I’m getting married. We decided to take marriage vows very seriously, and write vows that we will be fully committed to uphold. These vows are going to be a commitment no weaker than any promise I ever made or any contract I ever signed. Therefore, it is very important to avoid serious errors in their content.
I’m interested to hear feedback of the form “making these vows might turn out to be a big mistake for you, and here is why”[1] or of the form “here is how the spirit of these vows can be implemented better”. Given that this is a community which nurtures security mindset, I have great expectations :) More precisely, I am less interested in extreme nitpicking / rule-lawyering, since that should be neutralized by the Vow of Good Faith anyway (but tell me if you think I’m wrong about this!) and more in serious problems that can arise in at least semi-realistic situations. (Of course, since many of us here expect a Singularity in a few decades, semi-realistic is not a very high bar ;)
Without further ado, the vows:
[EDIT 2022-07-15: The following text has been edited to match the final version of the Vows (that we took on 2021-08-29)]
I, [name], solemnly pledge to [name] three sacred Vows as I take [pronoun] to be my [spouse]. These vows are completely sincere, literal, binding and irrevocable from the moment both of us take the Vows, unless the marriage is dissolved or my [spouse] unconscionably[2] breaks [pronoun]’s own Vows which I believe in all likelihood will never happen. Let everyone present be my witness.
The First Vow is that of Honesty. I will never set out to deceive my [spouse] on purpose without [pronoun]’s unambiguous consent[3], without exception. I will also never withhold information that [pronoun] would in hindsight prefer to know[4]. The only exception to the latter is when this information was given to me in confidence by a third party as part of an agreement which was made in compliance with all Vows[5]. If for any reason I break my vow, I will act to repair the error as fast as reasonably possible.
The Second Vow is that of Concord. Everything I do will be according to the policy which is the Nash solution to the bargaining problem defined by my [spouse]’s and my own priors and utility functions, with the disagreement point set at the counterfactual in which we did not marry. I will act as if we made all precommitments that would a priori be beneficial from a Nash bargaining point of view[6]. If our utility functions change, we will effectively perform another Nash bargaining with the previous policy as the disagreement point. Moreover, if I deviate from this policy for any reason then I will return to optimal behavior as soon as possible, while preserving my [spouse]’s a priori expected utility if at all possible[7]. Finally, a hypothetical act of dissolving this marriage would also fall under the purview of this Vow[8].
The Third Vow is that of Good Faith, which augments and clarifies all three Vows. The spirit of the Vows takes precedence over the letter. When there’s some doubt or dispute as to how to interpret the Vows, the chosen interpretation should be that which my [spouse] and I would agree on at the time of our wedding, in the counterfactual in which the source of said doubt or dispute would be revealed to us and understood by us with all of its implications at that time as well as we understand it at the time it actually surfaced[9].
Conditional on the assumption that my decision to marry is about as well-grounded as one can expect. I am not soliciting criticism of my choice of spouse!
That is, information X such that if the spouse knew X, they would believe it’s good that they found out about it. This excludes information which is not important (knowing X is practically useless) and infohazards (knowing X is actively harmful).
If I enter an agreement with a third party in violation of the Vow of Concord, the Vow of Honesty takes precedence over the agreement and I might have to violate the latter and pay whatever fine is necessary.
We are taking an “updateless” perspective here. The disagreement point is fixed in the counterfactual in which we didn’t marry in the first place, it does not move to the counterfactual of divorce. Notice also that marriage is guaranteed to be an a priori Pareto improvement over no-marriage because this is our current estimate, even if it turns out to be false a posteriori.
If the violation shifts the Pareto frontier such that the previous optimum is outside of it, the new Pareto optimum is chosen s.t. the violating party bears the cost.
This makes all of the Vows weightier than they otherwise would be. The Vows can be unmade by dissolving the marriage, but the act of dissolving the marriage is in itself subject to the Vow of Concord, which limits the ability to dissolve it unilaterally.
In other words, interpretation is according to the extrapolated volition of us at the time of our wedding, where the extrapolation is towards our knowledge and intellectual ability at the time of making the judgment.
My Marriage Vows
I’m getting married. We decided to take marriage vows very seriously, and write vows that we will be fully committed to uphold. These vows are going to be a commitment no weaker than any promise I ever made or any contract I ever signed. Therefore, it is very important to avoid serious errors in their content.
I’m interested to hear feedback of the form “making these vows might turn out to be a big mistake for you, and here is why”[1] or of the form “here is how the spirit of these vows can be implemented better”. Given that this is a community which nurtures security mindset, I have great expectations :) More precisely, I am less interested in extreme nitpicking / rule-lawyering, since that should be neutralized by the Vow of Good Faith anyway (but tell me if you think I’m wrong about this!) and more in serious problems that can arise in at least semi-realistic situations. (Of course, since many of us here expect a Singularity in a few decades, semi-realistic is not a very high bar ;)
Without further ado, the vows:
[EDIT 2022-07-15: The following text has been edited to match the final version of the Vows (that we took on 2021-08-29)]
Conditional on the assumption that my decision to marry is about as well-grounded as one can expect. I am not soliciting criticism of my choice of spouse!
Meaning that it’s a grave or persistent violation rather than a minor lapse.
Consent is mentioned to allow us to e.g. play tabletop games where you’re supposed to deceive each other.
That is, information X such that if the spouse knew X, they would believe it’s good that they found out about it. This excludes information which is not important (knowing X is practically useless) and infohazards (knowing X is actively harmful).
If I enter an agreement with a third party in violation of the Vow of Concord, the Vow of Honesty takes precedence over the agreement and I might have to violate the latter and pay whatever fine is necessary.
We are taking an “updateless” perspective here. The disagreement point is fixed in the counterfactual in which we didn’t marry in the first place, it does not move to the counterfactual of divorce. Notice also that marriage is guaranteed to be an a priori Pareto improvement over no-marriage because this is our current estimate, even if it turns out to be false a posteriori.
If the violation shifts the Pareto frontier such that the previous optimum is outside of it, the new Pareto optimum is chosen s.t. the violating party bears the cost.
This makes all of the Vows weightier than they otherwise would be. The Vows can be unmade by dissolving the marriage, but the act of dissolving the marriage is in itself subject to the Vow of Concord, which limits the ability to dissolve it unilaterally.
In other words, interpretation is according to the extrapolated volition of us at the time of our wedding, where the extrapolation is towards our knowledge and intellectual ability at the time of making the judgment.