Some interesting stuff in here. The rationalist theory elements are particularly intriguing.
I suspect I'd enjoy breaking that down and discussing it after you write the update. I don't want to bias you yet, but I'm proud of how much is packed into that notion.
My real world formulation of the idea also includes '
Aumann's Agreement Theorem', 'tit-for-tat with forgiveness', a bunch of stuff from information theory (esp. ideas on signal to noise ratio and error coding), and the idea that '
Bayesian priors ought to allow for the possibility that you are the other person', but that seems a bit much to ask of Hazou.
At least the veil of ignorance thought experiment, the concept of an iterated prisoner's dilemma, and the tit-for-tat strategy are explainable w/ Hazou's level of sophistication, even if the elemental nations probably doesn't have that exact vocabulary.
OTOH, I get this ominous feeling I'm misinterpreting what this is meant to accomplish. Who knows.
I think you might be a bit.
It's mostly what I've said before, about how Orochimaru doesn't really have equals, intellectually or power, so us making this argument to him doesn't really stop him from conducting what are (to us) unethical experimentation on other peoples.
It doesn't
stop him from acting unethically, no. He'll still be able to push his weight around and end up benefiting his terminal values. This isn't a magic argument that will make his motives less selfish.
I am trying to provide Orochimaru with a different perspective/framing on the problem of making good decisions. One where the key sticking point is figuring out what definition of 'sufficiently similar' is best for him.
Hazou has relatively little power and goals that benefit significantly from cooperation, hence his optimal definition of 'sufficiently similar' is going to be rather wide and cover lots of agents with similar structure and goals.
Orochimaru on the other hand will probably have a narrower optimal definition of 'sufficiently similar' since the space of agents whom he could usefully cooperate with is smaller. If we're unlucky Oro's definition of 'sufficiently similar' won't include any others who actually exist, but even then it will still cover many possible agents who aren't Orochimaru.
Basically, it's a line of thought that leads from purely selfish motives to actions that approximate altruism for groups that he's capable of cooperating with.
A bunch of other stuff in the plan is there to signal that:
- We're a group he can usefully cooperate with
- We are capable of usefully cooperating with many more actors than he is.
- If he works with us he can derive benefit from not just our capabilities but those of our allies, even those allies he couldn't interface with directly.
If he doesn't enjoy dealing with most people (which seems likely) then that places us as valuable intermediary between him and the outside world. (esp. stupid people and politics)
And I did see your response to my initial remark, and I still don't agree that Orochimaru would see us as part of the same intellectual reference class;
Ahh, I think you mistake me. I'm not trying to convince Orochimaru that we're as intelligent or as powerful as he is, just that we think like him. In particular, that we are enough like him that straightforward cooperation is the best tactic, as opposed to manipulation, force, or some other non win-win tactic.
we're nowhere near him in that time frame, we probably won't get to that position without becoming significantly more powerful and having a few more impossible things to our name, and if Orochimaru does start to see us like that, he'll also start thinking of ways to accidentally us to keep us from interfering with his research.
The point is to stop him from seeing us an an enemy/obstacle in the first place or, failing that, make his optimal move be defecting ASAP in an arena with limited stakes. If he's trustworthy then I want to convince him that we're trustworthy too, if he's not trustworthy then I want to find out as soon as possible.