As with most cognitive biases, there's an inverse to this, where we ignore low-probability high-impact scenarios. E.g. people drive drunk or without a seatbelt, because it'll *probably* be fine. And they repeatedly have that assumption confirmed--until one day it isn't.
I had one friend who would leave his bike chained partially blocking a fire exit, because "what are the odds the fire inspector will come today?" But the fire inspector comes once a year, and if your bike is chained there 99% of the time, odds are you're going to get a fine. He couldn't see the logic. He got fined.
After reading the first half of your comment, I was afraid the second half was going to end with something like "Then there was a fire and 3 people died because the exit was blocked."
Getting fined doesn't sound so bad -- if it was like $100, your friend could just be treating it as a $0.30/day fee for convenient parking. But you both seemed to ignore the really high-impact potential outcome. So I guess that proves your point.
>Getting fined doesn't sound so bad -- if it was like $100, your friend could just be treating it as a $0.30/day fee for convenient parking.
I had a professor in college that said she basically treated speeding tickets like that. She had to commute on the weekends from central IL to the Chicago and just accepted that she'd probably get 1-2 speeding tickets per year.
Traffic in general is riddled with this. People don't understand the risks they're taking during their everyday driving and get offended when you comment on it.
Typical folks cutting in front of me while I am barely at safe breaking distance from the car in front of me, on speeds > 100kmh. This is of course always in at least semi-dense traffic, and them immediately obscuring view further means I have less than second to react to any stronger breaking or I slam into them.
I honk them, then they often get aggressive that I dared to react to their perfectly cool maneuver that gave them those precious extra 5 seconds. Bloody a-holes. Had few almost-collisions even this year due to too aggressive drivers riding too close, some were literally car in front of us or next one behind. Keep your distance, I can't emphasize this enough.
Have you considered buying a dashcam? Then in case you do actually hit them because you weren't able to break fast enough, it's very likely that insurance will end up siding with you, as opposed to the usual outcome where you're the rear-ender.
My hand gesture for "Hey did you hear about the inverse Pascal Scam? It suggests that low-probability high-impact risks are easy to ignore, and I think that's what you're doing right now and that's not going to be good for your health, or mine for that matter, so maybe think about that a bit more in the future" is to raise my middle finger. Unfortunately it inevitably makes the situation worse somehow.
I have switched to smiling at things people do well in traffic and trying to give positive reinforcement when I see people who care about others. Some even do it in ways that might seem stupid, e.g a guy in the wrong side of the street was actually just being very considerate.
Nice! I tried thumbs up (ie sarcasm) but that's snarky too, and somehow never realized that you could actually do the same thing non-sarcastically. Srsly wow :-) Gonna try, thanks
In a similar spirit, I knew someone who claimed to not pay for parking permits at our university, and just parked wherever he liked. The parking permits were $100+ per month and the parking fines were ~$300 per citation, so if he gets caught less than once per quarter, he would come out ahead.
He tells me later that it didn't quite work out in terms of saving money, but because he sometimes parked in spots that he could not get permits for, it actually saved time.
Up until recently, fares for the LRT system in my city were enforced by a random check by transit police, typically by having an officer board trains and check riders' tickets at random times during random days and handing out fines to fare evaders who they caught.
Between around mid-2006 and the end of 2008 I rode the train to work downtown every day. The trains were so crowded during rush hour that it was impossible for Transit police to board trains to check fares, and even outside rush hour, fare checks were very occasional. A monthly pass at the time was around $75 and a fine for fare evasion was around $200 (the first violation was less than $200, and I think it increased until a cap of something like $250 for repeat offenders). I'd worked it out that if I was caught without paying a fare less than once every three months, it would be cheaper to just pay the fine if/when I got caught rather than buy a pass. So I didn't buy a pass and decided to see how long it would take to actually get caught.
The answer was about 18 months. Got a $170 fine. Which I then forgot about and never actually paid. The statute of limitations on that fine has long since expired.
You're lucky with that system of fine capping. I had the same mentality with the very expensive trains here, probably saved upwards of 2-3000£ until I got caught. Problem is they actually prosecute you in the courts for repeat offenders so I can no longer risk it.
