I work as an electrical engineer in the energy sector, and one thing that my coworkers and I talk about on a near weekly basis is how unprepared the United States is for the increase in energy demand as a result of data centers. Electricity as a whole in the United States is split into three groups: generation, transmission, and distribution. When we do projects on the generation side, we often ask their engineers what the holdup is with other forms of energy. Solar presents a whole slew of challenges between needing clean solar panels, only being able to produce energy during the day, and needing huge amounts of land to generate any meaningful amount of electricity. With nuclear, we often hear that a) There is just no investment in it, and b) people are scared of it. I worry about electricity prices for the average consumer in 5-10 years often.
I mean even over the past year they've gone up substantially! and that's something the average person really takes for granted. I really hope that we can sort it out, and thanks for doing such important work.
One challenge to a “maintenance economy that is designed to care for our aging population” is that we are notoriously poor at maintaining things in this country (see bridges, roads, etc). In a decidedly non-collectivist society, how long before “maintenance” becomes a synonym for burden?
I work with an aging population in pain management and I can tell you the western medical world cannot save us. I agree with most everything and I am more socially liberal than anyone could imagine. But I have seen the trenches, boomers rely on pills and their docs, which they go round and round rarely improving their quality of life. when their docs are just trying to get them in and out to cover their ever growing overhead. Millennials and younger use toxic substances without consideration of their long term consequences and they are easier to get and in more different varieties then ever. The answer is not “fund cancer research” we KNOW what causes cancer ! and no one can stop their consumption out of short term effect, convenience or ignorance. We have round up in our tampons, medications in our water supply and heavy metals in our adipose tissue.
In Tyler Cowen's presentation to Google Deep Mind, he said he anticipates that AI-enabled improvements in healthcare would result in many people living into their late 90s. However, he did not make any comments about the implications of longer life expectancy on our Social Security and Medicare programs which are already on the verge of insolvency. Neither party wants to put forth a credible and balanced approach to financing the "maintenance economy" that Kyla describes.
The steeled (futurist?) argument might be that semaglutides and their ilk get better every generation; thus, maybe positive health outcomes increase. When does AI begin to make the connections between disparate data to propose disease cures? Also, there is promising research on myostatin regulators leading some to suggest we are entering an aesthetic revolution where we can have the body we want with little effort. As with all technology, this is the new low water mark.
I do think a lot of things AI is being used for isn't providing much value. But I think people underestimate the talent shortage for good software engineers and researchers. I work in robotics and we never remotely have enough people. I think it is reasonable to bet on making engineers so much more efficient that we can rebuild the parts of the economy that are struggling (manufacturing, R&D outside of AI).
There's some really important research on diminishing returns for research investment because all the easy stuff has been discovered. A key to solving this is scaling up brainpower/talent non-linearly with population.
I completed a second master's in ME (emphasis on controls/dynamics) but had to go back to my old industry (aerospace/radar) because I could not get past non-AI HR in job applications in the more modern industry. I suspect AI has not helped the application process. Engineers still need to maintain (live) networking skills!
This article is a fantastic synthesis of what's going on with AI, the economy and society. Really one stop shopping. About the only thing it doesn't include is what I read about just before this one posted: the collapse of the market for suburban office parks, though, I wonder if they could be repurposed as data centers to make more fake money and dumbass chatbots.
Well thought out and well written. It's too much, though. It could easily be three newsletters, one for each type of America.
On AI, I wonder about all the noise about it making us stupid. Did dictionaries make us stupid? Does Grammarly? Or does it just push us faster towards a sharper definition of what our human value-added is? I know that makes people nervous.
Every article you publish just seems to compel me to read it and stop everything else I’m doing. Amazing stuff and such clarity, it’s like a veil is lifted and all the unformed thoughts floating around in my head and nibbling at me suddenly come together.
With regard to Trump's tweet excoriating the CEO of Intel and his warm praise of Tim Cook, it makes me think we are entering a period of State Capitalism. Businesses that grovel the best (Cook's clever but tacky award) and promise to invest billions of dollars in the U.S., get Trump's blessing. Companies like Intel, which is having problems, don't.
