Marked for Death: A Rational Naruto Quest

[Noumero]

Which idea exactly? I think I've lost the thread here

Click to shrink...

This one, I think:

Consider warning the Akatsuki about Orochimaru's assault and runic abilities if Oro kills us.

• Having a really serious deadman switch makes meeting with Oro much safer.

Click to shrink...

I am against that. If we do that, it would become good decision theory for Orochimaru to ignore our dead man's switches and kill us anyway.

As-is, our dead man's switches are not, actually, deterrents. They're contingencies which turn a universe in which Oro kills us from a lose:win situation for us:Oro, into a win:lose. If Oro kills us despite our dead man's switches, we'd still prefer them to go off, because they would destroy Oro's monopoly on runecrafting and give it to people who are our allies/closer to us in values, who'd be able to act as checks on Oro. Creating these dead man's switches improves our outcomes across all possible worlds, regardless of Oro's actions.

On the other hand, empowering the Akatsuki relative to Orochimaru is not something we actually want, even if we're dead. A rune-monopolist!Orochimaru might still help Leaf defeat the Akatsuki and create an Uplift-adjacent future, whereas the Akatsuki are likely to set off a world-ending ritual and kill everyone. Therefore, if Orochimaru kills us despite our dead man's switch, we would not want our messages warning the Akatsuki about runecrafting to actually go out, because it would create a worse-by-our-values world. This kind of dead man's switch doesn't actually improve the outcomes for us, it only worsens outcomes for Orochimaru. Which means the only reason for us to create such a dead man's switch is to try and intimidate Orochimaru into yielding to us.

Which means Orochimaru ought to ignore that dead man's switch and kill us anyway, so as to create a deterrent against us building that dead man's switch to begin with. (And this deterrent might take the form not of rational calculation, but of impulsive behavior where he gets so pissed off at our perceived idiocy he lashes out and kills us on the spot, say. Two can play the game of "I am sooo crazy and irrational, you better not cross me".)

It's even worse if we're combining that with sending supplies for learning runecrafting to Leaf/our other allies, since then not only Orochimaru would be depowered relative to the Akatsuki, but our own allies as well.

Now, granted, such tactics might still work. Orochimaru is not necessarily a logical-decision-theory agent; he might yield to threats. Certainly plenty of real-life people do. But we don't actually know how competent/incompetent he is at decision theory, and modifying our dead man's switches in this way creates unnecessary additional failure states in worlds where Orochimaru is competent enough.

 

[_The_Bomb]

@eaglejarl @Velorien @Paperclipped

A few questions for y'all:

• If Hazō were to ask Cannai whether Kakashi had ever summoned him, what would Cannai say? (The intent of this question is to gain more information on Cannai's CP requirement.)
• In Hazō's opinion, how long would it take him to prepare sufficient reference material for someone to rediscover Primordial Sealing without Hazō's direct guidance? (The intent of this question is to see the feasibility of creating a deadman switch with Cannai that would disincentivize Orochimaru to kill Hazō to keep Runecraft to himself.)

 

[RandomOTP]

I am against that.

Click to shrink...

Seconding this.

 

[faflec]

I don't actually know - was wondering whether to raise the possibility. @faflec , any insight?

Click to shrink...

"Kakashi was a lone wolf with little ability to make friends among his own pack, yet he did well among our Clan. He was a powerful ally who fought beside us against the Leopards and their arrogant invasions of our territory. He was wealthy on the Human Path and able to provide us with things we lack here. He had the favor of your Hokage, meaning that the Dog Clan had Leaf itself as an ally that could be dispatched to further our interests. As a wealthy and clanless jōnin he was able to spend significant amounts of time here, meaning that his hands were useful to us.

Click to shrink...

"Still surprised by basic courtesy, I see." Cannai shook his massive head. "The Human Path must be a terrible place."

"Have...have you ever been there, sir?"

"Several times, over the span of centuries. Generally I am only there for a few minutes when a Summoner needs me to kill someone. I try to keep in mind that these experiences are not representative of an entire Path and its people."

Click to shrink...

Not definite but what I could find.

