In his latest “Pivot to AI” OpenAI faces cash crunch in 2026 as bills come due, @davidgerard (accurately) says: “#OpenAI works by setting as much money as it can on fire, as fast as possible.”
-
In his latest “Pivot to AI” OpenAI faces cash crunch in 2026 as bills come due, @davidgerard (accurately) says: “#OpenAI works by setting as much money as it can on fire, as fast as possible.”
But I want to know: if we had $200B in, say, $100 notes, and we literally set them on fire:
- would it dispose of the money faster? How long would it take?
- would the impact on the environment be worse or less bad?
I gotta think there is someone on the #fediverse with the wherewithal to figure this out. Surely if we boost this, it will nerdsnipe the right person and we will learn the answer.
-
In his latest “Pivot to AI” OpenAI faces cash crunch in 2026 as bills come due, @davidgerard (accurately) says: “#OpenAI works by setting as much money as it can on fire, as fast as possible.”
But I want to know: if we had $200B in, say, $100 notes, and we literally set them on fire:
- would it dispose of the money faster? How long would it take?
- would the impact on the environment be worse or less bad?
I gotta think there is someone on the #fediverse with the wherewithal to figure this out. Surely if we boost this, it will nerdsnipe the right person and we will learn the answer.
@paco @davidgerard@circumstances.run this would be a perfect submission to XKCD what-if (although I dunno if it still takes submissions)
-
@paco @davidgerard@circumstances.run this would be a perfect submission to XKCD what-if (although I dunno if it still takes submissions)
-
@paco @davidgerard@circumstances.run this would be a perfect submission to XKCD what-if (although I dunno if it still takes submissions)
200 billion in $100 notes is 2 billion notes or about 2,000 tons of rag paper and polymer.
Which would burn to release less than 10% of what OpenAI took to do a training run to make ChatGPT4.
(50 gigawatt hours of electricity => a few times that in thermal energy.)
-
In his latest “Pivot to AI” OpenAI faces cash crunch in 2026 as bills come due, @davidgerard (accurately) says: “#OpenAI works by setting as much money as it can on fire, as fast as possible.”
But I want to know: if we had $200B in, say, $100 notes, and we literally set them on fire:
- would it dispose of the money faster? How long would it take?
- would the impact on the environment be worse or less bad?
I gotta think there is someone on the #fediverse with the wherewithal to figure this out. Surely if we boost this, it will nerdsnipe the right person and we will learn the answer.
-
@michael_w_busch Thanks. But how long will it take to burn? I’m guessing days or weeks, not years.
@diazona -
In his latest “Pivot to AI” OpenAI faces cash crunch in 2026 as bills come due, @davidgerard (accurately) says: “#OpenAI works by setting as much money as it can on fire, as fast as possible.”
But I want to know: if we had $200B in, say, $100 notes, and we literally set them on fire:
- would it dispose of the money faster? How long would it take?
- would the impact on the environment be worse or less bad?
I gotta think there is someone on the #fediverse with the wherewithal to figure this out. Surely if we boost this, it will nerdsnipe the right person and we will learn the answer.
@paco @davidgerard Former chemist, so a nerd snipe on anything to do with rapid oxidation and other fun material transformations is possible, but I'm going to give it the ol' Fermi problem try first.
A US note (regardless of denomination) is 156x66 mm, or about 0.01 square meter. Let's start by laying them all flat, in a single one-bill layer, to keep things simple. Assuming we're going to burn $100 bills (to maximize our literal cash burn), that's about $10,000 per square meter or 20,000,000 square meters to equal $200B.
This works out to a circular disk of $100 bills about 2523 meters in radius.
Individual bills are thin. They'd burn pretty fast, once lit. If we light the middle of the circle and the flame front expands radially outward by 25 mm (about an inch) per second, that's enough to consume an entire bill (lengthwise) in 6 seconds. Sounds about right, based on how fast thin paper seems to burn. At that rate, it'll take 2523 meters / 0.025 meters/second ~= 100,000 seconds or about 1.15 days to burn the entire disk of flammable currency.
But I suspect that past a certain point, the expanding fire might be hot enough to start igniting things further ahead of the immediate flames, in which case the flame front would expand much faster. Also, you'd likely get some updrafts that would carry burning Benjamins further afield, which would start spot fires in other parts of the gigadollar disk some distance away, each one burning at a similar rate. Spot fires could sharply reduce the total time, easily a factor of 10 or more, especially if they started fairly early on. My Fermi-inspired guess is you're looking at "a few hours, maybe longer if the fire is extremely well-behaved, maybe less if it goes total chaos muppet", to torch things that way.
I don't have a good sense of whether making stacks (more to burn per area, but less area) makes things go faster, and that's a branch of material science I'm not super familiar with, so I'll leave it at that.
