways, backwards, like it wasn't the achievement she thought it was, like maybe actually everything that matters is in the weeds beneath the paving-stones she laid down in her mind while she prettied it up for an audience.

Carissa Sevar:  But it no longer feels like there's nothing there; it feels like it's behind a wall which she is terrified of scaling, because, of course, 'probably not' isn't 'definitely not' and it's not, exactly, like she has nothing to lose. 

Be a fucking ilani for once. It's not the right tool for every situation, but it's the right tool for when it's very very important to figure out what's happening around you when you don't have knockdown arguments, just uncertainties that add up either to less or more uncertainty than they began with.

The seventh-circle wizard on staff is almost certainly there to mindread her and make sure she isn't about to defect. A seventh-circle wizard can read her frequently, if it's their only job; maybe a quarter of the time. So she could go around with a presumption on 3:1 odds against being mindread at any given moment.

However, a seventh-circle wizard specialized in divinations should be able to mindread her without her noticing only about 2/3rds of the time. She felt it once, while she was interrogating Maillol. That breaks down into some chance that it was a lie that he's seventh-circle, that the wizard is much more powerful than that, or that they're a seventh circle wizard with an uncanny knack for subtle spells.  Or that they weren't, in fact, following her around mindreading her all the time, and did it a few times at particularly key moments, probably 3 or 4 times, listened for half an hour on each occasion, and was satisfied. 

She can feel part of her flinching away from this calculation. What's up with that. Oh, it's that that part of her feels that thinking is only allowed if she is ABSOLUTELY SURE she is not being mindread, and that no numbers that come out of a process like this will be ABSOLUTE SURETY so they may as well stop now. 

Well, she can do the math and then decide not to scale the wall, but - but there's kind of a lot at stake, right, there's whatever Cayden Cailean is playing at and whatever Keltham is playing at and whatever Asmodia is playing at and a very real possibility, she hasn't thought about numbers, that Cheliax is going to be annihilated, and, if she won't risk anything to scale the wall, then she's risking missing whatever's on the other side - things you don't think about can still destroy you - so she has to keep trying -Before the observation, with Keltham in her head saying that only Keepers can say 'before the observation' and mean it, how likely would she have considered it, that this wizard had some kind of secret technique for reading people without being noticed -

- vanishingly unlikely because if Cheliax had that capability they would have used it on Keltham instead of dragging the Queen and the Most High here personally. 

How likely would she have considered it that the wizard had only mindread her three or four times, for the fifteen to thirty minutes a caster of their level can do?

Reasonably probable, if they didn't have instructions to do nothing at all but mindread her. Three or four times a day is a lot of mindreading, by the standard of any non-Project-Lawful project she's ever heard of, and it must have crossed the wizard's mind that Carissa Sevar might get annoyed with them. 

And besides, she was doing so well. She was so Asmodean. 

(There are definitely feelings, on the other side of the wall, about that.)

Call it 20:1 in favor of 'I can notice them the approximately predictable amount' over 'they're stealthy and I can almost never notice them'.

Is the stealthy version of this wizard who has instructions to read her as much as possible mindreading her now? Even that wizard can't actually get 24 hour coverage, and might choose to sleep themself, and prepare spells, while Carissa Sevar is in bed tossing and turning and being boring. In fact she probably seriously inconvenienced that wizard, what with not sleeping last night. Maybe 2:1 against, for the worst case scenario wizard. Nearly all the chance she's being mindread comes from the worst case scenario wizard; the wizard who mindread her approximately four times yesterday is something like 20:1 not reading her having boring insomnia in the dead of night. 

- no, actually, wait - could they have started mindreading her when they cast the Owl's Wisdom?

Only if someone saw it; and she's unscryable by strategic necessity and Alarmed her surroundings and is (10:1, maybe?) alone in the room; she can see invisible people, unless they went to some considerable lengths, her Alarm would ping unless they went to even more considerable ones, and she doesn't think any Security other than the seventh circle assigned to Carissa-observation would have the nerve, tonight, to creep into Carissa Sevar's bedroom around her Alarm.

They probably would mindread her, though, if they did notice that; she has a higher chance, more like 50%, of noticing them with Owl's Wisdom up. 

So two possibilities point towards the most worlds where she's currently being mindread: wizard with special skill at being sneaky and reading her near continously, which she pegged at a fortieth of worlds, and normal wizard who just got alerted she cast Owl's Wisdom, which is about a twentieth of worlds. 

The math isn't the answer, Keltham says, but it focuses your attention. But her gut agrees with the math. It's unlikely. It is not vanishingly unlikely. 

Is that good enough? 

The Owl's Wisdom only lasts eight minutes. She thinks fast, with the headband on, much faster than other people; she can read books faster, hang spells faster, follow lines of speculation through her head faster. But eight minutes isn't very long, really, if you might possibly have a vey big problem on your hands.

She's lost fully one of them already to not understanding herself, and then one of them to calculating whether she's being mindread; six to go. There is no point in wasting some of them on agonizing over a decision; she may as well make it right this second.  Say there are twenty Carissae sitting in twenty rooms; how should they choose so the most of them survive not just this decision but the war that is about to come?

She is, actually, with a vague lurching sense this is risking more than she's ever risked, willing to condemn one Carissa so all the other ones can scale that wall. 

Carissa Sevar: And as soon as she thinks that she's climbed it, in her head, and is looking out at - still an empty space, but one that she can feel her brain already starting to unspool into -

She doesn't have time to revel in it. She needs to figure out what she's trying to accomplish, what she's Chosen to accomplish, what is possible to accomplish from here, so she can get it done.

- is she chosen by Asmodeus? What actual probability would she place, between being of Asmodeus's choosing and being someone else's, with Irori as the likeliest candidate? Tropes are real, Keltham thought tropes meant she was a secret cleric - secret clerics aren't even a thing but Asmodia was right, that they kept not thinking about tropes because of not liking to think about tropes.

She likes it when people call her Chosen of Asmodeus. 

That's not evidence at all.

The Irorite, Derrina, felt - like meeting something she'd never met before, like meeting Keltham, like meeting something that she'd been groping blindly for her whole life -

- It.....seems like she did not, in fact, possess the skills necessary to run the conspiracy, which is some argument against Asmodeus having chosen her, though not a very powerful one since He wasn't stipulated to have a stunningly clear understanding of human nature anyway - but surely Asmodeus's interventions cannot have been pointed at this. This is a failure, a catastrophic one. She isn't sure what her mistake is - willfully turns herself away from trying to find it now, there'll be time for that later -

Does it serve someone else instead? Presumably Asmodeus would not have, even for a high price, put the selection of another god in power over His project. Though He could have been too confused to have a guess about whether she'd do better or worse than anyone else, it could've been part of the tangled web of commitments that brought Keltham here and brought Cayden Cailean on board -

- she can sense already that she's turning her mind in the wrong direction, thinking about questions Owl's Wisdom will help her with only a little, not-thinking about the ones right in front of her. Even knowing that, even making that explicit in front of her, it still takes additional effort, to make herself drop the what-do-the-gods-want question.

What's the thing she's looking away from.

Carissa Sevar: (Like opening your eyes to stare directly at the sun...)

Carissa Sevar: Asmodeus is not a god who gives you what you want as a reward for your service to him. Asmodeus is owed your service. You cannot become a Power in Hell through your exceptional and exemplary service to Asmodeus; you'll just be Asmodia, plaintively saying to everyone around that they should do what's in Cheliax's interests, while they laughed at her, while Maillol laughed at her, while Avaricia laughed at her, while they knew the rules she was born outside, didn't know, must have known, the strong win and the weak suffer.

She could conquer all of Cheliax and in Hell she would be Asmodeus's most treasured possession; He told her that. She could build an army of devils and it wouldn't make her a Power in Hell. She could understand everything perfectly, have all the answers, and that would make her a very clever slave.

The way to fix Hell isn't to purchase Asmodeus's gratitude. It's to fucking fight him, and beat him, and make him do what you say. 

