Okay, maybe not me, but someone I know, and that's what the title would be if he wrote it.  Newcomb's problem and Kavka's toxin puzzle are more than just curiosities relevant to artificial intelligence theory.  Like a lot of thought experiments, they approximately happen.  They illustrate robust issues with causal decision theory that can deeply affect our everyday lives.

Yet somehow it isn't mainstream knowledge that these are more than merely abstract linguistic issues, as evidenced by this comment thread (please no Karma sniping of the comments, they are a valuable record).  Scenarios involving brain scanning, decision simulation, etc., can establish their validy and future relevance, but not that they are already commonplace.  For the record, I want to provide an already-happened, real-life account that captures the Newcomb essence and explicitly describes how.

So let's say my friend is named Joe.  In his account, Joe is very much in love with this girl named Omega… er… Kate, and he wants to get married.  Kate is somewhat traditional, and won't marry him unless he proposes, not only in the sense of explicitly asking her, but also expressing certainty that he will never try to leave her if they do marry. 

Now, I don't want to make up the ending here.  I want to convey the actual account, in which Joe's beliefs are roughly schematized as follows: 

1. if he proposes sincerely, she is effectively sure to believe it.
2. if he proposes insincerely, she will 50% likely believe it.
3. if she believes his proposal, she will 80% likely say yes.
4. if she doesn't believe his proposal, she will surely say no, but will not be significantly upset in comparison to the significance of marriage.
5. if they marry, Joe will 90% likely be happy, and will 10% likely be unhappy.

He roughly values the happy and unhappy outcomes oppositely:

1. being happily married to Kate:  125 megautilons
2. being unhapily married to Kate:  -125 megautilons.

So what should he do?  What should this real person have actually done?¹  Well, as in Newcomb, these beliefs and utilities present an interesting and quantifiable problem…

Related to: The True Prisoner's Dilemma, Newcomb's Problem and Regret of Rationality

In The True Prisoner's Dilemma, Eliezer Yudkowsky pointed out a critical problem with the way the Prisoner's Dilemma is taught: the distinction between utility and avoided-jail-time is not made clear.  The payoff matrix is supposed to represent the former, even as its numerical values happen to coincidentally match the latter.  And worse, people don't naturally assign utility as per the standard payoff matrix: their compassion for the friend in the "accomplice" role means they wouldn't feel quite so good about a "successful" backstabbing, nor quite so bad about being backstabbed.  ("Hey, at least I didn't rat out a friend.")

For that reason, you rarely encounter a true Prisoner's Dilemma, even an iterated one.  The above complications prevent real-world payoff matrices from working out that way.

Which brings us to another unfortunate example of this misunderstanding being taught.

Followup to:  Newcomb's Problem and Regret of Rationality, Towards a New Decision Theory

Wei Dai asked:

"Why didn't you mention earlier that your timeless decision theory mainly had to do with logical uncertainty? It would have saved people a lot of time trying to guess what you were talking about."

...

All right, fine, here's a fast summary of the most important ingredients that go into my "timeless decision theory".  This isn't so much an explanation of TDT, as a list of starting ideas that you could use to recreate TDT given sufficient background knowledge.  It seems to me that this sort of thing really takes a mini-book, but perhaps I shall be proven wrong.

The one-sentence version is:  Choose as though controlling the logical output of the abstract computation you implement, including the output of all other instantiations and simulations of that computation.

The three-sentence version is:  Factor your uncertainty over (impossible) possible worlds into a causal graph that includes nodes corresponding to the unknown outputs of known computations; condition on the known initial conditions of your decision computation to screen off factors influencing the decision-setup; compute the counterfactuals in your expected utility formula by surgery on the node representing the logical output of that computation.

Suppose you're out in the desert, running out of water, and soon to die - when someone in a motor vehicle drives up next to you.  Furthermore, the driver of the motor vehicle is a perfectly selfish ideal game-theoretic agent, and even further, so are you; and what's more, the driver is Paul Ekman, who's really, really good at reading facial microexpressions.  The driver says, "Well, I'll convey you to town if it's in my interest to do so - so will you give me $100 from an ATM when we reach town?"

Now of course you wish you could answer "Yes", but as an ideal game theorist yourself, you realize that, once you actually reach town, you'll have no further motive to pay off the driver.  "Yes," you say.  "You're lying," says the driver, and drives off leaving you to die.

If only you weren't so rational!

This is the dilemma of Parfit's Hitchhiker, and the above is the standard resolution according to mainstream philosophy's causal decision theory, which also two-boxes on Newcomb's Problem and defects in the Prisoner's Dilemma.  Of course, any self-modifying agent who expects to face such problems - in general, or in particular - will soon self-modify into an agent that doesn't regret its "rationality" so much.  So from the perspective of a self-modifying-AI-theorist, classical causal decision theory is a wash.  And indeed I've worked out a theory, tentatively labeled "timeless decision theory", which covers these three Newcomblike problems and delivers a first-order answer that is already reflectively consistent, without need to explicitly consider such notions as "precommitment".  Unfortunately this "timeless decision theory" would require a long sequence to write up, and it's not my current highest writing priority unless someone offers to let me do a PhD thesis on it.

However, there are some other timeless decision problems for which I do not possess a general theory.

For example, there's a problem introduced to me by Gary Drescher's marvelous Good and Real (OOPS: The below formulation was independently invented by Vladimir Nesov; Drescher's book actually contains a related dilemma in which box B is transparent, and only contains $1M if Omega predicts you will one-box whether B appears full or empty, and Omega has a 1% error rate) which runs as follows:

Suppose Omega (the same superagent from Newcomb's Problem, who is known to be honest about how it poses these sorts of dilemmas) comes to you and says:

"I just flipped a fair coin.  I decided, before I flipped the coin, that if it came up heads, I would ask you for $1000.  And if it came up tails, I would give you $1,000,000 if and only if I predicted that you would give me $1000 if the coin had come up heads.  The coin came up heads - can I have $1000?"

Previously in series:  Beware of Other-Optimizing
Followup to:  Bystander Apathy

Yesterday I convered the bystander effect, aka bystander apathy: given a fixed problem situation, a group of bystanders is actually less likely to act than a single bystander.  The standard explanation for this result is in terms of pluralistic ignorance (if it's not clear whether the situation is an emergency, each person tries to look calm while darting their eyes at the other bystanders, and sees other people looking calm) and diffusion of responsibility (everyone hopes that someone else will be first to act; being part of a crowd diminishes the individual pressure to the point where no one acts).

Which may be a symptom of our hunter-gatherer coordination mechanisms being defeated by modern conditions.  You didn't usually form task-forces with strangers back in the ancestral environment; it was mostly people you knew.  And in fact, when all the subjects know each other, the bystander effect diminishes.

So I know this is an amazing and revolutionary observation, and I hope that I don't kill any readers outright from shock by saying this: but people seem to have a hard time reacting constructively to problems encountered over the Internet.

Perhaps because our innate coordination instincts are not tuned for:

• Being part of a group of strangers.  (When all subjects know each other, the bystander effect diminishes.)
• Being part of a group of unknown size, of strangers of unknown identity.
• Not being in physical contact (or visual contact); not being able to exchange meaningful glances.
• Not communicating in real time.
• Not being much beholden to each other for other forms of help; not being codependent on the group you're in.
• Being shielded from reputational damage, or the fear of reputational damage, by your own apparent anonymity; no one is visibly looking at you, before whom your reputation might suffer from inaction.
• Being part of a large collective of other inactives; no one will single out you to blame.
• Not hearing a voiced plea for help.