How AI conquered the US economy: A visual FAQ

https://www.derekthompson.org/p/how-ai-conquered-the-us-economy-a

by rbanffy • 12 months ago

287 231 12 months ago

derekthompson.org

croes2• 12 months ago

I see hardware and AI companies‘ revenues rise.

Shouldn’t the customers‘ revenue also rise if AI fulfills its productivity promises?

Seems like the only ones getting rich in this gold rush are the shovel sellers. Business as usual.

mewpmewp22• 12 months ago

If it's automation it could also reduce costs of the customers. But that is a very complex question. It could be that there isn't enough competition in AI and so the customers are getting only marginal gains while AI company gets the most. It could also be that for customers the revenue / profits will be delayed as implementation will take time, and it could be upfront investment.

thecupisblue4• 12 months ago

The biggest problem is the inability of corporate middle management to actually leverage GenAI.

sofixa• 12 months ago

Shouldn’t the customers‘ revenue also rise if AI fulfills its productivity promises

Not necessarily, see the Jevons paradox.

metalliqaz2• 12 months ago

Applying the Jevons paradox to this scenario should still result in revenues going up, assuming the employee labor being optimized adds value to the company. (they would add more)

croes2• 12 months ago

Jevron is about higher resource consumption and costs but the output and therefore the revenue should rise too.

Maybe not the profit but at least the revenue.

amunozo• 12 months ago

This is going to end badly, I am afraid.

m_ke2• 12 months ago

Could all pop today if GPT5 doesn’t benchmark hack hard on some new made up task.

mewpmewp22• 12 months ago

I don't expect GPT-5 to be anything special, it seems OpenAI hasn't been able to keep its lead, but even current level of LLMs to me justifies the market valuations. Of course I might eat my words saying that OpenAI is behind, but we'll see.

falcor844• 12 months ago

I don't see how it would "all pop" - same as with the internet bubble, even if the massive valuations disappear, it seems clear to me that the technology is already massively disruptive and will continue growing its impact on the economy even if we never reach AGI.

paulnpace• 12 months ago

Everywhere I use "A.I.", I am forced to use "A.I.", and the thing I am forced to use it for has been made worse, as a result of the "A.I."

For those comparing the "A.I." bubble to the .com bubble, they are missing the point that even a mostly normie user such as myself, logging on with my 14.4 kbps modem I instantly had something that was new and useful. With "A.I.", I haven't found anything, at least for myself, that is useful.

stackbutterflow• 12 months ago

Predicting the future is always hard.

But the only thing I've seen in my life that most resembles what is happening with AI, the hype, its usefulness beyond the hype, vapid projects, solid projects, etc, is the rise of the internet.

Based on this I would say we're in the 1999-2000 era. If it's true what does it mean for the future?

baxtr• 12 months ago

"It is difficult to make predictions, especially about the future" - Yogi Berra (?)

But let’s assume we can for a moment.

If we’re living in a 1999 moment, then we might be on a Gartner Hype Cycle like curve. And I assume we’re on the first peak.

Which means that the "trough of disillusionment" will follow.

This is a phase in Hype Cycle, following the initial peak of inflated expectations, where interest in a technology wanes as it fails to deliver on early promises.

baggachipz• 12 months ago

Classic repeat of the Gartner Hype Cycle. This bubble pop will dwarf the dot-bomb era. There's also no guarantee that the "slope of enlightenment" phase will amount to much beyond coding assistants. GenAI in its current form will never be reliable enough to do so-called "Agentic" tasks in everyday lives.

This bubble also seems to combine the worst of the two huge previous bubbles; the hype of the dot-com bubble plus the housing bubble in the way of massive data center buildout using massive debt and security bundling.

thecupisblue4• 12 months ago

GenAI in its current form will never be reliable enough to do so-called "Agentic" tasks in everyday lives

No, but GenAI in it's current form is insanely useful and is already shifting the productivity gears into a higher level. Even without 100% reliable "agentic" task execution and AGI, this is already some next level stuff, especially for non-technical people.

brookst• 12 months ago

The Internet in its 1999 form was never going to be fast enough or secure enough to support commerce, banking, or business operations.

ben_w• 12 months ago

Mm. Partial agree, partial disagree.

These things, as they are right now, are essentially at the performance level of an intern or recent graduate in approximately all academic topics (but not necessarily practical topics), that can run on high-end consumer hardware. The learning curves suggest to me limited opportunities for further quality improvements within the foreseeable future… though "foreseeable future" here means "18 months".