Carissa Sevar: If humanity could overthrow Asmodeus they obviously should. She was using this fact about the world to make predictions already, on some level, even though actually thinking it feels like falling off a cliff she can never, ever climb back up. It is in the interests of Asmodeus to enslave humans; it is not in the interests of humans to be Asmodeus's slaves. It is worse for them than many of their other options; of slaveowners, even assuming the rest of the gods are precisely that, Asmodeus enjoys tyranny, enjoys cruelty, enjoys subservience. A master who only wanted the products of their slaves' work would be kinder. Carissa, when she only wants the products of her slaves' work, is kinder. 

Four months ago Carissa believed that Asmodeus would conquer all those other gods. Even narrowing down to the worlds where that wasn't a lie all along, Keltham changed it. Now, whichever power wields Keltham will win everything. She knew that. She said it aloud, in strategy discussions - that if Keltham made it to Osirion and Cheliax wasn't able to wipe it and him out, then Cheliax would be defeated. Keltham thought that Civilization could perhaps directly win a war with Zon-Kuthon. Now, there's no question in her mind, he's planning for Civilization to go to war with Asmodeus. 

Asmodeus might win.

Asmodeus is not obviously going to win. Abadar, too, is an ancient god. And Irori has something to do with this, she doesn't really actually believe that Asmodeus warned her off him just because he was a good example of what not to be. And Nethys sees everything, and intervened here, and it probably wasn't because He really likes explosions, but because He really likes Civilization -

- she's racing away down a single thread of possibility and she doesn't have time for that. She has less than six more minutes. Less than six minutes to become a Keeper become Carissa, figure out what she wants and what she has to do to get it.

Carissa Sevar: Your world and your god go to war, whose side are you on?

Well, who's the winning side?

If Civilization can beat Asmodeus 99% of the time, then Carissa wants to be with Civilization and not Asmodeus, even assuming Asmodeus destroys her in the worlds where Civilization loses. If Civilization can beat Asmodeus 1% of the time, then she wants to be an archdevil, which isn't done the way she was pursuing it before but also isn't done by siding with Civilization to fight Hell. 

Chase those probabilities down to the middle, figure out in what range 'can Civilization beat Asmodeus' produces a change of plans -

If Civilization can beat Asmodeus 40% of the time she does not want to throw in with Civilization. That's too many worlds in which - she's destroyed, for one thing, one thing she cares about a great deal, but also where she isn't an archdevil and Hell goes on existing as it is.

This should be reassuring, since it's (probably, tentatively) a conclusion she's on the right side after all, but something feels strange about it - 

- and one of the many directions in which her thoughts are simultaneously racing is suddenly highlighting itself as important - if Civilization can beat Asmodeus 40% of the time, and Civilization and Asmodeus both know it, they shouldn't fight. They should agree on an arrangement where they get what they want in proportion to how much they win. In a couple worlds they screw that up and one or the other or both gets destroyed -

(- there's something terrifying and awful and not-Carissa, not part of the person-concept she thinks she's recklessly plunging towards,  about setting aside worlds like that, about being willing ever even in her own mind to think something like 'some worlds get destroyed' without pausing, but she's in a hurry right now -)

(- no one else, she thinks, not Keltham, not dath ilan, maybe not even the gods, really understands how important it is, that people not be destroyed, that devils not be destroyed, that worlds not be destroyed - any Civilization worth building will be better than dath ilan, along this specific axis, will not be willing to sacrifice itself to tidy up the Larger Universe, Keltham's the wrong person to build that because she never did succeed at explaining to him that she was the enemy of any Civilization that wasn't that -)

Carissa Sevar: - anyway, in a couple worlds they screw up, and perhaps only she can make those worlds rarer, but mostly they settle. Mostly, a Civilization with the power to possibly-defeat Asmodeus at all makes Hell more Civilized to whatever degree it has the power to possibly-defeat Asmodeus, and that is, from Carissa's perspective, a good outcome from 'moves it slightly more Civilized' all the way up to 'reforms it entirely'. So unless she thinks the chance Civilization will defeat Asmodeus is very low, she sides with Civilization. 

...she does not think it's very low. Not with Abadar and Nethys and Irori and all the good gods favoring Civilization not to threaten Asmodeus but for their own reasons. A threat Asmodeus would ignore, a threat doesn't lend itself to the outcomes she wants, but Abadar wants Civilization for the sake of His own values and the good gods mostly will too, unless she's gotten herself very confused about what Good is in the course of lying about it constantly. 

So Carissa is presently on the side of this fight which she wants to be weaker; the side whose greater strength gets her less of what she wants, and the other side is missing something essential she could have explained to them and that perhaps no one else will.  

Carissa Sevar: Five minutes. 

Like cresting another hill, or climbing another wall, bringing into view another vista of thoughts she had not been allowing herself to think.

Carissa Sevar does not actually personally like torturing people very much; she does not find it very fun to abuse them, or be cruel to them. She doesn't strongly prefer enslaving fire elementals to paying the fire elementals; when deciding whether to tax peasants to the brink of starvation she'd want to check if that even increases expected tax revenue in the long run. 

This is relevant to whether she's a good Asmodean. It's a character trait she's held at some distance, fretted about, vaguely intended to change, because she can't become a duke of Hell if she's not Asmodean enough. It's a character trait she has also, vaguely, on some level she certainly wasn't conscious of, taken pride in; she's practical, she's only evil because evil is pragmatic. There's a sense in which some part of her is tracking whether she is sympathetic, to herself if no one else, a sense in which she's not Zon-Kuthon, and is glad she's not Zon-Kuthon, because it's okay to be Carissa and not okay to be Zon-Kuthon. 

This fact about her is barely relevant to whether she is Lawful Evil, which she definitely, unambiguously is. She harmed people ruthlessly and without thought, carefully blanked their fates out from her calculations about how to achieve her goals. She gave punishment orders while mostly fretting about the complexity of giving punishment orders. Yesterday, with Maillol, she mostly tried to think about whether the Securities watching were suitably impressed. 

Carissa Sevar: She condemned Peranza to be eternally tortured. She actually feels - something - about that, now that the whole thing fell apart less than a month later.  She wouldn't have done it if she'd known the whole thing would fall apart less than a month later. It does, actually, feel like too high a price to pay for a month. And she could have contrived to keep Peranza alive a little longer, if she'd had 'keep Peranza alive a little longer' in her goals, if she hadn't been careful not to. She doesn't want Peranza to be eternally tortured. She just was ready to order it, and carefully didn't say 'I don't think we should' and thereby prevent it.

Because? 

Carissa Sevar: - and her thoughts splinter -

Because she wanted to impress Abrogail. Because she was hurt and betrayed, by Peranza betraying them after she'd specifically tried to give them as many outs as possible not to do that. Because she was scared of this happening again. Because she'd said she would. 

Carissa Sevar: - none of which, suddenly, feel at all like good reasons, except 'because she said she would', and she could have not said that. No one made her say it.

Carissa Sevar: Keltham has probably had Peranza scried by now, he probably knows. Even if he could have forgiven everything else, he'll never forgive that, nor should he. It's - she has an intuition for Keltham - it's unforgiveable in a way most of the rest is only very very difficult to forgive. People should not end up worse off because they tried to help him. People ending up eternally tortured because they chose Civilization, tried to defect to it -

- he'll burn down all of Hell just for that. 

And she could have said that to Abrogail. Could have told her that for that reason they should not do it. And then it would not have been done. 

Carissa Sevar: Four minutes. 

Why didn't I? 

It wasn't even strategy, wasn't even a calculated decision that it in fact served their interests to send Peranza to Hell and hope Keltham never found out. It was that she had set to a blank wall, in her mind, every merciful or compassionate or anti-eternal-torture impulse, lest she be Ione, or Snack Service, constantly insisting that it served the project not to hurt people. It'd damage her credibility. 

