I’m not aware of a satisfactory treatment of blackmail (in the context of reflective decision theory). The main problem appears to be that it’s not clear what “blackmail” is, exactly, how to formally distinguish blackmail from trade.
I think blackmail can be taken as a form of threat. To use Schelling’s definition (Strategy of Conflict), a threat has the property that after the person being threatened fails to perform the specified action, the threatener does not want to carry out the threat any more. In other words, the threatener has a credibility problem: he has to convince the target of the threat that he will carry it out once he desires not to. This requires some form of pre-commitment, or an iterated game, or a successful bluff, or something along those lines.
What do you see as the distinguishing difference between blackmail and a threat? (I assume that blackmail is a subset of threats, but I suppose that might not be universally agreed.)
A man seduces a female movie star into a one night stand and secretly records a sex tape. He would prefer to blackmail the movie star for lots of money, but if that fails he would rather release the tape to the press for a smaller amount of money + prestige than he would just do nothing. The movie star’s preference ordering is for nothing to happen, for her to pay out, then lastly for the press to find out. The optimal choice is for her to pay out, because if she pre-commits to not give in to blackmail, she will receive the worst possible outcome.
This seems to fall squarely under blackmail, yet requires no pre-committment, iteration, or bluffing.
I would say that blackmail is the intersection of {things that are indistinguishable from threats} and {things that are indistinguishable from trade}. And yes, from the perspective of the blackmailers that includes both some things that are trade and some that are threats.
Yep. Even without the payoff—the blackmail immune agent is going to interact poorly with stupid blackmailer agent which simply doesn’t understand that the blackmail immune agent won’t pay. Or with evil blackmailer agent that just derives positive utility from your misfortune, which is the case with most blackmailers in practice.
The winning strategy depends to the ecosystem. The winning strategy among defect is defect, but among mixed tit-for-tat works great. The decision systems just tend to converge to a sort of self fulfilling prophecy solutions.
edit: i.e. there’s a rock-paper-skissors situation between different agents, they are not well ordered in ‘betterness’.
Actually, I’m not sure this does fall squarely under blackmail.
Consider the case where someone has a tape I don’t want shown to the press, and sells that tape to the press for money + prestige, and never gives me any choice in the matter. That’s clearly not blackmail. I’m not sure it becomes blackmail when they give me a choice to pay them instead, though the case could be made.
Or consider the case where it turns out I don’t mind having the tape shown (I want the publicity, say), and so the person sells the tape to the press, and everyone gets what they want. Also not blackmail. Not even clearly attempted blackmail, though the case could be made.
My point being that it seems to me that for me to legitimately call something “blackmail” it needs to be something the blackmailer threatens to do only because it makes me suffer more than paying them, not something that the blackmailer wants to do anyway for his own reasons that just happens to make me suffer.
I disagree that the essential element to blackmail is it must be done only to make me suffer. To this end I offer a scenario. (I’ve made it a little more like a story just for giggles).
You’ve just won the lottery, and the TV people interviewed you and your wife. Hurray! Shortly after, Angela your mistress calls you up. ”Congratulations. I saw on TV that you won the lottery… I also saw you had a wife. Things are over between us!” “Angela. I’m sorry for lying baby, but you’ve got a husband.” ”Irrelevant. But, you know, you’re rich now. How about you give me $4,000 a week for the next ten years, and I don’t tell your wife.” “Oh come on, you’ve got a husband!” ”Yeah, and you’re rich. I expect the money in a week.” click Well that’s what you get for cheating. Suddenly, Julia, your other mistress calls you up. ”Congratulations. I saw on TV that you won the lottery… I also saw you had a wife. Things are over between us!” “Julia. I’m sorry for lying baby.” ”Fuck you. I should tell her right now. But, you know, you’re rich now. How about you give me $4,000 a week for the next ten years, and I don’t tell your wife.” “Oh come on, not again” ”What’s that mean? Besides, you’re rich. I expect the money in a week.” click Way to go, two timer. However, unfortunately for you, the TV show calls you up and lets you know that there was a printing error and you did NOT in fact win the lottery. You quickly call your mistresses back. ”Angela, baby, turns out I didn’t win the lottery. I can’t possibly pay you.” “Don’t baby me. Well, I guess I won’t tell your wife then as long as you don’t tell my husband. I don’t want anyone to know this.” ”Thanks Angela, you’re the best.” Hurray, bullet dodged! ”Julia, baby, turns out I didn’t win the lottery. I can’t possibly pay you.” “Don’t baby me. Well, I guess I’ll tell your wife then asshole. Since you won’t pay me I’m gonna go post on facebook now!” ”But, I can’t pay you because the money never belonged to me.” ″Irrelevant. Sucks to be you!”
