3 ms·

You're presenting a future where there is (in the limit) no point in earning money. Lately there was a discussion about how in the future it makes no sense to b

You're presenting a future where there is (in the limit) no point in earning money. Lately there was a discussion about how in the future it makes no sense to buy insurance if someone is willing to sell it to you, you are always better off saving and you won't get insured if you actually need it[0]. I understand all these specific scenarios may be kinda-dismissed with some commonsensical argumentation, but has someone somewhere (in the academia?) explored how perfect information could break the economy/capitalism in some way? It looks like an interesting problem theoretically. (I don't feel like arguing either way, just would like a deeper look on this.) It's my understanding that economists actually think that perfect information should be a prerequisite/an effect of efficient markets, but did they really mean the same thing when this was formulated in the last century? [0] this tree may make this impression https://news.ycombinator.com/item?id=21778890 https://news.ycombinator.com/item?id=21778890

wutbrodo 7y ago
> there was a discussion about how in the future it makes no sense to buy insurance if someone is willing to sell it to you, you are always better off saving and you won't get insured if you actually need it[0]. This is a misunderstanding of how insurance works. It makes sense to self-insured in expected $ value, but not necessarily in expected utility. There's nonzero value to lowering volatility[1], and the insurance company creates surplus by coordinating the pooling of risk across many people. For an easy example of the difference, for an easy example of the difference, I wouldn't take a 50/50 coin flip that would either cost me $1 million dollars or gain me $1.1 million; even though the expected value is $50k, because I couldn't afford such a high chance of financial ruin. I would however, sell the chance at this bet to an insurance company for $40k, and if they were large enough to absorb the loss (or bought many such policies), they would take it and gain $10k minus the (lower) value they place on the risk. (All numbers are made up, obviously). [1] this is a simple enough concept that you can derive it graphically from a utility/$ graph, and requires only the assumption of diminishing marginal utility of $
- stereolambda 7y ago
  Yeah, I wanted to give an example and disclaim redoing that specific discussion if possible. But just to provide a different view, what you're saying is fine theoretically, but in practice often people see the insurance situation as "some bad costly thing X can happen to me, I want to pay reasonable money to not have to eat the whole cost if it happens". Now if insurance companies had some super-accurate data and models on you (a futuristic perspective here), they could just infer if X will happen to you and take some free money if it won't and refuse insurance/make you pay everything plus premium otherwise. One similar case nowadays is health and old age pension, though these tend to be very regulated or state-run (i.e., tax funded). The point is that the scope of things where this applies can become very wide in the future. Someone in the linked comment tree said "dispersion of the cost in wider population" is "not insurance, but a social solidarity scheme" and you may agree. But it's possible that we are currently running some "social solidarity schemes" only because of the imperfect information.
  - wutbrodo 7y ago
    It would require more than imperfect information, but imperfect information _about_ the imperfect information. Ie, if people knew the insurance company had perfect information, they wouldn't buy the policy. At that point, the insurance company could just directly sell the information and still create surplus by reducing uncertainty. This meta-imperfect information wouldn't last very long: People are generally skeptical of things like insurance as a baseline anyway, since they're unintuitive ("why am I paying for nothing every month?"). In this hypothetical world of perfect insurers, people would just sum up the anecdota of X insurance literally never paying out (and everyone rejected having X happen), and it wouldn't take very long for conventional wisdom to switch to "insurance is never worth buying".
clairity 7y ago
you're grossly underestimating the absolute ingenuity of 7-10 billion people jostling against each other for status. even if people were taxed such that net income was exactly the same for everyone, higher earners would still try to claim superior status from it. and that theoretical limit of perfect, simultaneous, accurate information is akin to the heat death of the universe, and really not worth practical (as opposed to theoretical) discussion. for all practical purposes, at the very least there are rate limits related to the propogation of and reaction to perfect info, not to mention all the ways people will actively obstruct it.
- stereolambda 7y ago
  I dunno, the status appeal of money is very much what you can buy with it. I can technically imagine stores both online (good old tracking) and brick-and-mortar (looking at clothing and subtle body cues: inferring health, psychical well-being etc. - if not matching to external data with facial recognition) being able to quote, with machine learning, higher prices for you at check-out if your financial situation is better. Now of course wealthy people spend less on such "trivial" consumption in comparison to capital investments, but for middle class this could be significant. I mean a kind of thing like this.
  - clairity 7y ago
    the rich person would just hire poor people to buy their trivial consumption items for them, both to get better prices and to put downward pressure on the practice. that's not even a particularly ingenious solution, but that's the kind of thing i mean by underestimating our collective ingenuity. the transactional value of money is not inconsequential, but people absolutely do all kinds of (silly) things for status (see social media), moreso than just for transactional value.