Trump is operating a form of State Capitalism in which the central government (Trump) decides winners and losers. This type of system, unsurprisingly, lends itself to corruption and misallocation of capital.
Hi Kyla! Love this article as a synthesis of our weird economy. I’ve been thinking about getting in on the Labubu game…
One note - I think the OpenAI usage graph does not reflect a school-related drop in usage. It’s a graph from OpenRouter, an API for LLM calls, and most of the drop is from usage of one model (o4-mini). It’s probably more likely to be a single large OpenRouter customer moving to a different model or routing provider.
Phenomenal piece. Thanks for clarifying so many gaps in my knowledge.
If possible, I would like to see a future piece explaining how the military complex fits into this economy. Defence contract company stocks are booming. At $1 trillion dollars a year, I feel this also plays a huge inflationary role in the over-all economy, after all, bombs produce nothing, and don't generate wealth.
I’m sorry that I wasn’t following you until I saw you on The Daily Show. I immediately bought your book, listened to the podcast with Ezra Klein (then got his book) and am hanging on every word. Thanks for bridging the gap between this 70 year old and the future. Hope is in short supply. Keep up the good work!
I have nothing of substance to add to the conversation. I’m just here to stan Kyla for being a voice of reason in an age of insanity.
thank you!!
No girl, thank YOU
Must read stuff every time
thanks for reading!!
I work as an electrical engineer in the energy sector, and one thing that my coworkers and I talk about on a near weekly basis is how unprepared the United States is for the increase in energy demand as a result of data centers. Electricity as a whole in the United States is split into three groups: generation, transmission, and distribution. When we do projects on the generation side, we often ask their engineers what the holdup is with other forms of energy. Solar presents a whole slew of challenges between needing clean solar panels, only being able to produce energy during the day, and needing huge amounts of land to generate any meaningful amount of electricity. With nuclear, we often hear that a) There is just no investment in it, and b) people are scared of it. I worry about electricity prices for the average consumer in 5-10 years often.
I mean even over the past year they've gone up substantially! and that's something the average person really takes for granted. I really hope that we can sort it out, and thanks for doing such important work.
Thank god someone else is thinking about this
One challenge to a “maintenance economy that is designed to care for our aging population” is that we are notoriously poor at maintaining things in this country (see bridges, roads, etc). In a decidedly non-collectivist society, how long before “maintenance” becomes a synonym for burden?
I think that is going to be a very scary conversation
I work with an aging population in pain management and I can tell you the western medical world cannot save us. I agree with most everything and I am more socially liberal than anyone could imagine. But I have seen the trenches, boomers rely on pills and their docs, which they go round and round rarely improving their quality of life. when their docs are just trying to get them in and out to cover their ever growing overhead. Millennials and younger use toxic substances without consideration of their long term consequences and they are easier to get and in more different varieties then ever. The answer is not “fund cancer research” we KNOW what causes cancer ! and no one can stop their consumption out of short term effect, convenience or ignorance. We have round up in our tampons, medications in our water supply and heavy metals in our adipose tissue.
In Tyler Cowen's presentation to Google Deep Mind, he said he anticipates that AI-enabled improvements in healthcare would result in many people living into their late 90s. However, he did not make any comments about the implications of longer life expectancy on our Social Security and Medicare programs which are already on the verge of insolvency. Neither party wants to put forth a credible and balanced approach to financing the "maintenance economy" that Kyla describes.
I did a long piece on "health span" for freakonomics that got into what is going to have to be done - its all expensive
The steeled (futurist?) argument might be that semaglutides and their ilk get better every generation; thus, maybe positive health outcomes increase. When does AI begin to make the connections between disparate data to propose disease cures? Also, there is promising research on myostatin regulators leading some to suggest we are entering an aesthetic revolution where we can have the body we want with little effort. As with all technology, this is the new low water mark.