"The odds of X happening are so low that what's the point?", to which I respond "It only needs to happen once for me to be dead, so, the stakes are way too high for me to risk the odds".
>Suppose someone comes to me and says, "Give me five dollars, or I'll use my magic powers from outside the Matrix to run a Turing machine that simulates and kills 3^^^^3 people.
It was introduced to point out a place where the currently-best "decision theory" produces a seemingly-wrong recommended action in the hope that someone might come up with a better decision theory or alternatively to explain why the seemingly-wrong recommendation is actually right. A "decision theory" is a recipe for acting in the face of uncertainty that is precisely stated enough that it might eventually serve as the basis of a computer program.
This also touches mind theory. In particular it relies on the belief that running a Turing machine that kills 3^^^^3 people is as bad as breeding and killing the same number of people in the physical world we live in.
1% chance of humanity going extinct is exactly the pascals mugging doomerism. If someone thinks there’s like >10% chance they’re probably actively working on like AI safety or something else to stop it.
I argue that it's not that hilarious because the thinking is very tightly related. The very contemplations that lead to AI doomerism lead to pascals mugging.
One of my main gripes with AI doomerism is that it is downstream of being pascal's mugged into being a doomer
Rationalists and Effective Altruism people fall for this stuff _constantly_. Roko's Basilisk being the canonical example of it.
They assign infinite negative or postive values to outcomes and then it doesn't mighter what the likelihood or how much they uncertainty they have everywhere else, they insist that they need to do everything possible to cause or prevent whatever that outcome is.
Aside from other problems with it, there are a vast number of highly improbable and near-infinitely bad or good outcomes that might possibly occur which would require completey different actions if you're concerned about them.
This is exactly how I feel when Effective Altruism starts talking about the wellbeing of trillions of humans living in the far distant future that we should be devoting ourselves to now.
One of the things you need to do for this situation, as well as the stuff in the blog post, is apply what is commonly called "the time value of money" to the concept. However the concept extends beyond money into any attempt to modify the future, whether or not it involves money. Money just happens to function as a good, quantifiable example in its "score keeping" role here. Your ability to model the future and take actions based on it exponentially decays, and really quite rapidly.
Or to put it another way, everything fuzzes out into noise for me much sooner than humanity will have trillions of new members. There's no way for me to predict whatsoever what effect any action I take today will have a thousand years from now. Even in extreme cases, like, I push a magic button that instantly copies whatever you, the reader, believe is the optimal distribution of ideological beliefs out into the world (ignoring for the moment the possibility that your ideology might consider that unethical, this is just a thought experiment anyhow so no need to go that meta), you really don't know what that would do 1000 years from now, what the seventeenth-order effects of such a thing would be. I'm not even saying that it might not be as good as you think or something; I'm saying you just have no idea what it would be at all. So there's no way to hold people responsible for that, and no way to build plans based on it.
You have personified EA, but it isn't a person. Some EA people are into long-termism, but it is an error to pretend EA is a monolith that speaks with one voice.
I think the core idea is simply: since resources for helping the poor/sick is not unlimited, we should try to allocate those resources in the most effective way. Before EA charity evaluation came along, the only metric for most people was simply looking at the charity overhead via Charity Navigator. But that isn't a great metric. A charity with only a 1% overhead with a mission to make balloon animals for children dying in a famine will score well on Charity Navigator but does nothing to help the problem.
To be honest I haven't looked deeply into long-termism, but from what I've heard (eg, hearing Will MacAskill on a few podcasts) it seems to ignore a few things. Just like a bird in the hand is worth two in the bush, long-termers have no good way to estimate the likelihood of future events, and that discounting needs to increase greatly the further out one looks. At best many of these estimates are like the Drake Equation -- better than nothing, but with multiple orders of magnitude error bars.
There are other second-order reasons which don't seem to factor in, or at least haven't come across in the few hours of listening to long-termers talk about the issue. One is that by working to make a better world now, it effects the trajectory of future events much more directly than the low-probability guesswork they think my have an impact in the distant future.
That was always a convenient excuse to justify amassing lots of money, if necessary by theft (see SBF). They have no intention of actually doing good with it.
Let me reframe it. Among these trillions of people, there will be many who are 99% similar to you. Wouldn't you want that version of yourself to live a great life?