I definitely agree it's a bubble. Many of these companies are priced with the assumption that they get most of the market; they obviously can't all get most of the market, and because these models are accessible to the upper end of consumer hardware, there's a reasonable chance none of them will be able to capture any of the market because open models will be zero cost and the inference hardware is something you had anyway so it's all running locally.

Other than that, to the extent that I agree with you that:

GenAI in its current form will never be reliable enough to do so-called "Agentic" tasks in everyday lives

I do so only in that not everyone wants (or would even benefit from) a book-smart-no-practical-experience intern, and not all economic tasks are such that book-smarts count for much anyway. This set of AI advancements didn't suddenly cause all cars manufacturers to suddenly agree that this was the one weird trick holding back level 5 self driving, for example.

But for those of us who can make use of them, these models are already useful (and, like all power tools, dangerous when used incautiously) beyond merely being coding assistants.

Traubenfuchs• 12 months ago

I fully agree that there will be a pop, there must be. Current evaluations and investments are based on monumentally society destroying assumptions. But with every disappointing, incremental and non evolutionary model generation the chance increases that the world at large realizes that those assumptions are wrong.

What should I do with my ETF? Sell now, wait for the inevitable crash? Be all modern long term investment style: "just keep invested what you don't need in the next 10 years bro"?

This really keeps me up at night.

bspammer• 12 months ago

Last week I tested out the agent mode of chatGPT by asking it to plan a week's meals, then add all the ingredients to an online shopping basket for me. It worked pretty much flawlessly, the only problem was it ran out of time before it could add the last few ingredients, which doesn't exactly seem like an unsolvable problem.

keiferski2• 12 months ago

Well, there’s a fundamental difference: the Internet blew up because it enabled people to connect with each other more easily: culturally, economically, politically.

AI is more-or-less replacing people, not connecting them. In many cases this is economically valuable, but in others I think it just pushes the human connection into another venue. I wouldn’t be surprised if in-person meetup groups really make a comeback, for example.

So if a prediction about AI involves it replacing human cultural activities (say, the idea that YouTube will just be replaced by AI videos and real people will be left out of a job), then I’m quite bearish. People will find other ways to connect with each other instead.

dfedbeef2• 12 months ago

There's also the difference that the internet worked.

LinuxAmbulance• 12 months ago

Businesses are overly optimistic about AI replacing people.

For very simple jobs, like working in a call center? Sure.

But the vast majority of all jobs aren't ones that AI can replace. Anything that requires any amount of context sensitive human decision making, for example.

There's no way that AI can deliver on the hype we have now, and it's going to crash. The only question is how hard - a whimper or a bang?

AlexandrB3• 12 months ago

There's also an element of economic drag caused by AI that did not exist with the internet: for example, the proliferation of slop content and AI-driven scams. The internet did not almost immediately make previous modes of doing business more difficult but instead slowly replaced them with more efficient alternatives. AI has already enshittified things like Google search and product reviews. In many ways AI is undoing some of the gains of the internet without providing a replacement.

gilbetron2• 12 months ago

Connecting to other people, I think, is going to see a surge of desire. I've been feeling it, many I talk to are feeling. One strange piece of anecdata is that I've been working out (with my son) at a gym for the past 3 years, very consistently. Everybody has always just done their thing, almost entirely individuals working out, with a handful of pairs. I'm old, so I remember when working out at gyms would be a much more social situation, but times change.

There are a number of regulars that we've seen there for years now. Barely have interacted with them. Suddenly, the past week, through a chance interaction, two of them individually have talked to us and ended up introducing themselves. Plus I see other people talking more and more.

I think we've hit a on-your-own saturation point and the pendulum might swing the other way.

But I'm really staying away from many "predict the AI future" because we just don't know where it is going. I've read thousands of scifi books and nonfiction and futurist and any number of options are open and no one, and I mean no one, knows what will happen with any kind of certainty.

api• 12 months ago

I too lived through the dot.com bubble and AI feels identical in so many ways.

AI is real just like the net was real, but the current environment is very bubbly and will probably crash.

thewebguyd2• 12 months ago

It definitely feels identical. We had companies that never had any hope of being profitable (or even doing anything related to the early internet to begin with), but put .com in your name and suddenly you are flooded with hype and cash.

