The Infinity Fallacy
You’ve probably heard of short-termism, a mode of thinking where one lives like each day is their last, to predictable consequences. I am looking to label its extreme opposite—accumulation which fails to have an adequate (or any) spending phase.
Some examples:
Strategy/card game players often play “too greedy”. New slay the spire players may avoid wraith form or dead branch because these effects come with infinitely ramping downsides.
The EA pledge pushes us to give 10% even early in our careers, presumably in part to undermine our tendency to accumulate, accumulate, then die of old age.
One reason to oppose resource extraction is that it is irreversible. Once you take the oil and burn it, we permanently have less oil, and what if we need it in 10000 years?
Taxation—in a framework where low taxes bolster growth and pay for themselves with a bigger tax base, the policymaker still has to ask “when?”. If I am only interested in maximizing tax take, halving the tax rate to double the tax base is great if it happens in a year, less so if it takes a century.
One is committing the Infinity Fallacy if, given a choice between a short- and a long-term action, they always pick the long-term without considering if the payoff will actually materialize within their lifetime/planning horizon.
I use the word “infinity” here because I think that the most egregious errors of this type happen when we treat the opportunity cost or compound gain of an action as something that will follow us forever. “If you save $1 today, next year it will give you $0.07, which will earn you even more and so forth until eventually it turns into untold riches and infinite utilons.”
When committing the Infinity Fallacy, the long-term choice can always mug you by promising to provide infinite utils. Instead, we should calculate with a finite term—“I expect this $1 to compound over X years and ignore growth afterward because I will be dead/scammed/AI will make money useless.”
Prior to this post i did not hear about short-termism. So i took a look about it and i see that is a rather vague idea that can be summarized as “Prioritizing desicions that have an immediacy”. Interesting thing is that there are no set rules or definitions about the term. There is just a vague term given that is “immediacy”. I think it is safe to assume that this is considered as immediate relative to a human’s life time or maybe an individual’s own perspective of time. Sure. However to oppose you have proposed a new idea that is as vague as the original one. The infinity fallacy. The more i write the more arguments come to my mind.
For starters short-termism does not signal happening of without duration does it? As far as i understand it rather signals something happening in a rather short time related to human perception or life time. So i would assume short-termism does have at least some understanding of time duration. If that is true then the opposite of this cannot be infinity. It would rather need to be something like… long-termism? Something that takes relatably longer.
If my first thought is not correct and short termism does really mean non duration or as less duration as possible then maybe infinity as an opposite option does seem plausable.
Your argument is that as far as i understand people who are using the infinity fallacy are not recognizing where short-termism can be beneficial and are always choosing long term or infitine term over short-term thinking it is automatically more beneficial than short term.
However isn’t both long-termism and short-termism are a bit tunnel visioned approaches? I suspect there is a bigger and deeper fallacy or a problem behind both of these. It would be more like “Choosing what my set habit system over any other system or choice opportunity will always result in better ways because my system is the best.” wouldn’t it?
I think both a person who prefers short-termism and long-termism (who does the infinity fallacy) would be similar people in thought. They are both coming from the same flawed thought.
Also i do not think that both approaches can really be divided from each other since they are kind of part of each other aren’t they?
For example if you want to improve over a certain topic in the long term, you start by doing small projects that can be done, critiqued, revised and done again in the short term don’t you? In this way both concepts work for your benefit not the other way around.
So. Yes your arguments are sound that tasks that are more suited for short term approaches suffer from long-termism or infinity fallacy. However vise versa happens where long term approaches suffer from short-termism.
So really i think there would be a bigger deeper fallacy or a problem or a way of thinking,functioning that should be addressed before these in my opinion.
Finance has a way of thinking about these tradeoffs. It’s called the “discount rate”. For example if you want to value an apartment which rents for an amortized $100 per month, infinite fallacy would be to envision taking profits forever—how can you possibly put a finite value on that? Finance guys instead discount future cashflows by a variable yearly rate, determined arbitrarily based on perceived risk and comparable investment options. Then they also stop the calculation at 30 years into the future. The 30-year term afaik is arbitrarily determined.
I am not sure exactly which factors these two measures are meant to control for. I’ve seen them attributed to risk (no company lives forever), to inflation, to opportunity cost vs other investments, to time preference. But in practice they allow the finance industry to put finite, comparable valuations on cashflows that have no clear end date.
Lets imagine two people who make the opposite error. Lets say both of them want middle class lifestyles, marriage 2.5 kids etc. Short-term Sammy lives for today, he spends all his money on consumption and cuts his work hours to have more free time. Long-term Lenny scrounges away every penny, never goes out and has no chances of meeting women. Common advise to Sammy would be that he should think about his future, and common advise to Lenny would be that he should loosen up because he won’t be able to spend his money or start a family when he’s old.