Sure! It's just not actually reasonably possible to predict what will lead to that outcome, whereas I can sorta (badly) predict what will make me happy tomorrow, my wife happy next week, and what will best improve conditions for my current fellow citizens of the world.
You can certainly rationalize anything, but I fail to see what help that is to us.
"The great subverter of Pyrrhonism [radical skepticism] is action, and employment, and the occupations of common life. [...] I dine, I play a game of back-gammon, I converse, and am merry with my friends; and when after three or four hour's amusement, I wou'd return to these speculations, they appear so cold, and strain'd, and ridiculous, that I cannot find in my heart to enter into them any farther."
Regardless, it is an extremely uncertain proposition that we can do anything in the present that would have a reliably positive impact on their lives in the far future. It's hard enough to figure something out for the billions of people who actually exist right now.
They claim isn't that there will be trillions of people alive at the same time; they are integrating over the course of tens and hundreds of thousands of years.
Although we are at a peak population of a bit over 8B people at the moment, it is estimated that more than 100B people we would classify as humans have ever lived. The population long ago was much smaller than 1B, but thousands of generations have lived and died.
What is something specific that an individual should have done or should not have done, say, 2000 years ago, that would have made a positive impact on your life?
Certainly there are individuals who have had that kind of impact, but I don't think one could have foreseen the consequences at the time, e.g., Jesus and Mohamed.
Likewise, the guy who invented the Segway probably thought he was going to revolutionize personal transportation, but that didn't pan out (yet, at least).
1. Suppose you have a chart of the total past and future history of human population.
2. Cover up the chart so only the data from the past to present day is visible.
3. Note that most humans in that subset exist near or at the present. You are one of these people today, it should make sense for you to be born in one of the densest parts of the graph.
4. Now uncover the graph. If there are trillions of humans in the future, it seems almost impossibly unlikely that you would born in a part of the graph with "so few" humans as today, and not in the far future.
Therefore, you must conclude that the actual graph rapidly drops to zero in the near future. QED.
This "doomsday argument" is a pretty shit one, but not worse than others I've seen arguing the opposite.
As a kernel developer, I have always had to worry about what happens in the 0.000001% of cases. You're either correct all of the time or you likely have an exploitable security hole in your code that will turn into a CVE at some point in the future that will cost a lot of money to fix. One has to use strategies to reason through proof that your code does the right thing in all states, including the insertion of delays and other fault injection techniques to trigger the cases that aren't normally hit.
It's also kind of like saying that an LLM only pushes 0.01% of people off the deep end in a psychotic episode, so it doesn't really matter. That doesn't actually get rid of the liability.
All that said, I will always refuse to write life safety code. I am not a good enough developer to do so.
A lower-key variant of this frequently comes into play with consulting or other sales pitches. "You spend <big number> per year on this <necessary business expense>. Our service will easily shave 2% off this, making the cost of our service completely negligible and this purchase an obviously good decision."
For reasons explained in the article, we are bad at estimating small probabilities.
Similarly, we are bad at estimating small proportions ("easily shave 2%"). What is being claimed in the parentheses here is that there's a probability distribution of "how much costs are shaved" and that we can estimate where the bulk of its support is.
But we're not really good at making such estimates. Maybe there is some probability mass around 2%, but the bulk is around 0.5%. It seems like that's a small difference (just 1.5%!) but it's a factor of 4 in terms of savings.
So now we have a large number (annual spend), multiplied by a very uncertain number (cost shave, with poor experimental support), leading to a very uncertain outcome in terms of savings.
And it can be that, in reality, the costs of changing service turn out to overwhelm this outcome.
When modern advertising is a spectrum of “lies, damn lies, and statistics,” I don’t blame folks for crying foul and demanding a baseline level of truth in advertising. When folks trust but verify, this is seen as a change in the status quo by folks, and some of those folks who protest about it in those terms are trying to sell you something.
It's amusing to consider how much of a Rorschach test this article must be. But it's a great point, even if it arms us to abusively write off unwelcome ideas as scams. As the author points out, Pascal's reasoning is easily applied to an infinity of conceivable catastrophes - alien invasions, etc. That Pascal specifically applied his argument to the possibility of punishment by a biblical God was due to the psychological salience of that possibility in Pascal's culture - a truly balanced application of his fallacious reasoning would be completely paralyzing.