-
@paco @davidgerard Former chemist, so a nerd snipe on anything to do with rapid oxidation and other fun material transformations is possible, but I'm going to give it the ol' Fermi problem try first.
A US note (regardless of denomination) is 156x66 mm, or about 0.01 square meter. Let's start by laying them all flat, in a single one-bill layer, to keep things simple. Assuming we're going to burn $100 bills (to maximize our literal cash burn), that's about $10,000 per square meter or 20,000,000 square meters to equal $200B.
This works out to a circular disk of $100 bills about 2523 meters in radius.
Individual bills are thin. They'd burn pretty fast, once lit. If we light the middle of the circle and the flame front expands radially outward by 25 mm (about an inch) per second, that's enough to consume an entire bill (lengthwise) in 6 seconds. Sounds about right, based on how fast thin paper seems to burn. At that rate, it'll take 2523 meters / 0.025 meters/second ~= 100,000 seconds or about 1.15 days to burn the entire disk of flammable currency.
But I suspect that past a certain point, the expanding fire might be hot enough to start igniting things further ahead of the immediate flames, in which case the flame front would expand much faster. Also, you'd likely get some updrafts that would carry burning Benjamins further afield, which would start spot fires in other parts of the gigadollar disk some distance away, each one burning at a similar rate. Spot fires could sharply reduce the total time, easily a factor of 10 or more, especially if they started fairly early on. My Fermi-inspired guess is you're looking at "a few hours, maybe longer if the fire is extremely well-behaved, maybe less if it goes total chaos muppet", to torch things that way.
I don't have a good sense of whether making stacks (more to burn per area, but less area) makes things go faster, and that's a branch of material science I'm not super familiar with, so I'll leave it at that.
@dpnash @paco @davidgerard this clearly needs mythbusters to do a test.
-
@dpnash @paco @davidgerard this clearly needs mythbusters to do a test.
One advantage of doing the test this way is the uncertainty in the time estimate is high enough that nobody (except possibly your financial advisor and some carefully chosen charities) notices you only burned $199.99B instead of the intended $200B.
-
In his latest “Pivot to AI” OpenAI faces cash crunch in 2026 as bills come due, @davidgerard (accurately) says: “#OpenAI works by setting as much money as it can on fire, as fast as possible.”
But I want to know: if we had $200B in, say, $100 notes, and we literally set them on fire:
- would it dispose of the money faster? How long would it take?
- would the impact on the environment be worse or less bad?
I gotta think there is someone on the #fediverse with the wherewithal to figure this out. Surely if we boost this, it will nerdsnipe the right person and we will learn the answer.
@paco @davidgerard accordingbto the official dimensions, a US dollar bill of any denomination has a volume of just over a millilitre
with 2 billion of those, you could stack them into a cube with a side length of almost exactly 13.12 meters (in a perfect world; you might need to add some extra space for shelving and/or reinforcements) and a weight of just about 2000 tons
-
@paco @davidgerard Former chemist, so a nerd snipe on anything to do with rapid oxidation and other fun material transformations is possible, but I'm going to give it the ol' Fermi problem try first.
A US note (regardless of denomination) is 156x66 mm, or about 0.01 square meter. Let's start by laying them all flat, in a single one-bill layer, to keep things simple. Assuming we're going to burn $100 bills (to maximize our literal cash burn), that's about $10,000 per square meter or 20,000,000 square meters to equal $200B.
This works out to a circular disk of $100 bills about 2523 meters in radius.
Individual bills are thin. They'd burn pretty fast, once lit. If we light the middle of the circle and the flame front expands radially outward by 25 mm (about an inch) per second, that's enough to consume an entire bill (lengthwise) in 6 seconds. Sounds about right, based on how fast thin paper seems to burn. At that rate, it'll take 2523 meters / 0.025 meters/second ~= 100,000 seconds or about 1.15 days to burn the entire disk of flammable currency.
But I suspect that past a certain point, the expanding fire might be hot enough to start igniting things further ahead of the immediate flames, in which case the flame front would expand much faster. Also, you'd likely get some updrafts that would carry burning Benjamins further afield, which would start spot fires in other parts of the gigadollar disk some distance away, each one burning at a similar rate. Spot fires could sharply reduce the total time, easily a factor of 10 or more, especially if they started fairly early on. My Fermi-inspired guess is you're looking at "a few hours, maybe longer if the fire is extremely well-behaved, maybe less if it goes total chaos muppet", to torch things that way.
I don't have a good sense of whether making stacks (more to burn per area, but less area) makes things go faster, and that's a branch of material science I'm not super familiar with, so I'll leave it at that.