Same thing now with AI. The capital is going to dry up eventually, no one is profitable right now and its questionable whether or not they can be at a price consumers would be willing or able to pay.

Models are going to become a commodity, just being an "AI Company" isn't a moat and yet every one of the big names are being invested in as if they are going to capture the entire market, or if there even will be a market in the first place.

Investors are going to get nervous, eventually, and start expecting a return, just like .com. Once everyone realizes AGI isn't going to happen, and realize you aren't going to meet the expected return running a $200/month chatbot, it'll be game over.

bwfan123• 12 months ago

I lived through dot-com, and there are so many parallels. Large amount of money is chasing a dream which wont materialize in the near term.

Recent deja-vus are articles like this:

"The 20-Somethings Are Swarming San Francisco’s A.I. Boom" and

"Tech Billboards Are All Over San Francisco. Can You Decode Them?"

If I recall correctly, after the 2000 bust, folks fled silicon valley abandoning their leased BMWs at the SFO airport. 101 had no traffic jams. I wonder if that will repeat this time around.

oezi• 12 months ago

It means you should hold on to your investments and be prepared to exit the market.

gilbetron2• 12 months ago

Almost three years ago I gave a talk to a Staff+ group where I work and told them that AI felt like the internet in 1995 timeframe. It still does, and I agree we seem to be in the late-90s now. However, we need to be careful, this feels similar in a world-changing way, but it is different on so many levels. Amongst other things, the world is far more tech savvy than in the 90s. Plus nearly everyone alive and making decisions now were alive and making decisions in the 90s. That's why we're seeing this insane "all in" mentality - there is almost certainly going to be new Googles/Facebooks/Apples/Amazons that come out of this, or at least that is what capital believes.

I hate, hate, hate the Gartner Hype Cycle thing, it's just a dumb statement that things can get overhyped. Instead, I see it as a Cambrian explosion of everybody everywhere trying to find use cases for this new tech. Nearly all of them will fail, but many won't and we'll have a different world in a few years. There will be crashes and crazes and (hopefully) generally positive changes in the world.

It could be Terminator or Handmaid's Tale or 1984, too. The main difference between the 90s and now is that the world is much more educated on the downsides of tech and so there isn't that giddy sense of coolness from the 90s, and that bums me out.

hnhg• 12 months ago

I found this the most interesting part of the whole essay - "the ten largest companies in the S&P 500 have so dominated net income growth in the last six years that it’s becoming more useful to think about an S&P 10 vs an S&P 490" - which then took me here: https://insight-public.sgmarkets.com/quant-motion-pictures/o...

Can anyone shed light on what is going on between these two groups. I wasn't convinced by the rest of the argument in the article, and I would like something that didn't just rely on "AI" as an explanation.

rogerkirkness• 12 months ago

Winner takes most is now true at the global economy level.

moi2388• 12 months ago

They are 40% of the S&P 500, so it makes sense that they are primary drivers of its growth.

They are also all tech companies, which had a really amazing run during Covid.

They also resemble companies with growth potential, whereas other companies such as P&G or Walmart might’ve saturated their market already

andsoitis5• 12 months ago

They are also all tech companies, which had a really amazing run during Covid.

Only 8 out of the 10 are. Berkshire and JP Morgan are not. It is also arguable whether Tesla is a tech company or whether it is a car company.

nowayno5833• 12 months ago

It is a very complex phenomenon, with no single driving force. The usual culprit is uncertainty, which itself can have a ton of root causes (say, tariffs changing every few weeks, or higher inflation due to government subsidies).

In more uncertain scenarios small companies can't take risks as well as big companies. The last 2 years have seen AI, which is a large risk these big companies invested in, pay off. But due to uncertainty smallish companies couldn't capitalize.

But that's only one possible explanation!

automatic61312• 12 months ago

The last 2 years have seen AI, which is a large risk these big companies invested in, pay off

LOL. It's paying off right now, because There Is No Alternative. But at some point, the companies and investors are going to want to make back these hundreds of billions. And the only people making money are Nvidia, and sort-of Microsoft through selling more Azure.

Once it becomes clear that there's no trillion dollar industry in cheating-at-homework-for-schoolkids, and nvidia stop selling more in year X than X-1, very quickly will people realize that the last 2 years have been a massive bubble.

foolswisdom• 12 months ago

The primary goal of big companies is (/has become) maintaining market dominance, but this doesn't always translate to a well run business with great profits, it depends on internal and external factors. Maybe profits should have actually gone down due to tarrifs and uncertainty but the big companies have kept profit stable.

andsoitis5• 12 months ago

Maybe profits should have actually gone down due to tarrifs and uncertainty but the big companies have kept profit stable.