 

[Kyreneryk]

On the other hand, empowering the Akatsuki relative to Orochimaru is not something we actually want, even if we're dead. A rune-monopolist!Orochimaru might still help Leaf defeat the Akatsuki and create an Uplift-adjacent future, whereas the Akatsuki are likely to set off a world-ending ritual and kill everyone. Therefore, if Orochimaru kills us despite our dead man's switch, we would not want our messages warning the Akatsuki about runecrafting to actually go out, because it would create a worse-by-our-values world. This kind of dead man's switch doesn't actually improve the outcomes for us, it only worsens outcomes for Orochimaru. Which means the only reason for us to create such a dead man's switch is to try and intimidate Orochimaru into yielding to us.

Which means Orochimaru ought to ignore that dead man's switch and kill us anyway, so as to create a deterrent against us building that dead man's switch to begin with. (And this deterrent might take the form not of rational calculation, but of impulsive behavior where he gets so pissed off at our perceived idiocy he lashes out and kills us on the spot, say. Two can play the game of "I am sooo crazy and irrational, you better not cross me".)

Click to shrink...

My understanding of decision theory is not that extensive, but ... does this actually hold up? This is more-or-less a brinkmanship situation, if I understand it correctly; the payoff matrix is something like:

Us, Oro	-Akatsuki Switch	+Akatsuki Switch
-Oro kills us	0, -1	-1, -2
+Oro kills us	-100, 0	-(100+x), -y

Assuming that we indeed prefer Oro to beat Akatsuki even if he killed us before the assault (which I don't necessarily agree with) and that we're dealing with an Oro who would prefer to kill us before the assault (without which there's no reason to set up this switch.) Values are approximate, obviously, with the key assumption being that Oro doesn't care about killing us that much. x is how much we care about Akatsuki being more likely to resurrect Pain even if we're dead, y is how much Oro cares about Akatsuki being forewarned.

The complication is that x is probably not that large, especially not compared to y - we'd be dead anyway, while Oro would have to live with the consequences. Suppose y were 1000 and x were 20 - then would it make sense for us to set it up? It would still only have the purpose of intimidating Oro, but we'd be threatening him with something really bad and ourselves with something relatively mild; it seems reasonable to expect that Oro might yield first. Also, we have first-mover advantage, in that we've already locked ourselves into having the switch or not before we meet Oro and he gets the chance to kill us or not (though I suppose Oro could lock himself into killing us or not by precommitment).

As I've said, I'm not in favour of setting up this switch - I think the marginal benefit for us is tiny compared to just having the Cannai switch, and I think it might cause Oro to plan to kill us later even if he wasn't before - but it's not obvious to me that decision theory necessarily says Oro shouldn't yield.

Actually, insight - I think my key concern here is that Oro deciding to kill us regardless to disincentivize us from setting up a deadman's switch seems to me to only work if Oro expects us to be able to predict that he'd do that (à la Newcomb's paradox). I'm not at all convinced that we actually can predict Orochimaru's actions that well, especially in a scenario where Oro would be imposing such a massive cost on himself by killing us.

(Apologies if this has some major flaw that I just haven't seen - as I said, not all that familiar with decision theory.)

 

[Sir Stompy]

This one, I think:

I am against that. If we do that, it would become good decision theory for Orochimaru to ignore our dead man's switches and kill us anyway.

As-is, our dead man's switches are not, actually, deterrents. They're contingencies which turn a universe in which Oro kills us from a lose:win situation for us:Oro, into a win:lose. If Oro kills us despite our dead man's switches, we'd still prefer them to go off, because they would destroy Oro's monopoly on runecrafting and give it to people who are our allies/closer to us in values, who'd be able to act as checks on Oro. Creating these dead man's switches improves our outcomes across all possible worlds, regardless of Oro's actions.

On the other hand, empowering the Akatsuki relative to Orochimaru is not something we actually want, even if we're dead. A rune-monopolist!Orochimaru might still help Leaf defeat the Akatsuki and create an Uplift-adjacent future, whereas the Akatsuki are likely to set off a world-ending ritual and kill everyone. Therefore, if Orochimaru kills us despite our dead man's switch, we would not want our messages warning the Akatsuki about runecrafting to actually go out, because it would create a worse-by-our-values world. This kind of dead man's switch doesn't actually improve the outcomes for us, it only worsens outcomes for Orochimaru. Which means the only reason for us to create such a dead man's switch is to try and intimidate Orochimaru into yielding to us.