wutbrodo 7y ago

> there was a discussion about how in the future it makes no sense to buy insurance if someone is willing to sell it to you, you are always better off saving and you won't get insured if you actually need it[0]. This is a misunderstanding of how insurance works. It makes sense to self-insured in expected $ value, but not necessarily in expected utility. There's nonzero value to lowering volatility[1], and the insurance company creates surplus by coordinating the pooling of risk across many people. For an easy example of the difference, for an easy example of the difference, I wouldn't take a 50/50 coin flip that would either cost me $1 million dollars or gain me $1.1 million; even though the expected value is $50k, because I couldn't afford such a high chance of financial ruin. I would however, sell the chance at this bet to an insurance company for $40k, and if they were large enough to absorb the loss (or bought many such policies), they would take it and gain $10k minus the (lower) value they place on the risk. (All numbers are made up, obviously). [1] this is a simple enough concept that you can derive it graphically from a utility/$ graph, and requires only the assumption of diminishing marginal utility of $

stereolambda 7y ago
Yeah, I wanted to give an example and disclaim redoing that specific discussion if possible. But just to provide a different view, what you're saying is fine theoretically, but in practice often people see the insurance situation as "some bad costly thing X can happen to me, I want to pay reasonable money to not have to eat the whole cost if it happens". Now if insurance companies had some super-accurate data and models on you (a futuristic perspective here), they could just infer if X will happen to you and take some free money if it won't and refuse insurance/make you pay everything plus premium otherwise. One similar case nowadays is health and old age pension, though these tend to be very regulated or state-run (i.e., tax funded). The point is that the scope of things where this applies can become very wide in the future. Someone in the linked comment tree said "dispersion of the cost in wider population" is "not insurance, but a social solidarity scheme" and you may agree. But it's possible that we are currently running some "social solidarity schemes" only because of the imperfect information.
- wutbrodo 7y ago
  It would require more than imperfect information, but imperfect information _about_ the imperfect information. Ie, if people knew the insurance company had perfect information, they wouldn't buy the policy. At that point, the insurance company could just directly sell the information and still create surplus by reducing uncertainty. This meta-imperfect information wouldn't last very long: People are generally skeptical of things like insurance as a baseline anyway, since they're unintuitive ("why am I paying for nothing every month?"). In this hypothetical world of perfect insurers, people would just sum up the anecdota of X insurance literally never paying out (and everyone rejected having X happen), and it wouldn't take very long for conventional wisdom to switch to "insurance is never worth buying".

clairity 7y ago

you're grossly underestimating the absolute ingenuity of 7-10 billion people jostling against each other for status. even if people were taxed such that net income was exactly the same for everyone, higher earners would still try to claim superior status from it. and that theoretical limit of perfect, simultaneous, accurate information is akin to the heat death of the universe, and really not worth practical (as opposed to theoretical) discussion. for all practical purposes, at the very least there are rate limits related to the propogation of and reaction to perfect info, not to mention all the ways people will actively obstruct it.

stereolambda 7y ago
I dunno, the status appeal of money is very much what you can buy with it. I can technically imagine stores both online (good old tracking) and brick-and-mortar (looking at clothing and subtle body cues: inferring health, psychical well-being etc. - if not matching to external data with facial recognition) being able to quote, with machine learning, higher prices for you at check-out if your financial situation is better. Now of course wealthy people spend less on such "trivial" consumption in comparison to capital investments, but for middle class this could be significant. I mean a kind of thing like this.
- clairity 7y ago
  the rich person would just hire poor people to buy their trivial consumption items for them, both to get better prices and to put downward pressure on the practice. that's not even a particularly ingenious solution, but that's the kind of thing i mean by underestimating our collective ingenuity. the transactional value of money is not inconsequential, but people absolutely do all kinds of (silly) things for status (see social media), moreso than just for transactional value.