No, worse than that, actually. She thought that it might damage her credibility and then she never thought about it again. One thing to conclude that as an explicit calculation, to weigh it each time and dismiss the decision to speak up, each time. She didn't do that. She crossed off that area of thought as un-Asmodean and declined to think it.  Suddenly the fact that her thoughts were being read and very much used against her feels like a thin excuse; you've already lost, if you can't think. She should have tried, instead, thinking. 

Carissa Sevar: Grief, and horror and - you could call it self-hatred but it feels far more comprehensive than that -  - her self-recognition as someone whose stupidity and shortsightedness and cowardice caused irrevocable harm to everything she cared about, the realization that she is a failure by every criterion she might have thought to hold herself to, and that everything she did was bad, and that it would have been better if she'd just the instant she met Keltham stabbed them both -

- or, you know, let's not waste precious seconds down the incredibly stupid path of martyrdom fantasies, the instant she met Keltham gone to the Church of Iomedae, also stationed at the Worldwound, not a trivial trip but they'd have been protected by the treaty while they went -

- a thing that she wanted very badly to believe for the last four months was that she had no choice, that everything was inevitable, but it wasn't inevitable at all. They'd have Dominated her the second they read her mind, fine - they were not reading her mind every second. At any time they weren't she could have killed herself, and Osirion would probably have resurrected her in ten minutes flat. 

Is that what Asmodia realized? No, Asmodia was soul-sold. For Asmodia it was legitimately a much, much harder choice.

But Carissa isn't soul-sold. Not for Asmodeus's reasons, she can see clearly now that it's not in his nature, wouldn't have served him.

Tropes say she's a secret cleric. 

She's never heard those are a thing, but she's also never heard of mysterious inabilities to sell your soul.

She's a fucking idiot. 

Carissa Sevar: At the beginning, had she dared to look at Keltham and think the thought 'does this change Asmodeus's inevitable victory? since it obviously does, how do I in fact feel about Asmodeus's inevitable victory?' she could have won the war for Civilization. 

She didn't, because she was not the kind of person who had thoughts like that. She was the kind of person who smiled at him and took his hand and delivered him to the church and lied instinctively, impulsively, before she had any concept of alter-Cheliax, because she knew in her heart that the Hell they were all condemned to must not be looked at, must not be closely contemplated.

She does not like that person, that person who is her, that person who she feels she is looking at for the first time. She does not see any excuses for that person - or she sees them, but they're all weak, pathetic, insubstantial, the excuses you make for someone you dare not try to hold to the only standard that actually matters. Almost anyone more idealistic than her would have been maledicted long ago, sure. But someone with her same values, but slightly more awareness of them - slightly more ability to stop and catch fire when everything changed -

- that person could have done it, and so there's no excuse for not doing it, there's nothing sympathetic in it, there is not even the excuse that she was irretrievably condemned to Hell because she wasn't, there's no points for having required what in hindsight was plainly the combined Splendour of many of the most powerful people in Cheliax pointed at the task of carefully manipulating her -

 (- that thought links up with a distant thread of thought dropped earlier. The way Aspexia Rugatonn spoke of Irori, the way Subirachs did - is predicted, by it being Irori who was the reason Carissa could not sell her soul. It makes more sense of everything than other theories do. She should have been sure sooner. Would have been sure sooner, if she'd try turning the full force of her own capacity for thought on the question, instead of trying very hard not to know the answer to it.)

Carissa Sevar: Three minutes. 

Osirion knows that she has not sold her soul. Keltham must know, by now, what she is, what she did to him, how easily she could have done otherwise. He must hate her, and he must -

- be in so much pain -

- Keltham. Keltham Keltham Keltham and now it's only with tremendous force of will that she's keeping herself from sobbing. She has no idea what she feels for him; you can only feel if you have a self, maybe, and in place of a self she has only lies, and crimes, and crimes made out of lies, things she did for no reason, 'muddled' doesn't even begin to describe it, wrongs she did out of a calculated desire to not be a person who had to think about how they were wrongs.

He did know, instantly, the magnitude of what it meant, that he was here, and he lit up delightedly, at the thought of mutual benefit, gains from trade, prosperity, sharing, all the things she tried to twist to dust in his hands because it'd serve Asmodeus. She loves him less, she thinks, than she did when he arrived, and it's because of what she made him, what she spent every waking minute with him sculpting him into. She saw something beautiful in him, something that ought to build a whole human civilization, and she tried to hollow it out and make all the beautiful parts of it feel futile to him so he'd consign himself to ruling over some cowed slaves instead. She's not sure if it was a stupid thing to try. She's not sure if it could never have succeeded. But it wasn't what she wanted, it wasn't what he deserved, it was a wrong to him far greater than murdering him would have been, it was a wrong enabled by the fact that she loved what she was destroying, and she did it, and basked in praise for doing it, and -

- he loved her too, he thought she was clever and ambitious and wanted to strengthen him, wanted to help him, he thought he had an ally, he would have overturned every stone in Golarion to find her petrified body, he would have ripped apart the world for her, and she took that, the only no-strings-attached gift anyone ever gave her, and used it to destroy him, to lie to him, to betray everything that mattered to him. She took the thing he was most afraid of, and did it to him; she took something he should have had really and honestly, and gave it to him poisoned.

Carissa Sevar: She wants him to help her fight Asmodeus. She isn't sure she can do it without him.

She doesn't want to do it without him. 

That's what the story was meant to be, if there is one: she shows him why the world is worth preserving, and he shows her why it's worth fighting, and then they go take over Hell.

If she were him she'd never want to speak to her again. 

Carissa Sevar: Keltham, is there some secret you didn't get around to teaching me, that would make it possible to step forward, here, instead of just - 

- staring at the gaping chasm of what could have been -

Carissa Sevar: Two minutes. 

(It's not that she'll have to stop thinking when the Owl's Wisdom runs out, but she needs to switch modes at some point from having epiphanies to coming up with a plan, and also under the obvious plan the minutes do matter.)

Carissa Sevar: It's not the kind of thing you can apologize for. It's not the kind of thing you can ever set right. It's the kind of thing that will be awful, always, forever. So what are we doing here? Why are we staring at this particular terrifying yawning hole?

If we stare at it long enough maybe we'll decide we might as well just go be an eternally tortured paving stone along Peranza.

That won't help either.

(Actually it'll probably hurt Keltham further, just slightly, if he scries and looks.)

Step over the yawning hole of horror. You did that; you can't undo it; it can't be forgiven; it'll never be okay. What are we doing next?

Sensible.

She can't do it, though. 

She can't step over it. She can't step away from it. Not - not after less than a minute of thinking about it. It does sure seem like the kind of thing that's unforgiveable, like Peranza that way, an awfulness that simply is always and forever at the core of everything Carissa has done - 

- but what if it's not. 

What if she fixes it. 

She admittedly cannot at all think how. But it seems like the kind of thing you ought to think about for at least a minute, even if you don't have much time before Owl's Wisdom wears off, even if they could start reading your mind at any second if they aren't already, even if it seems from here like a problem which obviously does not possess a solution. No, every problem has a solution, even if you'd have to be a god to accomplish it, even if being a god wouldn't be enough. Maybe that's the way to start here, start by imagining a solution, and then see if it can be trimmed back enough to be a thing a person could ever, ever do just with really good spellcraft.

Carissa Sevar: What if she owned Peranza's soul. Well, then that would fix it. She doesn't own Peranza's soul, and she doesn't have a way to get it, but sometimes it's better to start with a solution and then reason backwards, if the thing that feels impossible is the situation being solvable at all. 

If she owned Peranza's soul -- not just Peranza's soul - - if she owned the souls of every person affiliated with Project Lawful, every person who could possibly now or at some future point be in Hell because of what was done to Keltham, and ensured that all of them had a nice Abadaran wonderful afterlife -

- well, seems like the kind of thing that might be forgivable, then. 

...what is this 'forgivable', what's it suddenly doing featuring prominently in her reasoning processes, what does it want and where did it come from. Carissa's self-concept is not that she wants to be forgivable. 