I argue that both of them were attempting to blackmail and Julia’s desire to follow through with it anyways doesn’t change anything. The actions would both feel like blackmail to me if I were on the receiving end, and the police would treat both of them as blackmail as well. Blackmail is just an attempt to get money in exchange for not releasing information; the mindset of the blackmailer does not affect it. This is why I agree with Vladimir Nesov, that blackmail exists in a blurry spot on the continuum of trade.
If you don’t classify Julia’s actions as blackmail, I would be curious what you do call it.
I classify Julia’s actions as inconsistent, mostly.
At time T1, Julia prefers to date me rather than end our relationship and tell my wife. At time T2, Julia prefers to end our relationship and tell my wife. The transition between T1 and T2 evidently has something to do with the transient belief that her silence was worth $4k/week, but what exactly it has to do with that belief is unclear, since by Julia’s own account the truth or falsehood of that belief is irrelevant.
If I take her account as definitive, I’m pretty clear that what Julia is doing is not blackmail… it reduces to “Hey, I’ve decided to tell your wife about us, and there’s nothing you can do to stop me.” It isn’t even a threat, it’s just early warning of the intent to harm me.
If I assume she’s lying about her motives, either consciously or with some degree of self-delusion, it might be blackmail. For example, if she believes I really can afford to pay her, and am just claiming poverty as a negotiating tactic, which she is countering by claiming not to care about the money, then it follows that she’s blackmailing me.
If I assume that she doesn’t really have relevant motives anymore, she just precommitted to reveal the information if I don’t pay her and now she’s following through on her previous precommitment, and the fact that the precommitment was made based on one set of beliefs about the world and she now knows those beliefs were false at the time doesn’t change the fact that the precommitment was made (“often wrong, never uncertain”), then she clearly blackmailed me once, and I guess it follows that she’s still blackmailing me… maybe? It seems that if she set up a mechanical device that posts the secret to Facebook unless fed $4k in quarters once a week, and then changed her mind and decided she’d rather just keep dating me, but was unable to turn the device off, we could in the same sense say she was still blackmailing me, albeit against her own will. That is at best a problematic sense of blackmail, but not clearly an incorrect one.
So it seems pretty clear that the blackmailer’s intentions play some role in my intuitions about what is or isn’t blackmail, albeit a murky one.
Of course, I could choose to ignore my linguistic intuitions and adopt a simpler definition which I apply formally. Nothing wrong with that, but it makes questions about what is and isn’t blackmail sort of silly.
For example, if I say any attempt to get money in exchange for not revealing information, regardless of my state of mind, is blackmail, then the following scenario is clearly blackmail:
I develop a practical, cheap, unlimited energy source in my basement. Julia says “Honey, I work for the oil company, and we will pay you $N/week for the rest of your life if you keep your mouth shut about that energy source.” I agree and take the money.
My native speaker’s intuitions are very clear that this is not blackmail, but I’m happy to use the term “blackmail” as a term of art to describe it if that makes communication with you (and perhaps Vladimir Nesov) easier.