Which means Orochimaru ought to ignore that dead man's switch and kill us anyway, so as to create a deterrent against us building that dead man's switch to begin with. (And this deterrent might take the form not of rational calculation, but of impulsive behavior where he gets so pissed off at our perceived idiocy he lashes out and kills us on the spot, say. Two can play the game of "I am sooo crazy and irrational, you better not cross me".)

It's even worse if we're combining that with sending supplies for learning runecrafting to Leaf/our other allies, since then not only Orochimaru would be depowered relative to the Akatsuki, but our own allies as well.

Now, granted, such tactics might still work. Orochimaru is not necessarily a logical-decision-theory agent; he might yield to threats. Certainly plenty of real-life people do. But we don't actually know how competent/incompetent he is at decision theory, and modifying our dead man's switches in this way creates unnecessary additional failure states in worlds where Orochimaru is competent enough.

Click to shrink...

Hmmmmm @Left-Hand Mutant I am pretty convinced by this. Any counterarguement?

 

[FaintlySorcerous]

[X] Action Plan: Desperate Measures
Word Count <399
Intended Duration: ~1 day

• Discuss the Orochimaru situation with the team, including Snowflake.
  • Hazou thinks that it's likely Orochimaru fled Leaf for the same reason we did -- to research in peace.
  • Hazou doesn't think that Orochimaru is likely to betray us at this point (prior to the Rift Assualt)
  • However, it makes sense to take some precautions. We'll go to the meeting with SCs only to minimize exposure to a potentially hostile Orochimaru. Unless there are objections.
    • Should we split the team to get into contact with Leaf sooner? Hazou favors Noburi (for runic preparations) and Yuno with him, and Mari to make contact with Leaf but composition is flexible.
    • Should Mari come with Hazou to detect if Orochimaru is being deceitful?
  • Hazou also thinks it would be prudent to prepare a couple deadman switches giving Leaf/TU resources to learn Runecrafting in case Orochimaru decides to eliminate his only runecrafting peer.
    • He favors giving one to Kei/Kagome and setting up one in Dog. Should be sufficiently hardened resist Orochimaru torturing Hazou into compliance.
  • All of our Rift Assualt strategies require extra chakra from Leaf, how likely is it that Orochimaru is attempting this without drawing on them for support?
  • Kei, how likely do you think Orochimaru betraying us is? Both before we assualt the Rift and afterwards. Does the assessment change if he doesn't want to involve Leaf?
  • Noburi, can you ask the Toad Sages and Gamabunta if they'd be willing to avenge Jiriaya?
• Research
  • Loop in just Kei and Snow
    • Intent: Asking for harm mitigation, not permission.
    • Kei, apologies, but our other ideas didn't pan out and we need to know if this is viable. We don't plan to inform anyone or complete research unless things are truely desperate.
    • Can you brainstorm ways to disguise the effect?
  • Spend one day no-prep yes-DoB researching Superchillers (referred to only as Project Twilight) and Kagome's Tears
    • Have Noburi use Pain Supression for the research rolls
    • Fully enclose the AoE of the prototype Superchiller with a minimum size Force Dome to prevent a storm from forming.
    • Bury the rune afterwards
• Misc
  • Set up the Dead Man's switches discussed above.

Click to shrink...

The below is well (~270) under 300 words - I'm not sure if the 399 is to leave room for expansion etc. but I think I've got the tone and content, plus a clarifying statement or two.