If you’re referencing Trump’s tariffs, they have only come into effect now, so the economic effects will be felt in the months and years ahead.

whitej125• 12 months ago

That which might be of additional interest... look at how the top 10 of the S&P 500 has changed over the decades[1].

At any point in time the world thinks that those top 10 are unstoppable. In the 90's and early 00's... GE was unstoppable and the executive world was filled with acolytes of Jack Welch. Yet here we are.

Five years ago I think a lot of us saw Apple and Google and Microsoft as unstoppable. But 5-10 years from now I bet you we'll see new logos in the top 10. NVDA is already there. Is Apple going to continue dominance or go the way of Sony? Is the business model of the internet changing such that Google can't react quick enough. Will OpenAI go public (or any foundational model player).

I don't know what the future will be but I'm pretty sure it will be different.

[1] https://www.visualcapitalist.com/ranked-the-largest-sp-500-c...

onlyrealcuzzo6• 12 months ago

There was always some subset of the S&P that mattered way more than the rest, just like the S&P matters way more than the Russel.

Typically, you probably need to go down to the S&P 25 rather than the S&P 10.

k-i-r-t-h-i• 12 months ago

power law explains the distribution, but the distribution is getting more extreme over the years, likely due to (market structure, macro conditions, tech economics, etc)

atleastoptimal• 12 months ago

People have more room to buy more digital goods. There's far less room to buy more physical goods. People aren't going to double their stomach size to eat more McDonalds, but there's no limit to how much more data, software or AI tokens a person could require.

snitzr• 12 months ago

Billion-dollar Clippy.

PessimalDecimal• 12 months ago

Trillions, right?

freetonik• 12 months ago

Interesting that the profits of those bottom 490 companies of S&P 500 do not rise with the help of AI technology, which is supposedly sold to them at a reduced rate as AI vendors are bleeding money.

roncesvalles• 12 months ago

Other than NVIDIA, the profits of the S&P 10 haven't risen either. It's just that the market is pricing them very optimistically.

IMO this is an extremely scary situation in the stock market. The AI bubble burst is going to be more painful than the Dotcom bubble burst. Note that an "AI bubble burst" doesn't necessitate a belief that AI is "useless" -- the Internet wasn't useless and the Dotcom burst still happened. The market can crash when it froths up too early even though the optimistic hypotheses driving the froth actually do come true eventually.

Workaccount23• 12 months ago

We are still in the "land grab" phase where companies are offering generous AI plans to capture users.

Once users get hooked on AI and it becomes an indispensable companion for doing whatever, these companies will start charging the true cost of using these models.

It would not be surprising if the $20 plans of today are actually just introductory rate $70 plans.

andsoitis5• 12 months ago

Other than NVIDIA, the profits of the S&P 10 haven't risen either.

That’s not correct. Did you mean something else?

jayd165• 12 months ago

I'm curious to see the bubble burst. I personally don't think it will be anything like the dotcom era.

The benefits have just not been that wide ranging to the average person. Maybe I'm wrong but, I don't AI hype as a cornerstone of US jobs, so there's no jobs to suddenly dry up. The big companies are still flush with cash on hand, aren't they?

If/when the fad dies I'd think it would die with a wimper.

onlyrealcuzzo6• 12 months ago

We'll never know what would've happened without AI.

1. There profits could otherwise be down.

2. The plan might be to invest a bunch up front in severance and AI Integration that is supposed to pay off in the future.

3. In the future that may or may not happen, and it'll be hard to tell, because it may pay off at the same time an otherwise recession is hitting, which smoothes it out.

It's almost as if it's not that simple.

andsoitis5• 12 months ago

Artificial intelligence has a few simple ingredients: computer chips, racks of servers in data centers, huge amounts of electricity, and networking and cooling systems that keep everything running without overheating.

What about the software? What about the data? What about the models?

Steve163842• 12 months ago

What about the data?

What, you want them to pay for all the copyrighted data they consume and regurgitate? They can't afford that!

dsign2• 12 months ago

I don't think AI is having much impact on the bits of the economy that have to do with labor and consumption. Folk who are getting displaced by AI are, for now, probably being re-hired to fix AI mess-ups later.