The transition between T1 and T2 evidently has something to do with the transient belief that her silence was worth $4k/week
Sorry. I think I communicated unclearly, which is the danger of using stories instead of examples and is my fault entirely. At the very start of the story, Julia learns about your wife at the same time she learns about the lottery. She had previously thought you were single and the new information shifted her preference ordering.
Regarding the example you used (oil company & energy), I also hold it is not blackmail. If I use the previous definition of Blackmail being the act of making an attempt to get money in exchange for not revealing information, then the attempt is the crucial part in this case (whether it succeeds or not). The oil company offering me money is okay; me trying to get money out of the oil company is blackmail.
(nods) As noted elsewhere, I missed this and was entirely mistaken about Julia’s motives. I stand corrected. You were perfectly clear, I just wasn’t reading attentively enough.
Re: blackmail… OK. So, if I develop the technology and I approach the oil company and say “I have this technology, I’ll guarantee you exclusive rights to it for $N/week,” that’s blackmail?
At time T1, Julia prefers to date me rather than end our relationship and tell my wife.
I think in Xachariah’s story Julia did not know prior to seeing you on TV that you have a wife. So there was no time at which she had the preference you describe here.
Ah! Good point, I forgot about that. You’re absolutely right… throughout, she presumably prefers to break up with me than date me if I’m married. My error.
Not sure if this is what confused you or not, but it has since been pointed out to me that I was wrong; Julia does not necessarily (and ought not be understood to) have this preference, as she did not know about my wife at T1.
If we avoid the overloaded term “blackmail” and talk of threats vs. trade, Angela is threatening you whereas Julia is offering a trade. I agree that this example shows that “makes you suffer” is not the distinguishing element. It’s also interesting that you may not now if the situation is threat or trade (you may not know whether the mistress wants to tell your wife anyway).
I’m not sure how threats and trade are a real dichotomy rather than two fuzzy categories. Suppose I buy food. That’s basic trade. But at the same time a monopoly could raise the price of food a lot, and I would still have to buy it, and now it is the threat of starvation.
I can go fancy(N), and say, I won’t pay more than X for food, I would rather starve to death and then they get no more of my money, and if I can make it credible, and if the monopoly reasons in fancy(N-1) manner, they won’t raise the price above X because I won’t pay, but if monopoly reasons in the fancy(N) manner, it does exact same reasoning and concludes that it should ignore my threat to starve myself to death and not pay.
Most human agents seem to be tit for tat and mirror what ever you are doing, so if you are reasoning “i’ll just starve myself to death not to pay” they reason like “i’ll just raise the price regardless and the hell with what he does not pay”. The blackmail resistant agent is also blackmail resistance resistant.
I’m not sure how threats and trade are a real dichotomy rather than two fuzzy categories.
This is my position as well, blackmail probably doesn’t need to be considered as a separate case, reasonable behavior in such cases will probably just fall out from a sufficiently savvy bargaining algorithm.
Good point; haggling is a good example of a fuzzy boundary between threats and trade.
If A is willing to sell a widget for any price above $10, and B is willing to buy a widget for any price below $20, and there are no other buyers or sellers, then for any price X strictly between $10 and $20, A saying “I won’t sell for less than X” and B saying “I won’t sell for more than X” are both threats under my model.
Which means that agents that “naively” precommit to never respond to any threats (the way I understand them) will not reach an agreement when haggling. They’ll also fail at the Ultimatum game.
So there needs to be a better model for threats, possibly one that takes shelling points into account; or maybe there should be a special category for “the kind of threats it’s beneficial to precommit to ignore”.
Hmm the pre-commitment to ignore would depend on other agents and their pre-pre-commitment to ignore pre-commitments. It just goes recursive like Sherlock Holmes vs Moriarty, and when you go meta and try to look for ‘limit’ of recursion, it goes recursive again… i have a feeling that it is inherently a rock-paper-skissors situation where you can’t cheat like this robot. (I.e. I would suggest, at that point, to try to make a bunch of proofs of impossibility to narrow expectations down somewhat).