Pascal’s gamble wasn’t just about probability, it was about storytelling: the promise of nearly infinite payoff with minimal risk. That same allure is still at play today whenever people chase “moonshot” returns on crypto or quick-rich schemes.
It underscores a timeless lesson: no matter how much data or logic we have, we’re still wired to fall for well-crafted optimism and that means skepticism remains the best defense.
> Your comment would have been better if you'd chosen and example that did not create many tens of thousands of millionaires.
Lotteries have also produced lots of millionaires. Crypto could produce lots of winners just from wealth transfer even if it was a zero sum or net negative game in terms of wealth creation.
Just bitcoin uses more energy than Finland, the energy sector in Finland makes lots of people rich. Of course people will earn money when people spend that amount of resources on something.
As with most cognitive biases, there's an inverse to this, where we ignore low-probability high-impact scenarios. E.g. people drive drunk or without a seatbelt, because it'll *probably* be fine. And they repeatedly have that assumption confirmed--until one day it isn't.
I had one friend who would leave his bike chained partially blocking a fire exit, because "what are the odds the fire inspector will come today?" But the fire inspector comes once a year, and if your bike is chained there 99% of the time, odds are you're going to get a fine. He couldn't see the logic. He got fined.
After reading the first half of your comment, I was afraid the second half was going to end with something like "Then there was a fire and 3 people died because the exit was blocked."
Getting fined doesn't sound so bad -- if it was like $100, your friend could just be treating it as a $0.30/day fee for convenient parking. But you both seemed to ignore the really high-impact potential outcome. So I guess that proves your point.
>Getting fined doesn't sound so bad -- if it was like $100, your friend could just be treating it as a $0.30/day fee for convenient parking.
I had a professor in college that said she basically treated speeding tickets like that. She had to commute on the weekends from central IL to the Chicago and just accepted that she'd probably get 1-2 speeding tickets per year.
Traffic in general is riddled with this. People don't understand the risks they're taking during their everyday driving and get offended when you comment on it.
Typical folks cutting in front of me while I am barely at safe breaking distance from the car in front of me, on speeds > 100kmh. This is of course always in at least semi-dense traffic, and them immediately obscuring view further means I have less than second to react to any stronger breaking or I slam into them.
I honk them, then they often get aggressive that I dared to react to their perfectly cool maneuver that gave them those precious extra 5 seconds. Bloody a-holes. Had few almost-collisions even this year due to too aggressive drivers riding too close, some were literally car in front of us or next one behind. Keep your distance, I can't emphasize this enough.
Have you considered buying a dashcam? Then in case you do actually hit them because you weren't able to break fast enough, it's very likely that insurance will end up siding with you, as opposed to the usual outcome where you're the rear-ender.
No but that seems like a good idea, thank you.
My hand gesture for "Hey did you hear about the inverse Pascal Scam? It suggests that low-probability high-impact risks are easy to ignore, and I think that's what you're doing right now and that's not going to be good for your health, or mine for that matter, so maybe think about that a bit more in the future" is to raise my middle finger. Unfortunately it inevitably makes the situation worse somehow.
I’ve switched to a thumbs down in traffic and can’t recommend it enough. Let’s then know how they should feel without escalating like a middle finger.
I have switched to smiling at things people do well in traffic and trying to give positive reinforcement when I see people who care about others. Some even do it in ways that might seem stupid, e.g a guy in the wrong side of the street was actually just being very considerate.
Nice! I tried thumbs up (ie sarcasm) but that's snarky too, and somehow never realized that you could actually do the same thing non-sarcastically. Srsly wow :-) Gonna try, thanks
In a similar spirit, I knew someone who claimed to not pay for parking permits at our university, and just parked wherever he liked. The parking permits were $100+ per month and the parking fines were ~$300 per citation, so if he gets caught less than once per quarter, he would come out ahead.
He tells me later that it didn't quite work out in terms of saving money, but because he sometimes parked in spots that he could not get permits for, it actually saved time.