I do think a lot of things AI is being used for isn't providing much value. But I think people underestimate the talent shortage for good software engineers and researchers. I work in robotics and we never remotely have enough people. I think it is reasonable to bet on making engineers so much more efficient that we can rebuild the parts of the economy that are struggling (manufacturing, R&D outside of AI).
There's some really important research on diminishing returns for research investment because all the easy stuff has been discovered. A key to solving this is scaling up brainpower/talent non-linearly with population.
great point!
I completed a second master's in ME (emphasis on controls/dynamics) but had to go back to my old industry (aerospace/radar) because I could not get past non-AI HR in job applications in the more modern industry. I suspect AI has not helped the application process. Engineers still need to maintain (live) networking skills!
Sorry, I'm still trying to get over learning about "Fartcoin."
surprisingly not backed by the asset of fart, as one would assume
This article is a fantastic synthesis of what's going on with AI, the economy and society. Really one stop shopping. About the only thing it doesn't include is what I read about just before this one posted: the collapse of the market for suburban office parks, though, I wonder if they could be repurposed as data centers to make more fake money and dumbass chatbots.
Well thought out and well written. It's too much, though. It could easily be three newsletters, one for each type of America.
On AI, I wonder about all the noise about it making us stupid. Did dictionaries make us stupid? Does Grammarly? Or does it just push us faster towards a sharper definition of what our human value-added is? I know that makes people nervous.
I think it's the frictionless aspect! it's just so easy to... not think.
Every article you publish just seems to compel me to read it and stop everything else I’m doing. Amazing stuff and such clarity, it’s like a veil is lifted and all the unformed thoughts floating around in my head and nibbling at me suddenly come together.
thank you!!
Those of us in the Fourth America (the non-AI, non-healthcare, non-memecoin version of America) are baffled and confused by all of this.
Great piece, but your Labubu point would be stronger if China was primarily selling toys to Americans.
They are not. North America was just 5.5% of Pop Mart’s sales last year: https://kr-asia.com/pop-marts-global-push-pays-off-as-revenue-doubles-in-2024
China is both building the future and building the present cultural distractions.
With regard to Trump's tweet excoriating the CEO of Intel and his warm praise of Tim Cook, it makes me think we are entering a period of State Capitalism. Businesses that grovel the best (Cook's clever but tacky award) and promise to invest billions of dollars in the U.S., get Trump's blessing. Companies like Intel, which is having problems, don't.
Trump is operating a form of State Capitalism in which the central government (Trump) decides winners and losers. This type of system, unsurprisingly, lends itself to corruption and misallocation of capital.
I'd have to disagree strongly about state capitalism, as China has been using such a system with great success.
MIT Technology Review has been keeping track of Chinese technology advances for years.
USA needs to wake up, roll up the sleeves and get serious, and not
Superfluous about "Hollow Economies".
Hi Kyla! Love this article as a synthesis of our weird economy. I’ve been thinking about getting in on the Labubu game…
One note - I think the OpenAI usage graph does not reflect a school-related drop in usage. It’s a graph from OpenRouter, an API for LLM calls, and most of the drop is from usage of one model (o4-mini). It’s probably more likely to be a single large OpenRouter customer moving to a different model or routing provider.
There’s more info in the comments on this Tweet: https://x.com/mazeincoding/status/1952787361768587741?s=46&t=z_MDL9xr7sk2Ci3yEh8k-A
Phenomenal piece. Thanks for clarifying so many gaps in my knowledge.
If possible, I would like to see a future piece explaining how the military complex fits into this economy. Defence contract company stocks are booming. At $1 trillion dollars a year, I feel this also plays a huge inflationary role in the over-all economy, after all, bombs produce nothing, and don't generate wealth.
I’m sorry that I wasn’t following you until I saw you on The Daily Show. I immediately bought your book, listened to the podcast with Ezra Klein (then got his book) and am hanging on every word. Thanks for bridging the gap between this 70 year old and the future. Hope is in short supply. Keep up the good work!