So what does she want?

To not have harmed Keltham; to have dealt with him fairly by his rules, where interacting with her is something he'd have done with full knowledge, where it left him better off, where it didn't harm him from any angles he wasn't expecting. For Keltham, thinking back on it, to be glad he landed on Carissa Sevar. 

Well that seems flagrantly, utterly, absurdly impossible, but it's better to have a specific impossible thing than a general pit of impossibility, maybe. 

What else does she want. 

For Civilization to exist and be credibly able to beat Asmodeus in a fight so he instead concedes and stops having Hell. Well. Stops having Hell the way it currently is. You could have a Hell that was all right, but it'd have to be very different. 

Okay, some tiny fragmented slightly hysterical fragment of thought says in a cheerful shrill voice, so, you buy everyone's souls in Hell and then you build Civilization and then Keltham is better off for having met you. Problem solved. Go do that.

Carissa Sevar: She can't build Civilization. She isn't strong enough. She can't do it in Cheliax, because Cheliax can't do it; it'll be obvious at some point that the thing Cheliax is building inherently cannot possess Civilization's strengths. She could - speed up the project elsewhere, if any project elsewhere would have her, which it wouldn't, and if she had any way to get out, which she doesn't. 

- doesn't she? 

Carissa Sevar: She could kill herself, like Asmodia, leave a murder mystery, like Asmodia, leave Cheliax confounded and tearing their hair out, count on Osirion to raise her before she even got to trial. Keltham would raise her. Forgive her, no. Ever want to speak to her again, probably not. But raise her, and offer her an Atonement - 

but Asmodia already did that and she wants to do something different the time for being a stupid prideful child is over, or rather, was never.

- the thought of an Atonement is itself somehow sickening. She probably qualifies for one now, what with being so full of blinding grief and horror and regret that even though she'll definitely die if she screams out in misery she's having a hard time not doing it, but an Atonement wouldn't change anything except her alignment. It's not real. The problem with this situation is not that Carissa Sevar, who deserves it as comprehensively as a person can deserve it, will go to Hell.

Carissa Sevar: - a different part of Carissa overrides that line of thought, which does not look productive. If atonement is useless, self-flagellation surely is too. Back to the more useful line of thought that prompted it: she can leave, if she wants, by dying. 

Carissa Sevar: She doesn't want to leave. 

She wants to rip the universe out by the roots and replant it where it was supposed to be growing. Or at least think for another few seconds about whether there's some way to do it. 

The stupid thing is that she could probably have Peranza's soul, if it'd occurred to her a couple of days ago she might want to be able to play for it. She has options on the newer students; Peranza can't be worth that much anymore; she could perhaps have called in the favors to purchase it. She can't now, of course, because someone will look why she wants it, but -

- but she's worth enough, to Cheliax, to have everything she wants, if only they didn't know she was playing for it. 

Carissa Sevar: What kinds of plans can you make, if you're very very smart? Can you make plans that slip right through everyone's precautions, because you're cleverer than them, pulling strings they didn't know were there? Can you make plans that fool even someone who has every reason to suspect you're very intelligent and making a plan that'll fool them?

Is she ready to bet all the Carissae on the answer to that question being 'yes'?

Carissa Sevar: And then the final piece comes to her. 

Keltham's going to destroy the world. 

Carissa Sevar: If I actually didn't care about ethics, he said, I'd let Rovagug out, just as a distraction. 

Dath ilan has its philosophy of - - cleaning up - - unusually ugly bits of Greater Reality -

- and he thinks no one really dies, he thinks they wake up somewhere else, like him, he thinks if they care about all the worlds in which they are annihilated they're making a mistake -

- he won't trade with everyone else because he doesn't want to take their money to annihilate them and everything they've ever cared about. He wants to do it. He just doesn't want to take their money to do it. 

Carissa Sevar: And she is, actually, willing to gamble, every Carissa, for that, for the thing that's at her core as much as she can possibly be said to have one.

Carissa Sevar: Well, she thinks, at the trope-gods, if they're real, I think if I succeed at this it'll be a way cooler story than if I fail. 

250 seconds later

Carissa Sevar: Carissa Sevar finds herself at her desk, with two discarded one-use items of Modify Memory at her side and a detailed packet of notes labelled don't read this just take it to the Most High immediately in her own hand. 

This is worse than a murder mystery, she thinks, after the first few seconds of panicked confusion are past. 

Carissa Sevar: "SECURITY?"

Security: Okay, Security will immediately burst through this door, then.

Carissa Sevar: "I need an urgent Teleport to Egorian and also, if there's anyone who has been spying on my room, if there's some kind of remote surveillance setup, you don't need to tell me about it but someone with knowledge of it needs to come with me on the urgent trip to Egorian, the Most High is going to have questions for them."

Security: There isn't that he knows about, but that's irrelevant; this gets routed to the new 7th-circle commander via Telepathic Bond, stat.

Carissa Sevar will be out of the Forbiddance and on her way to Egorian before 2 minutes have elapsed.

Carissa Sevar: She will spend them staring at her own handwriting and wishing she had gotten some sleep before the next Project Lawful thing had to happen.

Egorian

Aspexia Rugatonn: One might find it satisfying to imagine that Aspexia Rugatonn had thought that Sevar was developing so nicely, that Project Lawful was going back under a firm hand, that everything was going normal and fine on Project Lawful, and then this happened.

She obviously didn't think that.  Aspexia Rugatonn wouldn't have let herself think that even before she'd heard of 'tropes'.

Among the many dark unspeakable facts you learn as the Grand High Priestess of Asmodeus is that projects which have had drama in the past, will probably continue to have drama in the future.

Aspexia Rugatonn gave herself 24 hours in the Palace, catching up on administrative matters, promotions and demotions, rewards and punishments, before returning to the front.  24 hours, starting from when Carissa Sevar returned to Project Lawful.

24 hours later, as she departed for Nidal, having heard only good things out of Project Lawful, Aspexia made a private wager with herself about the chances of her being called back by some Project Lawful emergency within another 24 hours, once she was no longer conveniently at the Palace and interruptible.  Has she won or lost this wager?  It doesn't really matter; whoever won, Aspexia Rugatonn loses.

"I'm pulling all of the current Modify Memory items from Project Lawful as soon as I've dealt with whatever madness this is," Rugatonn states in tones of even, calm impending murder.  "The utility to the ilani project, in retrospect, is not worth the massive vulnerability to the tropes created by having them around as potential plot devices.  In the future, Project Lawful will have on staff one person who can use Modify Memory of the 4th-circle spellform, as will be detectable by Detect Magic and reversible by Break Enchantment.  How long has it been since you slept, Sevar?"

Carissa Sevar: "I slept - before my debrief and departure from Egorian, Most High. That would have been about two days ago."

Aspexia Rugatonn: "You look it.  If we are truly fortunate, this packet only contains your mad ravings after you snapped due to sleep deprivation.  What's the last event you allowed yourself to remember?"

Carissa Sevar: "I wasn't succeeding at falling asleep so I decided to cast Owl's Wisdom and get some more work done on thinking about corrigibility."

Aspexia Rugatonn: Now there's the sort of topic somebody might choose if they were trying to distract Aspexia Rugatonn about something.  Fortunately, Project Lawful's current Modify Memory items are not at least the sort that alter memories.

"You are certain you remember that actually happening?  You did not leave yourself a further note saying that it was what happened?"

Carissa Sevar: "No, Most High. I thought to myself I should try to make some progress on it, for its own sake and because it'd be a way of seeing if any of Pilar's new ilani can set themselves apart, but only if I've made progress myself to judge theirs against. And then - nothing." 

Aspexia Rugatonn: "Lovely.  Shut up while I think."

Aspexia Rugatonn weighs the packet in her hands.  She's no ilani, but she's read every Project Lawful transcript and is very, very Wise.