It’s not possible to coordinate in general against arbitrary opponents, like it’s impossible to predict what an arbitrary program does, but it’s advantageous for players to eventually coordinate their decisions (on some meta-level of precommitment). On one hand, players want to set prices their way, but on the other they want to close the trade eventually, and this tradeoff keeps the outcome from both extremes (“unfair” prices and impossibility of trade). Players have an incentive to setup some kind of Loebian cooperation (as in theseposts), which stops the go-meta regress, although each will try to set the point where cooperation happens in their favor.
I was thinking rather of Halting Problem—like impossibility, along with rock-paper-skissors situation that prevents declaring any one strategy, even the cooperative, as the ‘best’.
If difficulty of selecting and implementing a strategy is part of the tradeoff (so that more complicated strategies count as “worse” because of their difficulty, even if they promise an otherwise superior outcome), maybe there are “best” strategies in some sense, like there is a biggest natural number that you can actually write down in 30 seconds. (Such things would of course have the character of particular decisions, not of decision theory.)
Huh? Just start writing. The rule wasn’t “the number you can define in 30 seconds”, but simply “the number you can write down in 30 seconds”. Like the number of strawberries you can eat in 30 seconds, no paradox there!
Given a fixed state of knowledge about possible opponents and finite number of feasible options for your decision, there will be maximal decisions, even if in an iterated contest the players could cycle their decisions against updated opponents indefinitely.
That’s a promise, not a threat, by Schelling’s terminology. Once the movie start upholds her end of the bargain, the man has no incentive to keep his promise, and every incentive to break it.
Is there something game-theoretic about blackmail that makes it an identifiable subset of the group threats + promises? Note that Schelling also describes a negotiating position that is neither threat nor promise, but the combination of the two. And that’s not exactly blackmail either. I suspect you could come up with a blackmail scenario that fits any of the three groupings.
That’s not blackmail at all. It seems like blackmail because of the questionable morality of selling secretly recorded sex tapes, but giving the movie star the chance to buy the tape first doesn’t make the whole thing any less moral than it would be without that chance, and unlike real blackmail the movie star being known not to respond to blackmail doesn’t help in any way.
Consider this variation: Instead of a secret tape the movie star voluntarily participated in an amateur porno that was intended to be publicly released from the beginning, but held up for some reason, and all that happened before the movie star became famous in the first place. The producer knows that releasing the tape will hurt her career and offers her to buy the tape to prevent it from being released. This doesn’t seem like blackmail at all, and the only change was to the moral (and legal) status of releasing the tape, not to the trade.
Something similar to this happened to Cameron Diaz although the rights to resell the photos were questionable. She posed topless in some bondage shots for a magazine, but they were never printed. The photographer kept the shots and the recording of the photo shoot for ten years until one of the Charlie’s Angel’s films was about to come out. He offered them to her for a couple of million or he would sell them to the highest bidder. The courts didn’t buy that he was just offering her first right of refusal and sentenced him for attempted grand theft (blackmail), forgery, and perjury (for modifying release forms and lying about it). Link
Are you sure you aren’t just pattern matching to similarity to known types of blackmail? Do you think it would be useful for an AI to classify it the same way (which was the starting point of this thread)?
Your link doesn’t go into much detail, but it seems like he was convicted because he was lying and making up the negative consequences he threatened her with, and like he was going out of his way to make the consequences of selling to someone else as bad as possible rather than maximizing revenue (or at least making her believe so). That would qualify this case as blackmail under the definition above, unlike either of our hypothetical examples.
I’m assuming that the movie star is at least reasonably smart. The first thing that comes to mind is periodic payments that decrease over time, with the value hovering just above what magazines are willing to pay + bragging rights, since people are less impressed with a 10 year old sex tape than a brand new one.
Eventually the payments would be stopped when either the man has more to lose from releasing the tape than staying quiet (eg, he’s settled down and married now) or the movie star values money more than the loss of prestige from a scandal (eg, another scandal breaks or she stops getting roles anyway).
