Over the last six months, capital expenditures on AI—counting just information processing equipment and software, by the way—added more to the growth of the US economy than all consumer spending combined. You can just pull any of those quotes out—spending on IT for AI is so big it might be making up for economic losses from the tariffs, serving as a private sector stimulus program.
I'm not sure the comparison is apples to apples, but this article claims the current AI investment boom pales compared to the railroad investment boom in the 19th century.
> Next, Kedrosky bestows a 2x multiplier to this imputed AI CapEx level, which equates to a $624 billion positive impact on the US GDP. Based on an estimated US GDP figure of $30 trillion, AI CapEx is expected to amount to 2.08 percent of the US GDP!
Do note that peak spending on rail roads eventually amounted to ~20 percent of the US GDP in the 19th century. This means that the ongoing AI CapEx boom has lots of legroom to run before it reaches parity with the rail road boom of that bygone era.
People should keep in mind that there was no such thing as a GDP before the 1980's.
All that has been back-calculated, and the further back you go the more ridiculous it gets.
Excuses sounded plausible at the time but killed two birds with one stone.
Less rapid increase in government benefits which had become based on GNP for survival to cope with inflation, and further obscuring the ongoing poor economic performance of the 1980's going forward compared to how it was before 1970 numerically.
The people who were numerically smart before that and saw what things were like first hand were not fooled so easily.
Even using GDP back in the 1980's when it first came out, you couldn't get a good picture of the 1960's which were not that much earlier.
Consider other infrastructure such as the US highway system. There may be an expansive bubble, but infrastructure such as the increase in base power production needs to be factored as well.
look at the S&P 500 chart when ChatGPT came out. We were just on our way to flushing out the Covid excess money and then the AI narrative saved the market. AI narrative + inflation that is definitely way more than reported is propping up this market.
> There could be a crash that exceeds the dot com bust, at a time when the political situation through which such a crash would be navigated would be nightmarish.
If the general theme of this article is right (that it's a bubble soon to burst), I'm less concerned about the political environment and more concerned about the insane levels of debt.
If AI is indeed the thing propping up the economy, when that busts, unless there are some seriously unpopular moves made (Volcker level interest rates, another bailout leading to higher taxes, etc), then we're heading towards another depression. Likely one that makes the first look like a sideshow.
The only thing preventing that from coming true IMO is dollar hegemony (and keeping the world convinced that the world's super power having $37T of debt and growing is totally normal if you'd just accept MMT).
"Over the last six months, capital expenditures on AI—counting just information processing equipment and software, by the way—added more to the growth of the US economy than all consumer spending combined."
If this isn't the Singularity, there's going to be a big crash. What we have now is semi-useful, but too limited. It has to get a lot better to justify multiple companies with US $4 trillion valuations. Total US consumer spending is about $16 trillion / yr.
Remember the Metaverse/VR/AR boom? Facebook/Meta did somehow lose upwards of US$20 billion on that. That was tiny compared to the AI boom.
There has to be give and take to this as well. The AI increase is going to cost jobs. I see it in my work flow and our company. We used to pay artists to do artwork and editors to post content. Now we use AI to generate the artwork and AI to write the content. It's verified by a human, but it's still done by AI and saves a ton of time and money.
These are jobs that normally would have gone to a human and now go to AI. We haven't paid a cent for AI mind you -- it's all on the ChatGPT free tier or using this tool for the graphics: https://labs.google/fx/tools/image-fx
I could be wrong, but I think we are at the start of a major bloodbath as far as employment goes.... in tech mostly but also in anything that can be replaced by AI?
I'm worried. Does this mean there will be a boom in needing people for tradeskills and stuff? I honestly don't know what to think about the prospects moving forward.
The AI bubble is so big that it's draining useful investment from the rest of the economy. Hundreds of thousands of people are getting fired so billionaires can try to add a few more zeros to their bank account.
The best investment we can make would be to send the billionaires and AI researchers to an island somewhere and not let them leave until they develop an AI that's actually useful. In the meanwhile, the rest of us get to live productive lives.
I also wonder the extent to which "pseudo-black-box-AI" is potentially driving some of these crazy valuations now due to it actually being used in a lot algorithmic trading itself... seems like a prevalence of over-corrected models, all expecting "line go up" from recent historical data would be the perfect way to cook up a really "big beautiful bubble" so to speak...
If AI gets us into orbit ( https://news.ycombinator.com/item?id=44800051#44804687 ) or revitalizes nuclear, I'm fine with those things. It's true that AI usage can scale with availability better than most things but that's not a path to world domination.
Interesting piece, but the idea that this guy understands how oligarchs think seems way off. Jack Welch took General Electric from a global leader to a sad bag holder and he and his fans cheered progress with every positive quarterly report.