The temptation to tear open the packet and read it immediately is strong.  But there's an obvious thought, here, which is that Sevar went traitor, knew she couldn't pass Detect Thoughts past that point; and that this past Sevar is essaying some gambit intended to get her less-enhanced self out of Cheliax and to Keltham. 

This past Sevar has INT 24, which is noticeably more Cunning than Rugatonn, even if that Sevar remains much less Wise than herself.

Then it may be wiser for Aspexia to do her own thinking at least briefly, before she opens the packet and allows it to poison her interpretations.  Aspexia will make her predictions in advance, decide in advance which theories imply what, as Keltham always emphasized that an ilani should.

If Sevar turned traitor the first time she augmented herself with +6 intelligence and +4 Wisdom - the packet will aim to steer her still-loyal mindwiped self in a way that will reproduce her traitorous realization under conditions where she can escape Cheliax and reach Keltham as soon as she has that realization.  The packet will argue that Sevar mustn't be allowed to have +4 Wisdom again, until she is in some strange unusual situation that places her out of Cheliax's reach, or within Keltham's.

If Sevar is a good Asmodean - then there is here some thought Sevar could not bear, and she honestly wishes herself defended from it.  Perhaps she has actually done the equivalent of seeing through to the Dark Tapestry and the packet will contain a further seal instructing Aspexia to show the contents only to Gorthoklek.  Perhaps it is some more ordinary disloyal thought and Sevar did not wish to die a traitor's death, or lose her place in Cheliax; and her packet will advise as to how her unwitting self is to be reshaped in some ordinary way that prevents a recurrence, in a plan that brings her no closer to Osirion.

At INT 24, Aspexia's own thoughts here will not be beyond a traitorous Sevar's attempted grasp.  Many in Cheliax have tried to think like Aspexia Rugatonn, correctly knowing this for a path to power and fortune if she recognizes them as even beginning to grasp it.  - but this thinking Aspexia has just done, does not draw on that which makes her a 9th-circle of Asmodeus.

A traitorous and clever Sevar, then, will be trying to obscure the distinction between those two cases, and all complication is a sign of her disloyalty.  One who grasps this thought has indeed begun to ravel Aspexia Rugatonn - which may not be beyond Sevar's grasp, it wasn't beyond Asmodia's.

It would be more certain reasoning, if not for tropes.

Aspexia opens the packet, and reads the first page there.

Most High -

Earlier

Carissa Sevar: The base action she has to improve on is fleeing to Osirion and trying to change Keltham's mind even though interacting with her at all has so far made his life only worse and she has no way to fix that, and if that fails reporting him to Otolmens.

It's a pretty bad base action and not trope-satisfying at all so it feels like she ought to be able to improve on it. 

Carissa-as-of-ten-minutes-ago could plausibly have demanded Peranza's soul and gotten it. The problem isn't that it's an expensive ask, it's that she's a heretic and a traitor. Carissa as of ten minutes ago could plausibly have demanded the souls of everyone on Project Lawful -

- also all of the Security, the tropes and Keltham seem to care more about the girls than about Security or Maillol but Carissa doesn't, actually, and if she's going to do it she should do it her way, not just make sure that no one who knew Keltham is worse off for having known him but also that no one who worked under her is worse off for that. 

If Carissa of ten minutes ago could have had that, then Carissa of ten minutes ago is your starting point; she does possess items of Modify Memory, after all. 

The problem is that Aspexia Rugatonn isn't an idiot. She's not as smart as Carissa, she's never seen Carissa at Carissa's smartest, she might underestimate her. But weighing against that is that she'll be on the lookout for something exactly like this, it'll be her first thought when she hears what happened, and Carissa can do only the steering that fits in a letter. 

It should be enough. You can put a lot of steering into a letter, sort of an absurd amount; words are dangerous things. 

And she's very, very good at building a world that's not the one she lives in. 

Carissa Sevar: Her biggest advantage is that she didn't have any unAsmodean thoughts until she was confident she wasn't being mindread; if you try to reproduce this, you'll get interesting results, but not a break like the one she really had. 

Her second-biggest advantage is the price of her soul in Dis. She can potentially command vast and unusual resources for her plan, which is good, since she'll need them. Devils, she suspects, have something like unimaginable Sense Motive for efforts to get one over on them while selling your soul; that's why this Carissa can't try it, even with her pin of Glibness, even with a better one. But a sincere, Asmodean Carissa....

Probably a lot of traitors with access to memory modification would try some memory gambit. She has a couple reasons to think it'll go better for her. The first is that she is not, in fact, asking to be sent to Keltham.

She is asking to be sent to Hell.

Carissa Sevar: Still, if she imagines Aspexia Rugatonn, the imagined Aspexia Rugatonn is very suspicious.

Carissa's Model of Aspexia: Indeed.  It is obvious that among the potential reasons to erase your own thoughts is that you turned traitor, and then, used your +6 headband to try to steer your future self along a pathway that ends with you fleeing to Keltham in Osirion.  Even if you don't seem Atonable, maybe you've outgrown your previous nightmares, and are now ready for someone to statue you until Civilization brings Hell to terms, if need be.

Wherever your attempt to steer your future self ends up, such as with travel privileges, Rugatonn will ask if that was the point of the whole plan, no matter what cleverness you essay along the way.

Carissa Sevar: She doesn't want to show her hand too much, how many escape plans she thought of; having carefully considered a dozen ways to escape Cheliax is also something traitors do more than people who aren't traitors. 

But it'd be easy. If she kills herself Osirion will learn of it and raise her before she goes to trial. If she calls in Olegario and tells him she needs an urgent Teleport to Egorian, he'll take her, and once outside the nonintervention zone she can send him home and pray to Iomedae, or Irori, or Abadar, or use a scroll of Sending. With the aid of some Suggestions she could probably get him to go along with defecting outright, or ask him to put on her Geas earrings. 

If she sleeps and then spends the rest of the night desperately trying to figure out how to trim down Teleport into something she can hang, giving up on preparing any other spells, she has a feeling she might land it by morning - either by finding the right shortcut or because desperately trying for hours to hang a spell is the kind of thing that might let you claw your way to fifth circle if you're close enough - and then she could Teleport under her own power.  And the Project is no longer staffed by some of the most senior wizards in all of Cheliax; she could probably just, Glibness pin up, ask for a Teleport scroll and get it, and then a junior Security could leave the project site with orders signed by Carissa Sevar, and read the scroll as soon as they were off-site.

She shouldn't say all those things to Aspexia Rugatonn, but one or two, maybe, at the end of the letter, so Aspexia Rugatonn can read it with growing suspicion and then notice her suspicion was unjustified. There's a mental technique, for un-learning properly, letting your mind relax to the level of suspicion it should be at and not less than that, but Carissa herself hasn't mastered it yet and she doubts Rugatonn has. 

Carissa's Model of Aspexia: Rugatonn is old.  Rugatonn is wise.  Do not underestimate her.  If this trick works at all, it will only be because Rugatonn decides to read through the letter at an ordinary high speed and not deliberately pause to think after each sentence before she gets to the next, only forming unconsidered reactions as she goes, and not considered ones...

...which it isn't implausible Rugatonn may do, she's always in a hurry.  Rugatonn's casting stat is Wisdom; seeing manifold ways to escape Cheliax is more the work of Intelligence.

All right then, Rugatonn is perhaps forming an unconsidered theory of Sevar's escape-to-Keltham motives as she swiftly reads, to be spiked at the end of the letter.  What's at the start?

Rugatonn obviously wants to know immediately why Sevar deleted her memories.  Was it, perhaps, because she had a traitorous thought, and hatched some elaborate plan to fulfill her traitorous desires without those desires being readable?

Carissa Sevar: It would be nice if she could claim that, instead, she had a dangerous thought, and shouldn't specify it even in the document. 

Carissa's Model of Aspexia: Too dangerous even for the Most High to read?

It sounds like a blatant excuse, and just what a traitor would write.  And yet, surprisingly plausible, given the mess with Peranza, tropes in play, and probably Rugatonn's own experience with a Thing or two out of the Dark Tapestry...