But if, or when AI gets a little better, then we will start to see a much more pronounced impact. The thing competent AIs will do is to super-charge the rate at which profits don't go to labor nor to social security, and this time they will have a legit reason: "you really didn't use any humans to pave the roads that my autonomous trucks use. Why should I pay for medical expenses for the humans, and generally for the well-being of their pesky flesh? You want to shutdown our digital CEO? You first need to break through our lines of (digital) lawyers and ChatGPT-dependent bought politicians."

tehjoker• 12 months ago

Well, if you don't use humans, then you're using machine labor that will drive prices down to at-cost in a competitive environment and strip profits in the end. Profits come from underpaying labor.

OtherShrezzing• 12 months ago

Without AI, US economic growth would be meager.

The assumption here is that, without AI, none of that capital would have been deployed anywhere. That intuitively doesn't sound realistic. The article follows on with:

In the last two years, about 60 percent of the stock market’s growth has come from AI-related companies, such as Microsoft, Nvidia, and Meta.

Which is a statement that's been broadly true since 2020, long before ChatGPT started the current boom. We had the Magnificent Seven, and before that the FAANG group. The US stock market has been tightly concentrated around a few small groups for a decades now.

You see it in the business data. According to Stripe, firms that self-describe as “AI companies” are dominating revenue growth on the platform, and they’re far surpassing the growth rate of any other group.

The current Venn Diagram of "startups" and "AI companies" is two mostly concentric circles. Again, you could have written the following statement at any time in the last four decades:

According to [datasource], firms that self-describe as “startups” are dominating revenue growth on the platform, and they’re far surpassing the growth rate of any other group.

biophysboy4• 12 months ago

Its also not fair to compare AI firms with others using growth because AI is a novel technology. Why would there be explosive growth in rideshare apps when its a mature niche with established incumbents?

dragontamer3• 12 months ago

I think the explosive growth that people want is in manufacturing. Ex: US screws, bolts, rivets, dies, pcbs, assembly and such.

The dollars are being diverted elsewhere.

Intel a chip maker who can directly serve the AI boom, has failed to deploy its 2nm or 1.8nm fabs and instead written them off. The next generation fabs are failing. So even as AI gets a lot of dollars it doesn't seem to be going to the correct places.

thrance• 12 months ago

> Without AI, US economic growth would be meager.
The assumption here is that, without AI, none of that capital would have been deployed anywhere. That intuitively doesn't sound realistic.

That's the really damning thing about all of this, maybe all this capital could have been invested into actually growing the economy instead of fueling this speculation bubble that will burst sooner or later, bringing along any illusion of growth into its demise.

ryandrake• 12 months ago

Or all that money might have been churning around chasing other speculative technologies. Or it might have been sitting in US Treasuries making 5% waiting for something promising. Who knows what is happening in the parallel alternate universe? Right now, it feels like everyone is just spamming dollars and hoping that AI actually becomes a big industry, to justify all of this economic activity. I'm reminded of Danny DeVito's character's speech in the movie Other People's Money, after the company's president made an impassioned speech about why its investors should keep investing:

"Amen. And amen. And amen. You have to forgive me. I'm not familiar with the local custom. Where I come from, you always say "Amen" after you hear a prayer. Because that's what you just heard - a prayer."

At this point, everyone is just praying that AI ends up a net positive, rather than bursting and plunging the world into a 5+ year recession.

justonceokay2• 12 months ago

If the economy in my life has taught me anything, it’s that there will always be another bubble. The Innovators Dilemma mentions that bubbles aren’t even a bad thing in the sense that useful technologies are often made during them, it’s just that the market is messy and lots of people end up invested in the bubble. It’s the “throw spaghetti at the wall” approach to market growth. Not too different than evolution, in which most mutations are useless but all mutations have the potential to be transformative.

onlyrealcuzzo6• 12 months ago

1. People aren't going to take on risk and deploy capital if they can't get a return.

2. If people think they can get an abnormally high return, they will invest more than otherwise.

3. Whatever other money would've got invested would've gone wherever it could've gotten the highest returns, which is unlikely to have the same ratio as US AI investments - the big tech companies did share repurchases for a decade because they didn't have any more R&D to invest in (according to their shareholders).

So while it's unlikely the US would've had $0 investment if not for AI, it's probably even less likely we would've had just as much investment.

jayd165• 12 months ago

Why is it "unlikely" that the alternative is not US investment by these US companies?