@dpnash @paco @davidgerard I feel like this could be solved or at least animated with Blender
-
@paco @davidgerard Former chemist, so a nerd snipe on anything to do with rapid oxidation and other fun material transformations is possible, but I'm going to give it the ol' Fermi problem try first.
A US note (regardless of denomination) is 156x66 mm, or about 0.01 square meter. Let's start by laying them all flat, in a single one-bill layer, to keep things simple. Assuming we're going to burn $100 bills (to maximize our literal cash burn), that's about $10,000 per square meter or 20,000,000 square meters to equal $200B.
This works out to a circular disk of $100 bills about 2523 meters in radius.
Individual bills are thin. They'd burn pretty fast, once lit. If we light the middle of the circle and the flame front expands radially outward by 25 mm (about an inch) per second, that's enough to consume an entire bill (lengthwise) in 6 seconds. Sounds about right, based on how fast thin paper seems to burn. At that rate, it'll take 2523 meters / 0.025 meters/second ~= 100,000 seconds or about 1.15 days to burn the entire disk of flammable currency.
But I suspect that past a certain point, the expanding fire might be hot enough to start igniting things further ahead of the immediate flames, in which case the flame front would expand much faster. Also, you'd likely get some updrafts that would carry burning Benjamins further afield, which would start spot fires in other parts of the gigadollar disk some distance away, each one burning at a similar rate. Spot fires could sharply reduce the total time, easily a factor of 10 or more, especially if they started fairly early on. My Fermi-inspired guess is you're looking at "a few hours, maybe longer if the fire is extremely well-behaved, maybe less if it goes total chaos muppet", to torch things that way.
I don't have a good sense of whether making stacks (more to burn per area, but less area) makes things go faster, and that's a branch of material science I'm not super familiar with, so I'll leave it at that.
@dpnash @paco @davidgerard That scene in The Dark Night where the Joker burned a huge pile of money always bothered me, it seemed like most of the money on the inside of the pile would be fine.
-
@paco @davidgerard Former chemist, so a nerd snipe on anything to do with rapid oxidation and other fun material transformations is possible, but I'm going to give it the ol' Fermi problem try first.
A US note (regardless of denomination) is 156x66 mm, or about 0.01 square meter. Let's start by laying them all flat, in a single one-bill layer, to keep things simple. Assuming we're going to burn $100 bills (to maximize our literal cash burn), that's about $10,000 per square meter or 20,000,000 square meters to equal $200B.
This works out to a circular disk of $100 bills about 2523 meters in radius.
Individual bills are thin. They'd burn pretty fast, once lit. If we light the middle of the circle and the flame front expands radially outward by 25 mm (about an inch) per second, that's enough to consume an entire bill (lengthwise) in 6 seconds. Sounds about right, based on how fast thin paper seems to burn. At that rate, it'll take 2523 meters / 0.025 meters/second ~= 100,000 seconds or about 1.15 days to burn the entire disk of flammable currency.
But I suspect that past a certain point, the expanding fire might be hot enough to start igniting things further ahead of the immediate flames, in which case the flame front would expand much faster. Also, you'd likely get some updrafts that would carry burning Benjamins further afield, which would start spot fires in other parts of the gigadollar disk some distance away, each one burning at a similar rate. Spot fires could sharply reduce the total time, easily a factor of 10 or more, especially if they started fairly early on. My Fermi-inspired guess is you're looking at "a few hours, maybe longer if the fire is extremely well-behaved, maybe less if it goes total chaos muppet", to torch things that way.
I don't have a good sense of whether making stacks (more to burn per area, but less area) makes things go faster, and that's a branch of material science I'm not super familiar with, so I'll leave it at that.
@dpnash @paco @davidgerard here’s some prior art, a million £ takes about an hour to burn. https://en.wikipedia.org/wiki/K_Foundation_Burn_a_Million_Quid
-
@Taco_lad @paco @michael_w_busch @diazona Excellent analysis. If those notes burning end to end took ten seconds each, a little over 634 years.
-
@paco @davidgerard@circumstances.run this would be a perfect submission to XKCD what-if (although I dunno if it still takes submissions)
-
@paco @davidgerard Former chemist, so a nerd snipe on anything to do with rapid oxidation and other fun material transformations is possible, but I'm going to give it the ol' Fermi problem try first.
A US note (regardless of denomination) is 156x66 mm, or about 0.01 square meter. Let's start by laying them all flat, in a single one-bill layer, to keep things simple. Assuming we're going to burn $100 bills (to maximize our literal cash burn), that's about $10,000 per square meter or 20,000,000 square meters to equal $200B.
This works out to a circular disk of $100 bills about 2523 meters in radius.