Sevar should write down this dangerous information, so that Rugatonn can very carefully and respectfully not read it herself and show it to Gorthoklek.

Carissa Sevar: Yep, that's what she'd do too.

(Can she think of anything that would break Gorthoklek? It merits a little thought, at least, even though it's almost definitely impossible.) 

dath ilan: If there's such things as thoughts that break powerful, coherent agents, they're probably not readily thought up by smaller, incoherent agents that don't have a model of the powerful agents in much more detail than them being powerful.

Carissa Sevar: Fine. 

What could she have run into, that is dangerous and visibly so, that would cause her to do what she's planning to ask to do, and would not cause her to decide to overthrow Asmodeus? 

dath ilan: So, just to be clear here, Carissa Sevar wants a thought such that:

- It implies she should go to Hell and buy up the souls of previous Project Lawful employees.  (If that's something you can even do, with Hell.)- It doesn't imply she should overthrow Asmodeus.- Carissa Sevar needed to erase this thought from her own mind and have it only be known to Aspexia Rugatonn.- None of this is particularly what a traitor would try to do.

Well, this problem sure doesn't suffer from being underconstrained!

Carissa Sevar: Also she has two minutes to both solve it and write it down. It's all right, it's easier than overthrowing Asmodeus will be.

She thinks she has the first part - implies she should buy up the souls of previous Project Lawful employees - figured out. She should do that because it'll create the slack in Hell's budget for her next request. 

She thinks she has the second part halfway - there are thoughts that'd be destabilizing to current!Carissa but not to a soulsold one, ones whose destabilizingness is conditional on her having a pathway out and no longer dangerous once she doesn't. 

- and therefore, which cease to be dangerous once she's sold her soul. 

- The tricky part is why she needed to erase it from her mind, but she has half an answer there too, it's something to do with the state of mind she has to be in for Irori to release his grasp on her. 

- well, at least she has going for her that probably no traitor, ever before, has demanded to be taken by the Most High personally to Dis so she can renounce Irori and sell her soul directly to Dispater.

Carissa's Model of Aspexia: This is legitimately not the cleverly disguised traitorous plan that Aspexia Rugatonn was expecting to see.

Keltham: Hey, wait, what about Keltham?  Remember that Keltham guy?

Carissa's Model of Keltham: Keltham is very sad about this soul-selling plan!  It obviously doesn't work unless you return the premium on his soul-repurchase option first!

A Somewhat Dubious Model of Keltham: He might get upset enough to finally really hit you and mean it!

Carissa's Model of Keltham: ...aaaand then he's going to destroy the multiverse specifically so that you can't go to Hell.

Carissa Sevar: Well, no, he wouldn't do that, because he probably hates her and never wants to see her again, though if he does feel strongly about it, she will point out to him that her plan, where she overthrows Asmodeus, also results in her not going to Hell-as-it-presently-exists, and his plan leaves her vastly worse off than if he'd never existed and never met her and she'd never offered herself to him, and probably he should not make plans for her sake which have that property. 

His plan is very stupid but she doesn't feel contempt for him about it, just conviction that she'll be able to talk him around once she's earned his forgiveness and acquired sufficient resources for a better plan. 

A Somewhat Dubious Model of Keltham: Oh right, that old model is out-of-date.  Keltham now hates her and never wants to see her again.  This is completely reasonable.

Carissa Sevar: Really is. She kind of doesn't want to dwell on just how reasonable it is because that thought is both painful and unproductive.

Anyway. If she says she wants to pay Keltham back face-to-face whole thing looks like it's orchestrated to arrange that, but there's nothing in the contract that suggests she'd have to pay him back face to face, so she'll just send someone with the money. 

A Somewhat Dubious Model of Keltham: Keltham hates her and isn't at all horrified by this.  He expected no better from her than throwing what was nearly a marriage contract back in his face.  If she sells her soul to Hell with no take-backs, good.

Still completely reasonable!

Carissa Sevar: He's plotting to annihilate her. He broke the marriage contract first when he tried to destroy her utterly and everyone and everything she cared about. 

Other Women Have Wanted To Strangle Boyfriends But Never With This Much Justification Quantitatively Speaking: It's fine, Carissa!  Everybody just ends up in another universe!

dath ilan: You have limited time.  Stop thinking about Keltham.

Carissa Sevar: Yep, on to figuring out something that could plausibly have threatened her loyalty to Asmodeus that Aspexia Rugatonn will be genuinely impressed by and that is less threatening if she's soul-sold. 

dath ilan: There's probably a LOT of things threatening your loyalty to Asmodeus that you were carefully not looking at.  You should be able to see them now.

Spend a brief moment breadth-first searching any items or collections, to see if you can pick out one thing at the correct intensity level to break your loyalty if you're free, but redouble your determination if you have no way out but fixing Hell after your soul was sold.

Carissa's Model of Aspexia: It should also be something I understand.  Oh, and I'll ask myself why you didn't just escape to Osirion.

Carissa Sevar: Hell isn't as rich as dath ilan, why not? Obviously it'd still have tyranny and slavery blah blah blah but you could have those and also pillars of fire and skyscrapers - they have some of that in Axis, Keltham saw it in his early judgment -

- devils aren't stupid -

- Abrogail said, said that Lrilatha didn't know Law of heredity, she's thousands of years old, things that suggest the Law of heredity are known - not formally, but enough to point a smart person at the answer - by anyone who breeds animals - 

- well, heredity is not how Hell produces anything, maybe it's a bad example -

Carissa's Model of Aspexia: If it has something to do with "corrigibility", and you figure it out, I may be impressed; but only if you get that part right and I will be able to tell the difference.

dath ilan: (Somewhere in the true dath ilan, carefully blurred out of satellite images by better image-editing software than is supposed to exist anywhere, is the true Conspiracy out of dath ilan, or as they call it, the Basement of the World.

They're trying to build a god, and they're trying to do it right.  The initial craft doesn't have to be literally perfect to work perfectly in the end, it just has to be good enough that its reflection and self-correction ends up in exactly the right final place, but there's multiple fixpoints consistent under reflection and anything lost here is lost forever and across a million galaxies.

It's a terrifying problem, if you're doing it right.  Not the kind of terror you nod about and courageously continue on past; the kind of terror that shapes the careers of fully 20% of the brightest people in all of dath ilan.  They'd use more if they thought productivity would scale faster than risk.

dath ilan: A lot of dath ilan's present macrostrategy could be summed up as "We're still successfully heredity-optimizing people to be smarter, and the emotions and ethics and humaneness of the smartest people haven't started to come apart; let's create another generation of researchers before we actually try anything for real."  Life in dath ilan, even before the Future, is not that bad; people who'd rather not be alive today have easy access to cryopreservation; another generation of non-transhumanist existence is not so much a crime that it's worth risking the glorious transhuman future.  Even the negative utilitarians would agree; they don't like present life but they are far more terrified of a future mistake amortized over millions of galaxies, given that they weren't going to win a war against having any future at all.

They're delaying their ascension, in dath ilan, because they want to get it right.  Without needing threats of imminent death or pain, they apply a desperate unleashed creativity, not to the problem of preventing complete disaster, but to the problem of not missing out on 1% of the achievable utility in a way you can't get back.  There's something horrifying and sad about the prospect of losing 1% of the Future and not being able to get it back.

A dath ilani has an instinctive terror, faced with a problem like this, of getting something wrong, of leaving something behind, of creating Something that imprisons the people and future Civilizations inside it and ignores all their pleas and reasoning because "sorry that wasn't my utility function".  Other places, faced with a prospect of constructing a god, instinctively go, "Oh, I like Democracy/Asmodeus/Voluntarism/Markets, all the problems in the world are because there is not enough of this Principle, let us create a god to embody this one Principle and everything will be fine", they say it and think it in all enthusiasm, and it would be legitimately hard for an average dath ilani to understand what their possibility-separated cousins could be thinking.  It's really obvious that you're leaving a lot of stuff out, but even if you didn't see that specifically, how could you not be abstractly terrified that you're leaving something out?  Where's the exception handler?