The big US software firms have the cash and they would invest in whatever the market fad is, and thus, bring it into the US economy.

jlarocco• 12 months ago

it's probably even less likely we would've had just as much investment.

I doubt it. Investors aren't going to just sit on money and let it lose value to inflation.

On the other hand, you could claim non-AI companies wouldn't start a new bubble, so there'd be fewer returns to reinvest, and that might be true, but it's kind of circular.

metalliqaz2• 12 months ago

1. People aren't going to take on risk and deploy capital if they can't get a return.

This doesn't seem to align with the behavior I've observed in modern VCs. It truly amazes me the kind of money that gets deployed into silly things that are long shots at best.

bayarearefugee• 12 months ago

1. People aren't going to take on risk and deploy capital if they can't get a return.
2. If people think they can get an abnormally high return, they will invest more than otherwise.

Sounds like a good argument for wealth taxes to limit this natural hoarding of wealth absent unreasonably good returns.

whoknowsidont• 12 months ago

1. People aren't going to take on risk and deploy capital if they can't get a return.

I'm sorry, what? This has happened all the time, and in increasing volleys, since 2008.

sgt101• 12 months ago

A big dynamic in this is that the business models for two of the biggest players are at stake.

Google is facing a significant danger that its search advertising business is going to just disappear. If people are using AI to find stuff or get recommendations then then aren't using google. Why should a photographer continue to spend $200 a month on ads when their clients are coming from openAI? Especially when OpenAI is using the organic search results. Meta is facing the same sort of issue, if the eyeballs aren't on insta then the ad $$$ go somewhere else.

So if they have money, and can get money, they will invest it to protect their businesses - all of it.

MS and VC's are doing the opposite, they are investing with the idea of taking the attention that Google and Meta currently have, but they are also following the "I'm scared" signal that Google and Meta have put in the market.

thiago_fm• 12 months ago

I agree, in any time in US history there has always been those 5-10 companies leading the economic progress.

This is very common, and this happens in literally every country.

But their CAPEX would be much smaller, as if you look at current CAPEX from Big Tech, most of it are from NVidia GPUs.

If a Bubble is happening, when it pops, the depreciation applied to all that NVidia hardware will absolute melt the balance sheet and earnings of all Cloud companies, or companies building their own Data centers like Meta and X.ai

rwmj2• 12 months ago

And NVidia don't even make the GPUs! They're all made on a brave little island off the coast of China, while 7 huge US companies shuffle them around and exchange vast amounts of money for them.

rsanek• 12 months ago

assumption here is that, without AI, none of that capital would have been deployed anywhere. That intuitively doesn't sound realistic

For the longest time, capex at FAANG was quite low. These companies are clearly responding specifically to AI. I don't think it's realistic to expect that they would raise capex for no reason.

a statement that's been broadly true since 2020, long before ChatGPT started the current boom

I guess it depends on your definition of "long before," but the ChatGPT release is about mid-way between now and 2020.

As for the startup vs. AI company point, have you read Stripe's whitepaper on this? They go into detail on how it seems like AI companies are indeed a different breed. https://stripe.com/en-br/lp/indexingai

trod1234• 12 months ago

The sunsetting of research tax breaks would explain why they threw everything into this.

They also view labor as a replaceable cost, as most accountant based companies that no longer innovate act. They forget that if you don't hire people, and pay people, you don't have any sales demand and this grows worse as the overall concentration or intensity of money in few hands grows. Most AI companies are funded on extreme leverage from banks that are money-printing and this coincided with 2020 where the deposit requirements were set to 0% effectively removing fractional banking as a system.

conductr• 12 months ago

I think the money is chasing growth in a market that is mostly mature. Tech is really the only hope in that situation, so that's where the dollars land.

camgunz3• 12 months ago

I think it's more likely the assumption is you'd expect a far more diversified market. If we're really in a situation where the rational, good reasons move is to effectively ignore 98% of companies, that doesn't say good things about our economy (verging on some kind of technostate). You get into weird effects like "why invest in other companies" leading to "why start a company that will just get ignored" leading to even more consolidation and less dynamism.

istjohn2• 12 months ago

But startups are thriving. That doesn't suggest decreasing dynamism to me. It suggests that there are abundant gains to be had by exploiting technological progress, and the legacy economy is not availing itself of these opportunities. A thriving tech startup sector is surely key to dynamism.

johnnienaked2• 12 months ago

The capital would have, should have maybe, been deployed in stock buybacks and dividends. Investment already doesn't happen in the US, and that's an expected thing for a well-developed country with a trade deficit.

golly_ned• 12 months ago

Derek Thompson is not well suited to this kind of work. He is much better suited to his usual lame, predictable, centrist, lukewarm commentary on tired political and social topics.

johnnyanmac4• 12 months ago

The assumption here is that, without AI, none of that capital would have been deployed anywhere.