I’m sure there are other ways to solve the problem as well, but regardless it’s a technical hurdle rather than an absolute one.
Continuum behaviors are discussed in some detail by Schelling, and interestingly they can be used by both parties. Here they make the blackmail more effective. If the payment is lump-sum, the blackmailer can’t be trusted, and so the movie start won’t pay. The continuous payment option gives her a way to pay the blackmailer and expect him to stay quiet, which makes her more vulnerable to blackmail in the first place.
Continuous options can also be used to derail threats, when the person being threatened can act incrementally and there is no bright line to force the action (assuming the threat is to carry out a single action).
I’m not aware of a satisfactory treatment of blackmail (in the context of reflective decision theory). The main problem appears to be that it’s not clear what “blackmail” is, exactly, how to formally distinguish blackmail from trade.
I think blackmail can be taken as a form of threat. To use Schelling’s definition (Strategy of Conflict), a threat has the property that after the person being threatened fails to perform the specified action, the threatener does not want to carry out the threat any more. In other words, the threatener has a credibility problem: he has to convince the target of the threat that he will carry it out once he desires not to. This requires some form of pre-commitment, or an iterated game, or a successful bluff, or something along those lines.
What do you see as the distinguishing difference between blackmail and a threat? (I assume that blackmail is a subset of threats, but I suppose that might not be universally agreed.)
Counterexample:
A man seduces a female movie star into a one night stand and secretly records a sex tape. He would prefer to blackmail the movie star for lots of money, but if that fails he would rather release the tape to the press for a smaller amount of money + prestige than he would just do nothing. The movie star’s preference ordering is for nothing to happen, for her to pay out, then lastly for the press to find out. The optimal choice is for her to pay out, because if she pre-commits to not give in to blackmail, she will receive the worst possible outcome.
This seems to fall squarely under blackmail, yet requires no pre-committment, iteration, or bluffing.
It does, making ‘blackmail’ the wrong term to use when considering game theory scenarios. Some are ‘threats’, some are simply trade.
I would say that blackmail is the intersection of {things that are indistinguishable from threats} and {things that are indistinguishable from trade}. And yes, from the perspective of the blackmailers that includes both some things that are trade and some that are threats.
Yep. Even without the payoff—the blackmail immune agent is going to interact poorly with stupid blackmailer agent which simply doesn’t understand that the blackmail immune agent won’t pay. Or with evil blackmailer agent that just derives positive utility from your misfortune, which is the case with most blackmailers in practice.
The winning strategy depends to the ecosystem. The winning strategy among defect is defect, but among mixed tit-for-tat works great. The decision systems just tend to converge to a sort of self fulfilling prophecy solutions.
edit: i.e. there’s a rock-paper-skissors situation between different agents, they are not well ordered in ‘betterness’.
Actually, I’m not sure this does fall squarely under blackmail.
Consider the case where someone has a tape I don’t want shown to the press, and sells that tape to the press for money + prestige, and never gives me any choice in the matter. That’s clearly not blackmail. I’m not sure it becomes blackmail when they give me a choice to pay them instead, though the case could be made.
Or consider the case where it turns out I don’t mind having the tape shown (I want the publicity, say), and so the person sells the tape to the press, and everyone gets what they want. Also not blackmail. Not even clearly attempted blackmail, though the case could be made.
My point being that it seems to me that for me to legitimately call something “blackmail” it needs to be something the blackmailer threatens to do only because it makes me suffer more than paying them, not something that the blackmailer wants to do anyway for his own reasons that just happens to make me suffer.
I disagree that the essential element to blackmail is it must be done only to make me suffer. To this end I offer a scenario. (I’ve made it a little more like a story just for giggles).
I argue that both of them were attempting to blackmail and Julia’s desire to follow through with it anyways doesn’t change anything. The actions would both feel like blackmail to me if I were on the receiving end, and the police would treat both of them as blackmail as well. Blackmail is just an attempt to get money in exchange for not releasing information; the mindset of the blackmailer does not affect it. This is why I agree with Vladimir Nesov, that blackmail exists in a blurry spot on the continuum of trade.