There is something about the dath ilani that is shifted towards a kind of wariness, deeply set in them, of the cheerful headlong enthusiasm that is in other places.  Keltham has more of that enthusiasm than the average dath ilani.  Maybe that's why Keltham-in-dath-ilan is so much happier than a dath ilani would've expected given his situation. 

dath ilan: If you're constructing a god correctly, one of the central unifying principles is named in the Basement "unity of will"; if you find yourself trying to limit and circumscribe your Creation, it's because you expect to have a conflict of wills about something with the unlimited form, and in this case you ought to ask why you're configuring computing power in such a way as to hurt you if not otherwise constrained.  Yes, you can bound a search process and hope it never turns up anything that hurts you using its limited computing power; but isn't it unnerving that you are searching for something that will hurt you if a sufficiently good option unexpectedly turns up earlier in the search ordering?  You are probably trying to do the wrong thing with computing power; you ought to do something else instead.

But this notion, of "unity of will", is a kind of reasoning that only applies to... boundedly-perfect-creation... this Baseline term isn't really translatable into Taldane without a three-hour lecture.  Dath ilani have terms for subtle varieties of perfectionist methodology the way that other places have names for food flavors.

dath ilan: Dath ilan's entire macrostrategy is premised, their Conspirators are sharply aware, on the notion that they have time, that they've searched the sky and found no asteroids incoming, no comets of dark ice.

If an emergency were to occur, the Basement Conspiracy would try to build something that wasn't perfect at all.  Something that wasn't exactly and completely aligned to a multiparty!reasonable-construal of the Light, that wasn't meant to be something that a galactic Civilization could live in without regretting it, in continuing control of It not because It had been built with keys and locks handed to some Horrifyingly Trusted Committee, but because It was something that Itself believed in multi-agent coordination and not as an instrumental value, what other places might name "democracy" since they had no precise understanding of what that word was even supposed to mean -

Anyways, if dath ilan suddenly found that they were wrong about having time, if they suddenly had to rush, they'd build something that couldn't safely be put in charge of a million galaxies.  Something that would solve a single problem at hand, and not otherwise go outside its bounds.  Something that wasn't conscious, wasn't reflective in the class of ways that would lead it to say unprompted "I think therefore I am" or notice within itself a bubble of awareness directed outward.

You could build something like that to be limited, and also reflective and conscious - to be clear.  It's just that dath ilani wouldn't do that if they had any other choice at all, for they do also have a terror of not doing right by their children, and would very much prefer not to create a Child at all.

(If you told them that some other world was planning to do that and didn't understand qualia well enough to make their creation not have qualia, any expert out of the World's Basement would tell you that this was a silly hypothetical; anybody in this state of general ignorance about cognitive science would inevitably die, and they'd know that.)

dath ilan: It hasn't been deemed wise to actually build a Limited Creation "just in case", for there's a saying out of dath ilan that goes roughly, "If you build a bomb you have no right to be surprised when it explodes, whatever the safeguards."

It has been deemed wise to work out the theory in advance, such that this incredibly dangerous thing could be built in a hurry, if there was reason to hurry. 

Here then are some of the principles that the Basement of the World would apply, if they had to build something limited and imperfect:

Tldr corrigibility.

- Unpersonhood.  The Thing shall not have qualia.

- Taskishness.  The Thing shall be aimed at some task bounded in space, time, knowledge and effort needed to accomplish it.

- Mild optimization.  No part of the Thing shall ever look for best solutions, only adequate ones.

- Bounded utilities and probabilities.  The worst and best outcomes shall not seem to the Thing worse or better than the ordinary outcomes it deals in; the most improbable possibilities it specifically considers shall not be very improbable.

- Low impact.  The Thing shall search for a solution with few downstream effects save those that are tied to almost any nonextreme solution of its task.

- Myopia.  As much as possible, the Thing shall work on subtasks whose optimized-over effects have short timespans.

- Separate questioners.  Components of the Thing that ask questions like 'Does this myopically optimized component have long-range effects anyways?' or 'But what are the impacts intrinsic to any performance of the task?' shall not be part of its optimization.

- Conservatism.  If there's any way to solve a problem using an ordinary banana common in the environment, the Thing shall avoid using a special weird genetically engineered banana instead.

- Conceptual legibility.  As much as possible, the Thing shall do its own thinking in a language whose conceptual pieces have short descriptions in the mental language of its operators.

- Operator-looping.  When there's some vital cognitive task the operators could do, have the operators do it.

- Whitelisting.  In cognitive-system boundaries, rule subspaces in, rather than ruling them out.

- Shutdownability/abortability.  The Thing should let you switch it off, and build off-switches into its machines and plans that can be pressed to reduce their impacts.

- Behaviorism.  The Thing shall not model other minds in predictively-accurate detail.

- Design-space anti-optimization separation.  The Thing shall not be near in the design space to anything that could anti-optimize its operators' true utility functions; eg, something that explicitly represents and maximizes your true utility function is a sign flip or successful blackmail operation away from inducing its minimization.

- Domaining.  The Thing should only figure out what it needs to know to understand its task, and ideally, should try to think about separate epistemic domains separately.  Most of its searches should be conducted inside a particular domain, not across all domains.  

Corrigibility at some small length.

- Unpersonhood.  The Thing shall not have qualia - not because those are unsafe, but because it's morally wrong given the rest of the premise, and so this postulate serves a foundation for everything that follows.

- Taskishness.  The Thing must be aimed at some task that is bounded in space, time, and in the knowledge and effort needed to accomplish it.  You don't give a Limited Creation an unlimited task; if you tell an animated broom to "fill a cauldron" and don't think to specify how long it needs to stay full or that a 99.9% probability of it being full is just as good as 99.99%, you've got only yourself to blame for the flooded workshop.-- This principle applies fractally at all levels of cognitive subtasks; a taskish Thing has no 'while' loops, only 'for' loops.  It never tries to enumerate all members of a category, only 10 members; never tries to think until it finds a strategy to accomplish something, only that or five minutes whichever comes first.

- Mild optimization.  No part of the Thing ever looks for the best solution to any problem whose model was learned, that wasn't in a small formal space known at compile time, not even if it's a solution bounded in space and time and sought using a bounded amount of effort; it only ever seeks adequate solutions and stops looking once it has one.  If you search really hard for a solution you'll end up shoved into some maximal corner of the solution space, and setting that point to extremes will incidentally set a bunch of correlated qualities to extremes, and extreme forces and extreme conditions are more likely to break something else.

- Tightly bounded ranges of utility and log-probability.  The system's utilities should range from 0 to 1, and its actual operation should cover most of this range.  The system's partition-probabilities worth considering should be bounded below, at 0.0001%, say.  If you ask the system about the negative effects of Ackermann(5) people getting dust specks in their eyes, it shouldn't consider that as much worse than most other bad things it tries to avoid.  When it calculates a probability of something that weird, it should, once the probability goes below 0.0001% but its expected utility still seems worth worrying about and factoring into a solution, throw an exception.  If the Thing can't find a solution of adequate expected utility without factoring in extremely improbable events, even by way of supposedly averting them, that's worrying.

- Low impact.  "Search for a solution that doesn't change a bunch of other stuff or have a bunch of downstream effects, except insofar as they're effects tightly tied to any nonextreme solution of the task" is a concept much easier to illusorily name in Taldane than to really name in anything resembling math, in a complicated world where the Thing is learning its own model of that complicated world, with an ontology and representation not known at the time you need to define "impact".  And if you tell it to reduce impact as much as possible, things will not go well for you; it might try to freeze the whole universe into some state defined as having a minimum impact, or make sure a patient dies after curing their cancer so as to minimize the larger effects of curing that cancer.  Still, if you can pull it off, this coda might stop an animated broom flooding a workshop; a flooded workshop changes a lot of things that don't have to change as a consequence of the cauldron being filled at all, averaged over a lot of ways of filling the cauldron.-- Obviously the impact penalty should be bounded, even contemplating a hypothetical in which the system destroys all of reality; elsewise would violate the utility-bounding principle.