From a few weeks ago, see "Honey, AI Capex is Eating the Economy" / "AI capex is so big that it's affecting economic statistics" (365 points by throw0101c 18 days ago | hide | past | favorite | 355 comments):
They've been saying the same thing about whatever the trend of the moment is for years. Before this it was Magnificent 7 and before that it was FANG, and before that it was something else. Isn't this just sort of fundamental to how the economy works?
Imagine, the world's biggest economy propped up by hopes and dreams. Has anyone successfully monetized "AI" at a scale that generates a reasonable return on investment?
The article's comparison to the 19th century railroad boom is pretty spot on for how big it all feels, but maybe not so much for what actually happened.
Back then, the money poured into building real stuff like actual railroads and factories and making tangible products.
That kind of investment really grew the value of companies and was more about creating actual economic value than just making shareholders rich super fast.
More like the corpos are really excited about the post-human AI future, so they are pouring truckloads of money on it and this raises the GDP number. The well-being of the average folk is in decline.
52 comments
[ 3.4 ms ] story [ 66.6 ms ] threadNow, it does that at the expense of the average person, but it will definitely prop up the bubble just long enough for the next election cycle to hit.
As someone in an AI company right now - Almost every company we work with is using Azure wrapped OpenAI. We're not sure why, but that is the case.
https://wccftech.com/ai-capex-might-equal-2-percent-of-us-gd...
> Next, Kedrosky bestows a 2x multiplier to this imputed AI CapEx level, which equates to a $624 billion positive impact on the US GDP. Based on an estimated US GDP figure of $30 trillion, AI CapEx is expected to amount to 2.08 percent of the US GDP!
Do note that peak spending on rail roads eventually amounted to ~20 percent of the US GDP in the 19th century. This means that the ongoing AI CapEx boom has lots of legroom to run before it reaches parity with the rail road boom of that bygone era.
More like apples to octopus.
People should keep in mind that there was no such thing as a GDP before the 1980's.
All that has been back-calculated, and the further back you go the more ridiculous it gets.
Excuses sounded plausible at the time but killed two birds with one stone.
Less rapid increase in government benefits which had become based on GNP for survival to cope with inflation, and further obscuring the ongoing poor economic performance of the 1980's going forward compared to how it was before 1970 numerically.
The people who were numerically smart before that and saw what things were like first hand were not fooled so easily.
Even using GDP back in the 1980's when it first came out, you couldn't get a good picture of the 1960's which were not that much earlier.
Don't make me laugh trying for the 1860's :)
If the general theme of this article is right (that it's a bubble soon to burst), I'm less concerned about the political environment and more concerned about the insane levels of debt.
If AI is indeed the thing propping up the economy, when that busts, unless there are some seriously unpopular moves made (Volcker level interest rates, another bailout leading to higher taxes, etc), then we're heading towards another depression. Likely one that makes the first look like a sideshow.
The only thing preventing that from coming true IMO is dollar hegemony (and keeping the world convinced that the world's super power having $37T of debt and growing is totally normal if you'd just accept MMT).
If this isn't the Singularity, there's going to be a big crash. What we have now is semi-useful, but too limited. It has to get a lot better to justify multiple companies with US $4 trillion valuations. Total US consumer spending is about $16 trillion / yr.
Remember the Metaverse/VR/AR boom? Facebook/Meta did somehow lose upwards of US$20 billion on that. That was tiny compared to the AI boom.
These are jobs that normally would have gone to a human and now go to AI. We haven't paid a cent for AI mind you -- it's all on the ChatGPT free tier or using this tool for the graphics: https://labs.google/fx/tools/image-fx
I could be wrong, but I think we are at the start of a major bloodbath as far as employment goes.... in tech mostly but also in anything that can be replaced by AI?
I'm worried. Does this mean there will be a boom in needing people for tradeskills and stuff? I honestly don't know what to think about the prospects moving forward.
The AI bubble is so big that it's draining useful investment from the rest of the economy. Hundreds of thousands of people are getting fired so billionaires can try to add a few more zeros to their bank account.
The best investment we can make would be to send the billionaires and AI researchers to an island somewhere and not let them leave until they develop an AI that's actually useful. In the meanwhile, the rest of us get to live productive lives.
* https://paulkedrosky.com/honey-ai-capex-ate-the-economy/
* https://news.ycombinator.com/item?id=44609130
Back then, the money poured into building real stuff like actual railroads and factories and making tangible products.
That kind of investment really grew the value of companies and was more about creating actual economic value than just making shareholders rich super fast.
Post-Labor Economics Lecture 01 - "Better, Faster, Cheaper, Safer" (2025 update) https://www.youtube.com/watch?v=UzJ_HZ9qw14