If you don’t classify Julia’s actions as blackmail, I would be curious what you do call it.
I classify Julia’s actions as inconsistent, mostly.
At time T1, Julia prefers to date me rather than end our relationship and tell my wife.
At time T2, Julia prefers to end our relationship and tell my wife.
The transition between T1 and T2 evidently has something to do with the transient belief that her silence was worth $4k/week, but what exactly it has to do with that belief is unclear, since by Julia’s own account the truth or falsehood of that belief is irrelevant.
If I take her account as definitive, I’m pretty clear that what Julia is doing is not blackmail… it reduces to “Hey, I’ve decided to tell your wife about us, and there’s nothing you can do to stop me.” It isn’t even a threat, it’s just early warning of the intent to harm me.
If I assume she’s lying about her motives, either consciously or with some degree of self-delusion, it might be blackmail. For example, if she believes I really can afford to pay her, and am just claiming poverty as a negotiating tactic, which she is countering by claiming not to care about the money, then it follows that she’s blackmailing me.
If I assume that she doesn’t really have relevant motives anymore, she just precommitted to reveal the information if I don’t pay her and now she’s following through on her previous precommitment, and the fact that the precommitment was made based on one set of beliefs about the world and she now knows those beliefs were false at the time doesn’t change the fact that the precommitment was made (“often wrong, never uncertain”), then she clearly blackmailed me once, and I guess it follows that she’s still blackmailing me… maybe? It seems that if she set up a mechanical device that posts the secret to Facebook unless fed $4k in quarters once a week, and then changed her mind and decided she’d rather just keep dating me, but was unable to turn the device off, we could in the same sense say she was still blackmailing me, albeit against her own will. That is at best a problematic sense of blackmail, but not clearly an incorrect one.
So it seems pretty clear that the blackmailer’s intentions play some role in my intuitions about what is or isn’t blackmail, albeit a murky one.
Of course, I could choose to ignore my linguistic intuitions and adopt a simpler definition which I apply formally. Nothing wrong with that, but it makes questions about what is and isn’t blackmail sort of silly.
For example, if I say any attempt to get money in exchange for not revealing information, regardless of my state of mind, is blackmail, then the following scenario is clearly blackmail:
My native speaker’s intuitions are very clear that this is not blackmail, but I’m happy to use the term “blackmail” as a term of art to describe it if that makes communication with you (and perhaps Vladimir Nesov) easier.
Sorry. I think I communicated unclearly, which is the danger of using stories instead of examples and is my fault entirely. At the very start of the story, Julia learns about your wife at the same time she learns about the lottery. She had previously thought you were single and the new information shifted her preference ordering.
Regarding the example you used (oil company & energy), I also hold it is not blackmail. If I use the previous definition of Blackmail being the act of making an attempt to get money in exchange for not revealing information, then the attempt is the crucial part in this case (whether it succeeds or not). The oil company offering me money is okay; me trying to get money out of the oil company is blackmail.
And also sometimes okay. The distinction isn’t “okay” vs blackmail. It is blackmail vs not-blackmail and “okay” vs not-okay.
(nods) As noted elsewhere, I missed this and was entirely mistaken about Julia’s motives. I stand corrected. You were perfectly clear, I just wasn’t reading attentively enough.
Re: blackmail… OK. So, if I develop the technology and I approach the oil company and say “I have this technology, I’ll guarantee you exclusive rights to it for $N/week,” that’s blackmail?
I’d say it’s much closer to blackmail than the original oil company scenario.
I suppose I agree with that, but I wouldn’t call either of them blackmail. Would you?
I think in Xachariah’s story Julia did not know prior to seeing you on TV that you have a wife. So there was no time at which she had the preference you describe here.
Ah! Good point, I forgot about that. You’re absolutely right… throughout, she presumably prefers to break up with me than date me if I’m married. My error.
Something about this confused me.