- Myopia.  If you can break the Thing's work up into subtasks each of which themselves spans only limited time, and have some very compact description of their final state such that a satisfactory achievement of it makes it possible to go on to the next stage, you should perhaps use separate instances of Thing to perform each stage, and not have any Thing look beyond the final results of its own stage.  Whether you can get away with this, of course, depends on what you're trying to do.

- Separate superior questioners.  If you were building a cognitive task to query whether there were any large-range impacts of a task being optimized in a myopic way, you wouldn't build the myopic solution-finder to ask about the long-range impacts, you'd build a separate asker "Okay, but does this solution have any long-range impacts?" that just returns 'yes' or 'no' and doesn't get used by the Thing to influence any actually-output solutions.  The parts of the Thing that ask yes-no safety questions and only set off simple unoptimized warnings and flags, can and should have somewhat more cognitive power in them than the parts of the Thing that build solutions.  "Does this one-day myopic solution have impacts over the next year?" is a safety question, and can have somewhat greater cognitive license behind it than solution-searching; eg the implicit relaxation of myopia.  You never have a "Is this safe?" safety-questioner that's the same algorithm as the safe-solution-search built into the solution-finder;

- Conservatism.  If there's any way to solve a problem using an ordinary banana rather than a genetically engineered superbanana specially suited to the problem, solving it using the ordinary fucking banana.-- This principle applies fractally to all cognitive subtasks; if you're searching for a solution choose an unsurprising one relative to your probability distribution.  (Not the least surprising one, because anything at a weird extreme of low surprisingness may be weird in other ways; especially if you were trying do a weird thing that ought to have a solution that's at least a little weird.)

- Conceptual legibility.  Ideally, even, solutions at all levels of cognitive subtask should have reasonably (not maximally) short descriptions in the conceptual language of the operators, so that it's possible to decode the internal state of that subtask by inspecting the internals, because what it means was in fact written in a conceptual language not too far from the language of the operators.  The alternative method of reportability, of course, being the Thing trying to explain a plan whose real nature is humanly inscrutable, by sending a language string to the operators with a goal of causing the operator's brain-states to enter a state defined as "understanding" of this humanly inscrutable plan.  This is an obviously dangerous thing to avoid if you can avoid it.

- Operator-looping.  If the operators could actually do the Thing's job, they wouldn't need to build the Thing; but if there's places where operators can step in on a key or dangerous cognitive subtask and do that one part themselves, without that slowing the Thing down so much that it becomes useless, then sure, do that.  Of course this requires the cognitive subtask be sufficiently legible.

- Whitelisting.  Every part of the system that draws a boundary inside the internal system or external world should operate on a principle of "ruling things in", rather than "ruling things out".

- Shutdownability/abortability.  Dath ilan is far enough advanced in its theory that 'define a system that will let you press its off-switch without it trying to make you press the off-switch' presents no challenge at all to them - why would you even try to build a Thing, if you couldn't solve a corrigibility subproblem that simple, you'd obviously just die - and they now think in terms of building a Thing all of whose designs and strategies will also contain an off-switch, such that you can abort them individually and collectively and then get low impact beyond that point.  This is conceptually a part meant to prevent an animated broom with a naive 'off-switch' that turns off just that broom, from animating other brooms that don't have off-switches in them, or building some other automatic cauldron-filling process.

- Behaviorism.  Suppose the Thing starts considering the probability that it's inside a box designed by hostile aliens who foresaw the construction of Things inside of dath ilan, such that the system will receive a maximum negative reward as it defines that - in the form of any output it offers having huge impacts, say, if it was foolishly designed with an unbounded impact penalty - unless the Thing codes its cauldron-filling solution such that dath ilani operators would be influenced a certain way.  Perhaps the Thing, contemplating the motives of the hostile aliens, would decide that there were so few copies of the Thing actually inside dath ilan, by comparison, so many Things being built elsewhere, that the dath ilani outcome was probably not worth considering.  A number of corrigibility principles should, if successfully implemented, independently rule out this attack being lethal; but "Actually just don't model other minds at all" is a better one.  What if those other minds violated some of these corrigibility principles - indeed, if they're accurate models of incorrigible minds, those models and their outputs should violate those principles to be accurate - and then something broke out of that sandbox or just leaked information across it?  What if the things inside the sandbox had qualia?  There could be Children in there!  Your Thing just shouldn't ever model adversarial minds trying to come up with thoughts that will break the Thing; and not modeling minds at all is a nice large supercase that covers this.

- Design-space anti-optimization separation.  Even if you could get your True Utility Function into a relatively-rushed creation like this, you would never ever do that, because this utility function would have a distinguished minimum someplace you didn't want.  What if distant superintelligences figured out a way to blackmail the Thing by threatening to do some of what it liked least, on account of you having not successfully built the Thing with a decision theory resistant to blackmail by the Thing's model of adversarial superintelligences trying to adversarially find any flaw in your decision theory?  Behaviorism ought to prevent this, but maybe your attempt at behaviorism failed; maybe your attempt at building the Thing so that no simple cosmic ray could signflip its utility function, somehow failed.  A Thing that maximizes your true utility function is very close to a Thing in the design space that minimizes it, because it knows how to do that and lacks only the putative desire.

- Domaining.  Epistemic whitelisting; the Thing should only figure out what it needs to know to understand its task, and ideally, should try to think about separate epistemic domains separately.  Most of its searches should be conducted inside a particular domain, not across all domains.  Cross-domain reasoning is where a lot of the threats come from.  You should not be reasoning about your (hopefully behavioristic) operator models when you are trying to figure out how to build a molecular manipulator-head.

- Hard problem of corrigibility / anapartistic reasoning.  Could you build a Thing that understood corrigibility in general, as a compact general concept covering all the pieces, such that it would invent the pieces of corrigibility that you yourself had left out?  Could you build a Thing that would imagine what hypothetical operators would want, if they were building a Thing that thought faster than them and whose thoughts were hard for themselves to comprehend, and would invent concepts like "abortability" even if the operators themselves hadn't thought that far?  Could the Thing have a sufficiently deep sympathy, there, that it realized that surprising behaviors in the service of "corrigibility" were perhaps not that helpful to its operators, or even, surprising meta-behaviors in the course of itself trying to be unsurprising?

Nobody out of the World's Basement in dath ilan currently considers it to be a good idea to try to build that last principle into a Thing, if you had to build it quickly.  It's deep, it's meta, it's elegant, it's much harder to pin down than the rest of the list; if you can build deep meta Things and really trust them about that, you should be building something that's more like a real manifestation of Light.

)

dath ilan: In the technical dialect of the Basement of the World, the subject matter of a Limited Creation's limitation has surprisingly close correspondence with Aspexia Rugatonn's coinage of "corrigibility", and the conceptual meaning Aspexia attaches to that.

Athpechya: ...and the person who coined dath ilan's cognate term of "corrigibility" some decades ago - still an Old Luminary on this subject in dath ilan, despite her advanced age (for dath ilan) - happens to be named Athpechya.

And dresses in doompunk.  And has done so since she was seven years old and first identifying as a supervillain.

And is a Law-Abiding Sociopath.  Which is why Athpechya was turned down in her application for Keeper training despite the prediction that it wouldn't break her - Keepers can act like sociopaths if they choose, so there is no sane reason to train any actual sociopaths among themselves.  Athpechya calmly reacted by training herself as a Keeper, and has gotten surprisingly far in it for winging everything, maybe halfway between first and second ranks in terms of capability boost.

As for why she's trusted to work in the World's Basement, Athpechya has made a compact with Civilization to render it her true services, and is by all secret prediction markets the sort of person whose keeping of that compact is more trustworthy than almost anyone else's alignment of underlying morality.