Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Why is it so hard to be rational?

newyorker.com

Post reply on HN
Print view View original

221–230 of 309 posts

Re: Why is it so hard to be rational?

#221

Post by clairity »

mcguire wrote: post #178
nostromo wrote: post #79

Earlier quoted context omitted.

The sunk cost fallacy is particularly important to learn about and teach your children about. I see it everywhere, from my own decision making process to international politics. Just this morning I was thinking about it as I read the news about the US leaving Afghanistan, and last week talking with a friend who is staying at a bad job.

Here's a question for you: what is the difference between the sunk cost fallacy and persistence? And here's the answer: Persistence is good when it is successful. If the activity us unsuccessful, it's an example of the irrational sunk cost fallacy. (Making decisions without knowledge of future events is quite hard.) And the important lesson: If you bail at the first sign of adversity, no one can ever accuse you of be…

that's not irrationality, that's decision-making under uncertainty, which is the norm, not the exception. probabilities are dynamic, information is imperfect, and so decision-making must incorporate that uncertainty.

the sunk cost fallacy is simply considering existing loss when deciding on continued investment (in time, money and other resources), when you should only consider future cost for future benefit. it's thinking erroneously that existing loss is not already locked in, that it's salvageable somehow. but no, it's already lost.

in a project with continuously updating probabilities of success, and under imperfect information, the go-or-no-go decision should only be based on the likelihood of future gains exceeding future losses, not future+existing losses.

in this framework, persistence would be having credible evidence (e.g., non-public information), not just belief, of the likelihood of future net gain relative to opportunity cost. it'd be irrational to be persistent simply on belief rather than credible information and probability estimation.

Re: Why is it so hard to be rational?

#222

Post by carry_bit »

throwaway9690 wrote: post #30

I think part of the problem is that most people are conditioned into many beliefs from a young age I know a guy who hates foo (using a place holder). In fact he's downright foophobic. He is pretty convinced he has a natural unbiased hate of foo and is being rational when he expresses it. To me as an outsider it is pretty obvious that his hate of foo is the result of cultural conditioning. To him it is perfectly ratio…

It could be a case of implicit vs explicit knowledge. In the context of evolved culture beliefs, the foophobia may serve some real purpose, even if most/all of the enculturated individuals can't explicitly state what the real purpose is.

It could be that, like dietary restrictions to reduce the spread of disease, the foophobia is no longer needed, but keep Chesterton's fence in mind before you say it's unneeded.

Re: Why is it so hard to be rational?

#223

Post by amelius »

strulovich wrote: post #124
btilly wrote: post #14

I maintain that it isn't just hard, it is computationally impossible. We should all know that given a belief about the world, and evidence, Bayes' Theorem describes how to update our beliefs. But what if we have a network of interrelated beliefs? That's called a Bayesian net, and it turns out that Bayes' Theorem also prescribes a unique answer. However, unfortunately, it turns out that working out that answer is NP-h…

NP hard problems get abused for justifying things they cannot. An NP hard problem, even if it cannot be approximated does not mean the average input cannot be solved efficiently. Examples: - An NP hard problem is not sufficient for building crypto. - Type solving for many programming languages is EXP TIME complete, yet those languages prosper and compile just fine. Beware the idea of taking a mathematical concept and…

Ok, so what is the class of problems that is hard for any input?

Re: Why is it so hard to be rational?

#224

Post by paganel »

The rational powers that be were saying out loud 3 days ago that in an optimistic scenario the Afghan government would hang on for another 90 days, in a pessimistic scenario only for 30 days. As we all know it collapsed completely in just 2-3 days.

Early on during the pandemic (the first half of February 2020) the people writing on Twitter about covid in China were being labeled as conspiracy nuts, with some of them outright having their accounts suspended by Twitter. Covid/coronavirus was (I think purposefully) kept out of the trending charts on Twitter, the Oscars were seen as much more important.

And these are only two recent examples that came to my mind where the "rational" parts of our society (the experts and the media) failed completely, as such it's only rational not to trust these pseudo-rational entities anymore. Imo I think in a way the post-modernists were right, (almost) everything is negotiable or a social construct, there's no true or false, apart from death, I would say.

Re: Why is it so hard to be rational?

#226

Post by nicoburns »

Tenoke wrote: post #116
btilly wrote: post #14

I maintain that it isn't just hard, it is computationally impossible. We should all know that given a belief about the world, and evidence, Bayes' Theorem describes how to update our beliefs. But what if we have a network of interrelated beliefs? That's called a Bayesian net, and it turns out that Bayes' Theorem also prescribes a unique answer. However, unfortunately, it turns out that working out that answer is NP-h…

Nobody is disputing this. You can, however, clearly be more or less 'rational', adopt better or worse heuristics, etc. which is what you attempt to get from reading LessWrong or Kahneman.