Not sure if this is what confused you or not, but it has since been pointed out to me that I was wrong; Julia does not necessarily (and ought not be understood to) have this preference, as she did not know about my wife at T1.
No, it was just you talking about your wife in first person! :)
Ah. Well, my husband had one once, I suppose I might feel left out.
If we avoid the overloaded term “blackmail” and talk of threats vs. trade, Angela is threatening you whereas Julia is offering a trade. I agree that this example shows that “makes you suffer” is not the distinguishing element. It’s also interesting that you may not now if the situation is threat or trade (you may not know whether the mistress wants to tell your wife anyway).
I’m not sure how threats and trade are a real dichotomy rather than two fuzzy categories. Suppose I buy food. That’s basic trade. But at the same time a monopoly could raise the price of food a lot, and I would still have to buy it, and now it is the threat of starvation.
I can go fancy(N), and say, I won’t pay more than X for food, I would rather starve to death and then they get no more of my money, and if I can make it credible, and if the monopoly reasons in fancy(N-1) manner, they won’t raise the price above X because I won’t pay, but if monopoly reasons in the fancy(N) manner, it does exact same reasoning and concludes that it should ignore my threat to starve myself to death and not pay.
Most human agents seem to be tit for tat and mirror what ever you are doing, so if you are reasoning “i’ll just starve myself to death not to pay” they reason like “i’ll just raise the price regardless and the hell with what he does not pay”. The blackmail resistant agent is also blackmail resistance resistant.
This is my position as well, blackmail probably doesn’t need to be considered as a separate case, reasonable behavior in such cases will probably just fall out from a sufficiently savvy bargaining algorithm.
I agree with this, incidentally.
Good point; haggling is a good example of a fuzzy boundary between threats and trade.
If A is willing to sell a widget for any price above $10, and B is willing to buy a widget for any price below $20, and there are no other buyers or sellers, then for any price X strictly between $10 and $20, A saying “I won’t sell for less than X” and B saying “I won’t sell for more than X” are both threats under my model.
Which means that agents that “naively” precommit to never respond to any threats (the way I understand them) will not reach an agreement when haggling. They’ll also fail at the Ultimatum game.
So there needs to be a better model for threats, possibly one that takes shelling points into account; or maybe there should be a special category for “the kind of threats it’s beneficial to precommit to ignore”.
Hmm the pre-commitment to ignore would depend on other agents and their pre-pre-commitment to ignore pre-commitments. It just goes recursive like Sherlock Holmes vs Moriarty, and when you go meta and try to look for ‘limit’ of recursion, it goes recursive again… i have a feeling that it is inherently a rock-paper-skissors situation where you can’t cheat like this robot. (I.e. I would suggest, at that point, to try to make a bunch of proofs of impossibility to narrow expectations down somewhat).
It’s not possible to coordinate in general against arbitrary opponents, like it’s impossible to predict what an arbitrary program does, but it’s advantageous for players to eventually coordinate their decisions (on some meta-level of precommitment). On one hand, players want to set prices their way, but on the other they want to close the trade eventually, and this tradeoff keeps the outcome from both extremes (“unfair” prices and impossibility of trade). Players have an incentive to setup some kind of Loebian cooperation (as in these posts), which stops the go-meta regress, although each will try to set the point where cooperation happens in their favor.
I was thinking rather of Halting Problem—like impossibility, along with rock-paper-skissors situation that prevents declaring any one strategy, even the cooperative, as the ‘best’.
If difficulty of selecting and implementing a strategy is part of the tradeoff (so that more complicated strategies count as “worse” because of their difficulty, even if they promise an otherwise superior outcome), maybe there are “best” strategies in some sense, like there is a biggest natural number that you can actually write down in 30 seconds. (Such things would of course have the character of particular decisions, not of decision theory.)
There is not a biggest natural number that you can actually write down in thirty seconds—that’s equivalent to Berry’s paradox.