Up until recently, fares for the LRT system in my city were enforced by a random check by transit police, typically by having an officer board trains and check riders' tickets at random times during random days and handing out fines to fare evaders who they caught.
Between around mid-2006 and the end of 2008 I rode the train to work downtown every day. The trains were so crowded during rush hour that it was impossible for Transit police to board trains to check fares, and even outside rush hour, fare checks were very occasional. A monthly pass at the time was around $75 and a fine for fare evasion was around $200 (the first violation was less than $200, and I think it increased until a cap of something like $250 for repeat offenders). I'd worked it out that if I was caught without paying a fare less than once every three months, it would be cheaper to just pay the fine if/when I got caught rather than buy a pass. So I didn't buy a pass and decided to see how long it would take to actually get caught.
The answer was about 18 months. Got a $170 fine. Which I then forgot about and never actually paid. The statute of limitations on that fine has long since expired.
You're lucky with that system of fine capping. I had the same mentality with the very expensive trains here, probably saved upwards of 2-3000£ until I got caught. Problem is they actually prosecute you in the courts for repeat offenders so I can no longer risk it.
"Odds vs. Stakes"
"The odds of X happening are so low that what's the point?", to which I respond "It only needs to happen once for me to be dead, so, the stakes are way too high for me to risk the odds".
> low-probability high-impact
People often equate “risk” with “likelihood”, when it would be more effective to view risk = impact * likelihood.
Even then, high impact very low likelihood is still often worth avoiding.
Maybe (impact * likelihood) - cost to mitigate
It has a wiki page under a slightly different name (but the concept is the same): https://en.wikipedia.org/wiki/Pascal%27s_mugging
A better explanation would be https://www.lesswrong.com/w/pascal-s-mugging, but all most readers will need to know is this next quote from that last page:
>Suppose someone comes to me and says, "Give me five dollars, or I'll use my magic powers from outside the Matrix to run a Turing machine that simulates and kills 3^^^^3 people.
It was introduced to point out a place where the currently-best "decision theory" produces a seemingly-wrong recommended action in the hope that someone might come up with a better decision theory or alternatively to explain why the seemingly-wrong recommendation is actually right. A "decision theory" is a recipe for acting in the face of uncertainty that is precisely stated enough that it might eventually serve as the basis of a computer program.
This also touches mind theory. In particular it relies on the belief that running a Turing machine that kills 3^^^^3 people is as bad as breeding and killing the same number of people in the physical world we live in.
Kinda hilarious that this got invented by the same people who later Pascal's-Mugged half the world with AI doomerism.
If someone thinks there's only 1% change of AI doom they're definitely not a doomer, much less 1 in 1,000.
1% chance of humanity going extinct is exactly the pascals mugging doomerism. If someone thinks there’s like >10% chance they’re probably actively working on like AI safety or something else to stop it.
I argue that it's not that hilarious because the thinking is very tightly related. The very contemplations that lead to AI doomerism lead to pascals mugging.
One of my main gripes with AI doomerism is that it is downstream of being pascal's mugged into being a doomer
Rationalists and Effective Altruism people fall for this stuff _constantly_. Roko's Basilisk being the canonical example of it.
They assign infinite negative or postive values to outcomes and then it doesn't mighter what the likelihood or how much they uncertainty they have everywhere else, they insist that they need to do everything possible to cause or prevent whatever that outcome is.
Aside from other problems with it, there are a vast number of highly improbable and near-infinitely bad or good outcomes that might possibly occur which would require completey different actions if you're concerned about them.
This is exactly how I feel when Effective Altruism starts talking about the wellbeing of trillions of humans living in the far distant future that we should be devoting ourselves to now.
https://www.effectivealtruism.org/articles/cause-profile-lon...
One of the things you need to do for this situation, as well as the stuff in the blog post, is apply what is commonly called "the time value of money" to the concept. However the concept extends beyond money into any attempt to modify the future, whether or not it involves money. Money just happens to function as a good, quantifiable example in its "score keeping" role here. Your ability to model the future and take actions based on it exponentially decays, and really quite rapidly.