Individual bills are thin. They'd burn pretty fast, once lit. If we light the middle of the circle and the flame front expands radially outward by 25 mm (about an inch) per second, that's enough to consume an entire bill (lengthwise) in 6 seconds. Sounds about right, based on how fast thin paper seems to burn. At that rate, it'll take 2523 meters / 0.025 meters/second ~= 100,000 seconds or about 1.15 days to burn the entire disk of flammable currency.
But I suspect that past a certain point, the expanding fire might be hot enough to start igniting things further ahead of the immediate flames, in which case the flame front would expand much faster. Also, you'd likely get some updrafts that would carry burning Benjamins further afield, which would start spot fires in other parts of the gigadollar disk some distance away, each one burning at a similar rate. Spot fires could sharply reduce the total time, easily a factor of 10 or more, especially if they started fairly early on. My Fermi-inspired guess is you're looking at "a few hours, maybe longer if the fire is extremely well-behaved, maybe less if it goes total chaos muppet", to torch things that way.
I don't have a good sense of whether making stacks (more to burn per area, but less area) makes things go faster, and that's a branch of material science I'm not super familiar with, so I'll leave it at that.
@dpnash @paco @davidgerard
Not sure how air circulation will affect the burn rate. There’d be hot air rising from the burn area which would affect the outward spread of flames, effectively sucking air in from the periphery, slowing the outward spread. Any transient air from one side or another might push the flames horizontally. -
@paco @davidgerard Former chemist, so a nerd snipe on anything to do with rapid oxidation and other fun material transformations is possible, but I'm going to give it the ol' Fermi problem try first.
A US note (regardless of denomination) is 156x66 mm, or about 0.01 square meter. Let's start by laying them all flat, in a single one-bill layer, to keep things simple. Assuming we're going to burn $100 bills (to maximize our literal cash burn), that's about $10,000 per square meter or 20,000,000 square meters to equal $200B.
This works out to a circular disk of $100 bills about 2523 meters in radius.
Individual bills are thin. They'd burn pretty fast, once lit. If we light the middle of the circle and the flame front expands radially outward by 25 mm (about an inch) per second, that's enough to consume an entire bill (lengthwise) in 6 seconds. Sounds about right, based on how fast thin paper seems to burn. At that rate, it'll take 2523 meters / 0.025 meters/second ~= 100,000 seconds or about 1.15 days to burn the entire disk of flammable currency.
But I suspect that past a certain point, the expanding fire might be hot enough to start igniting things further ahead of the immediate flames, in which case the flame front would expand much faster. Also, you'd likely get some updrafts that would carry burning Benjamins further afield, which would start spot fires in other parts of the gigadollar disk some distance away, each one burning at a similar rate. Spot fires could sharply reduce the total time, easily a factor of 10 or more, especially if they started fairly early on. My Fermi-inspired guess is you're looking at "a few hours, maybe longer if the fire is extremely well-behaved, maybe less if it goes total chaos muppet", to torch things that way.
I don't have a good sense of whether making stacks (more to burn per area, but less area) makes things go faster, and that's a branch of material science I'm not super familiar with, so I'll leave it at that.
@dpnash @paco @davidgerard yet another countless example of why I fucking love the Fediverse.
-
@dpnash @paco @davidgerard
Not sure how air circulation will affect the burn rate. There’d be hot air rising from the burn area which would affect the outward spread of flames, effectively sucking air in from the periphery, slowing the outward spread. Any transient air from one side or another might push the flames horizontally.@qurlyjoe @paco @davidgerard Yeah, there are a *lot* of variables that are hard to envision. This particular scenario is the combustion equivalent of a spherical cow in many ways.
Rising hot air would definitely stir things up a lot, so you probably don’t have the nice clean circular flame front I was starting out with, your “fuel” is likewise going to get disturbed a bit, and you also get embers blown around that can start new fires elsewhere. All of which change the actual time quite a bit, either way.
Possibly actual measurements on dry grass fires (also low mass but rapidly burning fuel) would be informative (how long to burn a 20 million square meter area?) Only observational physics here for this one, not experimental.
-
@michael_w_busch @diazona @paco Also: how many trees need to be cut to produce those 2000 tons of rag paper?
-
In his latest “Pivot to AI” OpenAI faces cash crunch in 2026 as bills come due, @davidgerard (accurately) says: “#OpenAI works by setting as much money as it can on fire, as fast as possible.”
But I want to know: if we had $200B in, say, $100 notes, and we literally set them on fire:
- would it dispose of the money faster? How long would it take?
- would the impact on the environment be worse or less bad?
I gotta think there is someone on the #fediverse with the wherewithal to figure this out. Surely if we boost this, it will nerdsnipe the right person and we will learn the answer.
@paco @davidgerard I need to buy that almost free methane from US shell cracking industry and just let it fly into atmosphere. usually they just burn it because methane is more potent greenhouse gas, and it's almost free.
please invest in my idea, it's the most efficient way to burn money