If you want to rigorously check Sammy and Lenny’s work, I think you’d want to calculate with discount rates. Show that they are both using the wrong discount rate. The problem is that this is too much detail for everyday life. Shorthand concepts like “you are failing the marshmallow test” come to the rescue here. Everyone knows about the marshmallow test, but we lack a meme for the other extreme, and that’s what I’m trying to find.
I see your problem is not that there is one way of doing things but rather there is a good way to put your words across when one extreme happens but we do not have the same for the other one. Therefore it is not as easy to correct people when they do use such fallacies. However two very known concepts come into my head when i read this problem. Sunk cost fallacy and the concept of diminishing returns. Aren’t they are the “meme” you are searching for that can be used to correct someone when they do the infinity fallacy? What do they lack?
Another argument of yours suggest that not using shorthands is not practical in everyday life. So we cannot apply actual analysis of problems. It is more efficient to instead use shorthands that are seemingly working most of the time. However i do agree that sometimes it is not practical to apply actual analysis on everyday life but i would say that most of the time it is practical. In the heat of the moment yes it is reasonable to use shorthands. However after one deals with that problem they should start analyzing the situation and modify their strategies accordingly. I do think that this is practical. Therefore i would still say that abiding by any shorthand all of the time is not as practical as it seems. It would most likely be useful in just making the person have less mental load is all.
If that is still not practical, a new definition for basically finding “do what has to be done according to the situation” can be invented and can be used as a shorthand if it can be developed into a system. Wouldn’t that basically solve both the problems ?
I don’t see the connection to sunk cost or diminishing returns. You can commit the infinity fallacy even if the predicted returns or costs are completely correct. Consider the frugal old loner who dies a millionaire, or the slay the spire examples from the initial post. There is a category of error where even if you predict the long-term trend, it doesn’t matter because you won’t care by the time it happens.
I guess this means that ‘discount rates’ are not as good of a match as I thought?
Prior to this discussion i did not know what discount rate was. So i went and learned about it. Correct me if i am wrong. It is something like this:
Let’s say that i want to have 20.000 dollars exactly 1 year from now. I will invest into some fund that has an interest rate of 5%. How much money should i invest today to have 20.000 dollars exactly 1 year later?
The formula is this(source:https://openstax.org/books/principles-finance/pages/7-4-applications-of-tvm-in-finance?query=discount%20rate&target=%7B%22index%22%3A0%2C%22type%22%3A%22search%22%7D#para-00003):
PV = FV × 1/1(1+𝑟)^𝑛
PV = 20.000* 1/1(1+5%)^1
PV = 20.0000 * 1/1(1.05)^1
PV = 20.000 * 0,95238095
PV = 19,047619
I have to put 19,047619 dollars today to have 20.000 dollars exactly 1 year later.
If i understood it correctly your question to people who does the infinity fallacy is “You assume that you can put any number in the place of 𝑛 even infinity itself. That is illogical.”
Yes i agree. I still think that the law of diminishing returns is fitting for this problem isn’t it?
If you solely take the factors as money and numbers yeah sure maybe in some situations there might not be a diminshing return. Let us just assume that there is not diminishing return as far as money and resources goes. Then i would say that you are excluding the human from the equation. Human is the one who goes thru all of these to achieve a certain thing. If it is not use for the human even if everything else works out correctly. It still does not matter just like you said.
If you want to keep it mathematical think it like this:
Human happiness = H
Effort = E
Money = M
1H 1E 1M
10H 5E 20M
5H 40E 15M
I did not went thru a calculation to get these numbers but they are just there to explain a point. As you can see at first our human gets happy to achieve 10 happiness thru 20 money with only spending 5 effort.
But then he puts more effort. He puts 40 effort and he successfully gets 15 Money. However this time gaining more money simply made him less happy because the effort is too much now.
Isn’t this a diminishing return? You would argue that yes there is a diminishing return in the human happiness but because there IS a diminishing return in the system anyway. What happens when the system does not hit the diminishing return point ever?
I would argue that this changes nothing. Let us just assume that our human somehow gains infinite money with very low effort. The system itself never hits diminishing returns. However our human probably wanted to gain that money for something. The moment that money is way more than enough and that even the effort is so low that the system never hits diminishing returns, it will still hit diminishing returns once the effort hits high enough that it disturbs the targeted confort that the money is being used to acquire. We can even go further than that and can simply say that it should be enough to say that is a diminishing return point the moment the human itself finds it “boring” or “not to matter”.
So for that reason as a “meme” to explain that someone is doing infinity fallacy. Diminishing returns does apply to this situation and it is a well known phrase that can be used in a conversation.
Is it still not useful for this situation?