Or to put it another way, everything fuzzes out into noise for me much sooner than humanity will have trillions of new members. There's no way for me to predict whatsoever what effect any action I take today will have a thousand years from now. Even in extreme cases, like, I push a magic button that instantly copies whatever you, the reader, believe is the optimal distribution of ideological beliefs out into the world (ignoring for the moment the possibility that your ideology might consider that unethical, this is just a thought experiment anyhow so no need to go that meta), you really don't know what that would do 1000 years from now, what the seventeenth-order effects of such a thing would be. I'm not even saying that it might not be as good as you think or something; I'm saying you just have no idea what it would be at all. So there's no way to hold people responsible for that, and no way to build plans based on it.
You have personified EA, but it isn't a person. Some EA people are into long-termism, but it is an error to pretend EA is a monolith that speaks with one voice.
I think the core idea is simply: since resources for helping the poor/sick is not unlimited, we should try to allocate those resources in the most effective way. Before EA charity evaluation came along, the only metric for most people was simply looking at the charity overhead via Charity Navigator. But that isn't a great metric. A charity with only a 1% overhead with a mission to make balloon animals for children dying in a famine will score well on Charity Navigator but does nothing to help the problem.
To be honest I haven't looked deeply into long-termism, but from what I've heard (eg, hearing Will MacAskill on a few podcasts) it seems to ignore a few things. Just like a bird in the hand is worth two in the bush, long-termers have no good way to estimate the likelihood of future events, and that discounting needs to increase greatly the further out one looks. At best many of these estimates are like the Drake Equation -- better than nothing, but with multiple orders of magnitude error bars.
There are other second-order reasons which don't seem to factor in, or at least haven't come across in the few hours of listening to long-termers talk about the issue. One is that by working to make a better world now, it effects the trajectory of future events much more directly than the low-probability guesswork they think my have an impact in the distant future.
That was always a convenient excuse to justify amassing lots of money, if necessary by theft (see SBF). They have no intention of actually doing good with it.
Let me reframe it. Among these trillions of people, there will be many who are 99% similar to you. Wouldn't you want that version of yourself to live a great life?
Sure! It's just not actually reasonably possible to predict what will lead to that outcome, whereas I can sorta (badly) predict what will make me happy tomorrow, my wife happy next week, and what will best improve conditions for my current fellow citizens of the world.
By that kind of logic, I’m actually a Boltzmann brain floating alone in an infinite void, so I don’t have to worry about anyone but myself.
You can certainly rationalize anything, but I fail to see what help that is to us.
"The great subverter of Pyrrhonism [radical skepticism] is action, and employment, and the occupations of common life. [...] I dine, I play a game of back-gammon, I converse, and am merry with my friends; and when after three or four hour's amusement, I wou'd return to these speculations, they appear so cold, and strain'd, and ridiculous, that I cannot find in my heart to enter into them any farther."
Hume
> Wouldn't you want that version of yourself to live a great life?
The biggest positive change you can make, even for future generations, is to uplevel the people who are alive today.
That's more of a Bentham's mugging.
how are "trillions of humans living in the far distant future" improbable, or poor information or has a lack of information? seems pretty obvious
Regardless, it is an extremely uncertain proposition that we can do anything in the present that would have a reliably positive impact on their lives in the far future. It's hard enough to figure something out for the billions of people who actually exist right now.
It seems obvious to you that there will be trillions of people alive in the far distant future? Please explain how you know this.
They claim isn't that there will be trillions of people alive at the same time; they are integrating over the course of tens and hundreds of thousands of years.
Although we are at a peak population of a bit over 8B people at the moment, it is estimated that more than 100B people we would classify as humans have ever lived. The population long ago was much smaller than 1B, but thousands of generations have lived and died.
What is something specific that an individual should have done or should not have done, say, 2000 years ago, that would have made a positive impact on your life?
Certainly there are individuals who have had that kind of impact, but I don't think one could have foreseen the consequences at the time, e.g., Jesus and Mohamed.
Likewise, the guy who invented the Segway probably thought he was going to revolutionize personal transportation, but that didn't pan out (yet, at least).
1. Suppose you have a chart of the total past and future history of human population.
2. Cover up the chart so only the data from the past to present day is visible.
3. Note that most humans in that subset exist near or at the present. You are one of these people today, it should make sense for you to be born in one of the densest parts of the graph.