But what constitutes a better heuristic is context dependent. In particular, if I must make a decision in a time-constrained manner then any heuristic that blows the time budget is going to be worse even if it would be better given more time. And one can't really know in advance how much time spent thinking is optimal. So one has to pick a strategy. The fact that humans have evolved to use a variety of strategies fast and slow (depending on the human) suggests that there is no single optimal strategy.

See also Gigerenzer's Ecological Rationality.

Re: Why is it so hard to be rational?

#227

Post by AndrewKemendo »

btilly wrote: post #14

I maintain that it isn't just hard, it is computationally impossible. We should all know that given a belief about the world, and evidence, Bayes' Theorem describes how to update our beliefs. But what if we have a network of interrelated beliefs? That's called a Bayesian net, and it turns out that Bayes' Theorem also prescribes a unique answer. However, unfortunately, it turns out that working out that answer is NP-h…

This nails it.

I'd go further to say that there are real world issues that compound the variables. Namely that individual actions increasingly have global consequences eg. individual purchasing behaviors have externalities that the market is not pricing in and thus fall to the consumer to have to calculate.

Further, given that these global issues these kinds of calculations are game theoretic by their nature, making it even more complicated.

Re: Why is it so hard to be rational?

#228

Post by AndrewKemendo »

lalaithion wrote: post #26
btilly wrote: post #14

I maintain that it isn't just hard, it is computationally impossible. We should all know that given a belief about the world, and evidence, Bayes' Theorem describes how to update our beliefs. But what if we have a network of interrelated beliefs? That's called a Bayesian net, and it turns out that Bayes' Theorem also prescribes a unique answer. However, unfortunately, it turns out that working out that answer is NP-h…

That's why it's called "less wrong". The goal isn't to be perfect, the goal is to do better. To be less wrong.

I think this misses the point. Even being "less" wrong requires an amount of work that even the best/smartest etc... cannot consistently apply.

I do believe this is zero-sum in that improving on one set of decisions means no applying the same rigor to others.

This is often seen in the form of very smart people also believing conspiracy theories or throwing their hands up around other massive issues. As an example, the "Rationalist crowd" has de-emphasized work on climate change mitigation in favor of more abstract work on AI safety.

Re: Why is it so hard to be rational?

#229

Post by kazinator »

UnFleshedOne wrote: post #204
kazinator wrote: post #150

Earlier quoted context omitted.

> I maintain that it isn't just hard, it is computationally impossible. I further maintain that it's definitionally impossible. Before we find it computationally impossible, we will find that we can't write the a complete, detailed requirements specification defining what rational is. (Of course, we recognize egregious irrationality when we see it; that's not what I mean; you can't just define rationality as the oppo…

You can't say values are irrational -- they just are. If you really like paperclips, no amount of logic can tell you otherwise. What logic can tell you (and other people could), is that your values conflict with each other and you have to balance one against another. Turning whole universe into paperclips is counterproductive if you also value pins. If you literally have no value the other person is basing their argu…

> You can't say values are irrational

People basically do say that, though.

(Values can be contradictory/inconsistent. E.g. you say you value self-preservation, but you also enjoy whacking your head with a hammer. That would be a kind of irrational. That's not what I'm referring to though.)

Re: Why is it so hard to be rational?

#230

Post by wyager »

jitter_ wrote: post #214
wyager wrote: post #163

Earlier quoted context omitted.

“Logical fallacies” are mostly Boolean/Aristotelian and identifying them is completely useless and/or counterproductive in 99% of real world scenarios. Most of your reasoning should be Bayesian, not Boolean, and under Bayesian reasoning a lot of “fallacies” like sunk cost, slippery slope, etc. are actually powerful heuristics for EV optimization.

> under Bayesian reasoning a lot of “fallacies” like sunk cost, slippery slope, etc. are actually powerful heuristics for EV optimization. Can you elaborate on that? This really piqued my interest. I feel like logic is easy to apply retrospectively (especially so for spotting fallacies), but trying to catch myself in a fallacy in the present feels like excessive second quessing and overanalyzing. The sort that preven…

Sure. Fallacies, as usually stated, tell you when something that feels like a logical entailment isn’t actually a logical entailment.

Intuitively, people find “bob is an idiot so he’s wrong” a reasonable statement.

Technically, the implication does not hold (stupid people can be correct) and this is an ad hominem fallacy.

However, if we analyze this statement from a Bayesian standpoint (which we should), the rules of entailment are different and actually bob being stupid is evidence that he’s wrong. So maybe this is actually a pretty reasonable thing to say! Certainly reasonable people should use speakers’ intelligence when deciding how much to trust speakers’ claims, even though this is narrowly “fallacious” in an Aristotelian sense.

I’m not aware of any reading on this topic. It seems under-explored in my circles. However I know some other people have been having similar thoughts recently.

Post reply on HN