With this recessionary behavior, it might not be that far fetched an assumption. I'm not sure where else it would flow outside of being hoarded up in assets if there wasn't this big speculation everyone wants to take advantadge of. Especially when you consider that there's so much money flowing into AI, but AI is not as of yet profitable.

AlecSchueler• 12 months ago

That's it, it's unusual times and the AI boom could be covering things up. I personally completely divested from US businesses since the start of the year but I can understand the allure of wanting to ride that particular gravy train

pj_mukh• 12 months ago

“The assumption here is that, without AI, none of that capital would have been deployed anywhere.”

The question is could the capital be deployed somewhere more productive. If that was just used park it into REIT’s buying single family homes, would you be happier?

SideburnsOfDoom• 12 months ago

It depends on what "more productive" means to them. e.g. pyramid schemes are very productive if you exit early enough.

vannevar• 12 months ago

Nobody can say for sure whether the AI boom is evidence of the next Industrial Revolution or the next big bubble.

Like the Internet boom, it's both. The rosy predictions of the dotcom era eventually came true. But they did not come true fast enough to avoid the dotcom bust. And so it will be with AI.

GoatInGrey3• 12 months ago

My suspicion is that there's a there there, but it doesn't align with the predictions. This is supported by the tension between AI doom articles and the leading models experiencing diminishing performance gains while remaining error-prone. This is to speak nothing of the apparent LLM convergence limit of a ketamine-addled junior developer. Which is a boundary the models seem destined to approach indefinitely without ever breaching.

The "bust" in this scenario would hit the valuations (P/E ratio) of both the labs and their enterprise customers, and AI businesses dependant on exponential cost/performance growth curves with the models. The correction would shake the dummies (poorly capitalized or scoped businesses) out of the tree, leaving only the viable business and pricing models still standing.

That's my personal prediction as of writing.

owebmaster• 11 months ago

Although that is true, the big players just have enough money to go through it this time.

bravetraveler2• 12 months ago

They mention rate of adoption, compared to the internet. Consider the barriers to entry. Before we all got sick of receiving AOL CDs, the prospect of 'going online' was incredibly expensive and sometimes laborious.

More people subscribe to/play with a $20/m service than own/admin state-of-the-art machines?! Say it ain't so /s

thewebguyd2• 12 months ago

More people subscribe to/play with a $20/m service than own/admin state-of-the-art machines?! Say it ain't so /s

The problem is, $20/m isn't going to be profitable without better hardware, or more optimized models. Even the $200/month plan isn't making money for OpenAI. These companies are still in the "sell at a loss to capture marketshare" stage.

We don't even know if being an "AI Company" is viable in the first place - just developing models and selling access. Models will become a commodity, and if hardware costs ever come down, open models will win.

What happens when OpenAI, Anthropic, etc. can't be profitable without charging a price that consumers won't/can't afford to pay?

bravetraveler2• 12 months ago

Pop :) For some, at least. Barons tend to leave remnants

hackable_sand2• 12 months ago

What about food and housing? Why can't America invest in food and housing instead?

margalabargala5• 12 months ago

America has spent a century investing in food. We invested in food so hard we now have to pay farmers not to grow things, because otherwise the price crash would cause problems. Food in America is very cheap.

hackable_sand2• 12 months ago

It's reassuring to be reminded that every child in America must justify their existence or starve to death.

jimt1234• 12 months ago

Food in America is very cheap.

And that's a whole different problem. Cheap != inexpensive.

Steve163842• 12 months ago

Ironically, this is currently second in HN: https://news.ycombinator.com/item?id=44836219 ("Food, housing, & health care costs are a source of major stress for many people")

righthand2• 12 months ago

Is anyone starving in America? Why would there need to be focus on food production? We have huge food commodities.

vdupras• 12 months ago

Let them eat cake, right?

https://www.ers.usda.gov/topics/food-nutrition-assistance/fo...