4. Now uncover the graph. If there are trillions of humans in the future, it seems almost impossibly unlikely that you would born in a part of the graph with "so few" humans as today, and not in the far future.
Therefore, you must conclude that the actual graph rapidly drops to zero in the near future. QED.
This "doomsday argument" is a pretty shit one, but not worse than others I've seen arguing the opposite.
As a kernel developer, I have always had to worry about what happens in the 0.000001% of cases. You're either correct all of the time or you likely have an exploitable security hole in your code that will turn into a CVE at some point in the future that will cost a lot of money to fix. One has to use strategies to reason through proof that your code does the right thing in all states, including the insertion of delays and other fault injection techniques to trigger the cases that aren't normally hit.
It's also kind of like saying that an LLM only pushes 0.01% of people off the deep end in a psychotic episode, so it doesn't really matter. That doesn't actually get rid of the liability.
All that said, I will always refuse to write life safety code. I am not a good enough developer to do so.
A lower-key variant of this frequently comes into play with consulting or other sales pitches. "You spend <big number> per year on this <necessary business expense>. Our service will easily shave 2% off this, making the cost of our service completely negligible and this purchase an obviously good decision."
You've lost me; can you explain how these two relate?
For reasons explained in the article, we are bad at estimating small probabilities.
Similarly, we are bad at estimating small proportions ("easily shave 2%"). What is being claimed in the parentheses here is that there's a probability distribution of "how much costs are shaved" and that we can estimate where the bulk of its support is.
But we're not really good at making such estimates. Maybe there is some probability mass around 2%, but the bulk is around 0.5%. It seems like that's a small difference (just 1.5%!) but it's a factor of 4 in terms of savings.
So now we have a large number (annual spend), multiplied by a very uncertain number (cost shave, with poor experimental support), leading to a very uncertain outcome in terms of savings.
And it can be that, in reality, the costs of changing service turn out to overwhelm this outcome.
When modern advertising is a spectrum of “lies, damn lies, and statistics,” I don’t blame folks for crying foul and demanding a baseline level of truth in advertising. When folks trust but verify, this is seen as a change in the status quo by folks, and some of those folks who protest about it in those terms are trying to sell you something.
Came to read a programming language related polemic, stayed to read about philosophy.
This is a funny read in contrast to the latest Nate Silver book. He seems to have gone all in on justifying EV bets.
It's amusing to consider how much of a Rorschach test this article must be. But it's a great point, even if it arms us to abusively write off unwelcome ideas as scams. As the author points out, Pascal's reasoning is easily applied to an infinity of conceivable catastrophes - alien invasions, etc. That Pascal specifically applied his argument to the possibility of punishment by a biblical God was due to the psychological salience of that possibility in Pascal's culture - a truly balanced application of his fallacious reasoning would be completely paralyzing.
I often like to pair Pascal’s wager with Hitchens’s razor:
> Hitchens's razor is an epistemological razor that serves as a general rule for rejecting certain knowledge claims. It states:
> > What can be asserted without evidence can also be dismissed without evidence.
https://en.wikipedia.org/wiki/Hitchens%27s_razor
Pascal’s gamble wasn’t just about probability, it was about storytelling: the promise of nearly infinite payoff with minimal risk. That same allure is still at play today whenever people chase “moonshot” returns on crypto or quick-rich schemes.
It underscores a timeless lesson: no matter how much data or logic we have, we’re still wired to fall for well-crafted optimism and that means skepticism remains the best defense.
>people chase “moonshot” returns on crypto
Your comment would have been better if you'd chosen and example that did not create hundreds of thousands of millionaires.
> Your comment would have been better if you'd chosen and example that did not create many tens of thousands of millionaires.
Lotteries have also produced lots of millionaires. Crypto could produce lots of winners just from wealth transfer even if it was a zero sum or net negative game in terms of wealth creation.
Just bitcoin uses more energy than Finland, the energy sector in Finland makes lots of people rich. Of course people will earn money when people spend that amount of resources on something.
...and plenty more folks who lost their shirts- or even just their pizza money- on crypto scams in order to subsidize those millionaires.
Not to mention countries who subsidized the electricity.