Huh? Just start writing. The rule wasn’t “the number you can define in 30 seconds”, but simply “the number you can write down in 30 seconds”. Like the number of strawberries you can eat in 30 seconds, no paradox there!
I was reading “write down” more generally than “write down each digit of in base ten,” but I guess that’s not how you meant it.
Hmm if it was a programming contest I would expect non-transitive ‘betterness’.
Given a fixed state of knowledge about possible opponents and finite number of feasible options for your decision, there will be maximal decisions, even if in an iterated contest the players could cycle their decisions against updated opponents indefinitely.
That’s a promise, not a threat, by Schelling’s terminology. Once the movie start upholds her end of the bargain, the man has no incentive to keep his promise, and every incentive to break it.
Is there something game-theoretic about blackmail that makes it an identifiable subset of the group threats + promises? Note that Schelling also describes a negotiating position that is neither threat nor promise, but the combination of the two. And that’s not exactly blackmail either. I suspect you could come up with a blackmail scenario that fits any of the three groupings.
That’s not blackmail at all. It seems like blackmail because of the questionable morality of selling secretly recorded sex tapes, but giving the movie star the chance to buy the tape first doesn’t make the whole thing any less moral than it would be without that chance, and unlike real blackmail the movie star being known not to respond to blackmail doesn’t help in any way.
Consider this variation: Instead of a secret tape the movie star voluntarily participated in an amateur porno that was intended to be publicly released from the beginning, but held up for some reason, and all that happened before the movie star became famous in the first place. The producer knows that releasing the tape will hurt her career and offers her to buy the tape to prevent it from being released. This doesn’t seem like blackmail at all, and the only change was to the moral (and legal) status of releasing the tape, not to the trade.
I still classify it as blackmail.
Something similar to this happened to Cameron Diaz although the rights to resell the photos were questionable. She posed topless in some bondage shots for a magazine, but they were never printed. The photographer kept the shots and the recording of the photo shoot for ten years until one of the Charlie’s Angel’s films was about to come out. He offered them to her for a couple of million or he would sell them to the highest bidder. The courts didn’t buy that he was just offering her first right of refusal and sentenced him for attempted grand theft (blackmail), forgery, and perjury (for modifying release forms and lying about it). Link
Are you sure you aren’t just pattern matching to similarity to known types of blackmail? Do you think it would be useful for an AI to classify it the same way (which was the starting point of this thread)?
Your link doesn’t go into much detail, but it seems like he was convicted because he was lying and making up the negative consequences he threatened her with, and like he was going out of his way to make the consequences of selling to someone else as bad as possible rather than maximizing revenue (or at least making her believe so). That would qualify this case as blackmail under the definition above, unlike either of our hypothetical examples.
But you’re forgetting the man’s best option: get lots of money from the movie star and get a smaller amount from the press.
Edit: Ah, not lump-sum payment. I can see how that would work then.
And be a lot more vulnerable to criminal charges for the blackmail.
I’m assuming that the movie star is at least reasonably smart. The first thing that comes to mind is periodic payments that decrease over time, with the value hovering just above what magazines are willing to pay + bragging rights, since people are less impressed with a 10 year old sex tape than a brand new one.
Eventually the payments would be stopped when either the man has more to lose from releasing the tape than staying quiet (eg, he’s settled down and married now) or the movie star values money more than the loss of prestige from a scandal (eg, another scandal breaks or she stops getting roles anyway).
I’m sure there are other ways to solve the problem as well, but regardless it’s a technical hurdle rather than an absolute one.
Continuum behaviors are discussed in some detail by Schelling, and interestingly they can be used by both parties. Here they make the blackmail more effective. If the payment is lump-sum, the blackmailer can’t be trusted, and so the movie start won’t pay. The continuous payment option gives her a way to pay the blackmailer and expect him to stay quiet, which makes her more vulnerable to blackmail in the first place.
Continuous options can also be used to derail threats, when the person being threatened can act incrementally and there is no bright line to force the action (assuming the threat is to carry out a single action).