• Meet with everyone (including Snowflake)
  • Hazo believes…
    • Orochimaru fled Leaf to pursue research, like we did.
    • Orochimaru won't hurt us unless the Akatsuki are dead. It's unclear what happens afterwards but the Akatsuki are worse than Orochimaru.
    • Does everyone concur?
    • Kei/Snowflake, what's the likelihood Orochimaru betrays us after the fact to keep runes to himself? The Rift? If he does or doesn't plan on involving (or returning to) Leaf?
  • We plan on meeting him via SC to discuss plans.
    • Noburi and Yuno, would you accompany Hazo? You're good protection, counter-tracking, and chakra if we mount a runic assault.
    • Mari, should you come? Are we good enough to tell if Orochimaru is lying to us?
    • Should we try to contact Leaf simultaneously? It risks tipping off the Akatsuki. If so, Mari should lead that team.
  • We want deadman switches with everything required to learn runecrafting if we don't survive. It's too much power for Orochimaru alone.
    • One in Dog, one with Kei/Kagome?
  • Our Rift assault plans hinge on runes (RER barrage) and therefore lots of chakra.
    • We'd planned on Leaf - is Orochimaru likely to have enough, somehow? (Noburi: did Tsunade ever say anything?)
  • Noburi: would the Toad Sages/Gamabunta be willing to avenge Jiraiya?
• Kei/Snowflake:
  • Tone: remorseful, desperate.
  • We need to determine if Superchillers are viable. Can you think of any ways to disguise the effect?
• Research: one day no-prep yes-DoB researching Superchillers and Kagome's Tears.
  • Have Noburi use Pain Suppression for the research rolls
  • Fully enclose the Superchiller prototype's AoE with a minimum size Force Dome to prevent limit airflow. Bury it afterwards.
• Implement deadman switches as above.

Click to shrink...

I will vote against plans which have us threaten Orochimaru to his face, unprompted. We are a lootbox and we are not dangerous enough to merit a sudden assassination. If he decides he wants us dead he's going to be interrogating us first.

That said, I think I would vote for plans which set up contingencies providing the remainder of the Akatsuki with Runecrafting capabilities if Orochimaru takes possession of the Rift, kills us, and does not return to Leaf. If Orochimaru moves against us, we can let him know and go from there. I think this would be (relatively) straightforward on the Cannai side: if we die, read anyone except the Akatsuki in. If they appear, demand proof that Orochimaru never returned to Leaf - the Slug Princess' word will do just fine. If they can get it, then they can have Runecrafting.

 

[Noumero]

Assuming that [...] we're dealing with an Oro who would prefer to kill us before the assault (without which there's no reason to set up this switch.)

Click to shrink...

Oh yeah, that's another thing. We're almost certainly not dealing with such an Oro, so that part of the dead man's switch is unnecessary and only serves to antagonize Oro more.

Actually, insight - I think my key concern here is that Oro deciding to kill us regardless to disincentivize us from setting up a deadman's switch seems to me to only work if Oro expects us to be able to predict that he'd do that (à la Newcomb's paradox).

Click to shrink...

Yep. Well, technically, he doesn't need to expect us to predict him specifically, he needs to expect us to model how a smart-enough agent would behave and then assume that he is a smart-enough agent. Not yielding to threats is a convergently good strategy, such that many agents would independently reinvent it; it doesn't rely on the peculiarities of Oro's mind specifically.

Or, more in the realm of how human beings work: it would also work if Oro has learned social heuristics where he becomes irrationally defiant if he perceives someone trying to pump him for resources using unreasonable threats. Having a reputation for a "bad temper" is one possible way to implement something like the above, but in a more compute-efficient manner (i. e., via deontology, not consequentialism). And the local maximum here would be at "fly off the handle if I'm being threatened with an irrational lose/lose setup, behave rationally if someone set up a rational win/lose situation".

Like, think about it as an evolutionary process. Oro's mind generates candidate policies like:

• (1) "if someone tries to blackmail me, I give in".
• (2) "if someone tries to blackmail me, I kill them even if the dead man's switch is already created, and then I just eat the consequences".
• (3) "if someone tries to create situations where they win even if I kill them, I negotiate with them".
• (4) "if someone tries to create situations where they win even if I kill them, I kill them anyway and then just eat the consequences".

When a relevant situation comes up, it occurs to Oro to follow one of those policies, picked at random. After he acts on it, he traces the results, and sees whether the policy was beneficial or not. Over enough iterations, he'd see that following (2) instead of (1) decreases the rate of blackmail attempts he is facing, so (2) would be privileged over (1). On the other hand, using (4) over (3) would not actually decrease the extent to which people try to safeguard themselves against him; always killing people who defy him would only ensure that he's constantly burning himself triggering their countermeasures, and they keep building the countermeasures anyway.

So, even if no conscious calculation enters into it, he'd naturally gradient-descend towards using (2, 3) over (1, 4).