GoatInGrey3• 12 months ago

Because investing in housing means actually changing things. There's a "Don't just do something, stand there!" strategy of maximizing comfort and minimizing effort, that must be overcome.

daedrdev• 12 months ago

the US systematically taxes and forbids new housing in many ways as local voters desire. Setback requirements, 100K+ hookup costs, stairway standards, density limits, parking minimums and regulations, community input, allowing rejection of new housing despite it following all rules, abuse of environmental regulations (which ends up hurting the environment by blocking density), affordable housing requirements (a tax on each new housing block to fund affordable units on the side) all prevent new housing form being built.

booleandilemma• 12 months ago

Having food and housing would make the population too comfortable. They want everyone a little on edge. You don't work, you don't eat.

biophysboy4• 12 months ago

“The top 100 AI companies on Stripe achieved annualized revenues of $1 million in a median period of just 11.5 months—four months ahead of the fastest-growing SaaS companies.”

This chart is extremely sparse and very confusing. Why not just plot a random sample of firms from both industries?

I'd be curious to see the shape of the annualized revenue distribution after a fixed time duration for SaaS and AI firms. Then I could judge whether its fair to filter by the top 100. Maybe AI has a rapid decay rate at low annualized revenue values but a slower decay rate at higher values, when compared to SaaS. Considering that AI has higher marginal costs and thus a larger price of entry, this seems plausible to me. If this is the case, this chart is cherry picking.

doyouevensunbro• 12 months ago

because the oligarchs demanded it

There, summed it up for you.

voidhorse• 12 months ago

Not sure why you were down voted. It's true and it actually answers the question in the article title, which the article doesn't even explore, let alone answer.

ChrisArchitect• 12 months ago

AI is propping up the US economy

https://news.ycombinator.com/item?id=44802916

GoatInGrey3• 12 months ago

I'm noticing how that article is myopically discussing equity valuations rather than actual economic output and worker productivity.

krunck• 12 months ago

Please stop using stacked bar charts where individual lines(plus a Total) line would help the poor reader comprehend the data better.

GolfPopper• 12 months ago

Remember, the appropriate way to parse use of "the economy" in popular press is to read it as "rich peoples' yatch money".

csours• 12 months ago

Nobody can say for sure whether the AI boom is evidence of the next Industrial Revolution or the next big bubble. All we know is that it’s happening.

In hindsight, it will be clear, and future generations (if any exist) will ask: "Why didn't you understand what was happening at the time"

My answer: Noise. Just because you can find someone who wrote down the answer at the time, doesn't mean that they really understood the answer, at least not to the extent that we will understand with hindsight.

Future history is contingent.

jimmydoe• 12 months ago

This matches the tech job market: if you are not in top corp or labs, your hard work is most likely subsidizing the $ 1.5M paycheck for OpenAI employees.

maerF0x0• 12 months ago

“They’re generating unprecedented amounts of free cash flow,” Cembalest told me. “They make oodles and oodles of money, which is why they can afford to be pouring hundreds of billions of dollars of capital spending each year into AI-related R&D and infrastructure.”

IMO this should be a trigger for investors that they have not been receiving their profits, and instead the profits are being dumped into CEOs next big bets that will fuel their stockbased compensation gains. Also to blame is the government's culpability here for creating both tax incentives and a lack of laws that say profits must be returned as dividends (they can always be DRIP'd back into the company as new shares if desired, it's absurd to say its better for investors, when the alternative actually gives more choice).

jameslk• 12 months ago

I found this analysis insightful:

https://x.com/dampedspring/status/1953070287093731685

However, this pace is likely unsustainable going forward. The sharp acceleration in capex is likely behind us, and the recent growth rate may not be maintained. Any sustained weakness in final demand will almost certainly affect future investment, as AI demand ultimately depends on business revenues and profits, which are tied to nominal GDP. Realized and forecasted capex remain elevated, while free cash flow and cash and cash equivalents are declining for hyperscalers.

ThinkBeat• 12 months ago

The fact that there is massive spending in AI in the tech sector, isn't that just a -possible- sign of a other bust coming down the road?

We have seen it before, again and again.

johnnienaked2• 12 months ago

All I see is B-U-B-B-L-E

knowaveragejoe• 12 months ago

METR also conducted an in-depth study that asked experienced developers to code with a popular AI assistant. After they finished their tasks, the developers claimed that using the AI had made them 20 percent more productive. But independent evaluators in the study actually concluded that using AI did the opposite: it increased task completion time by about 20 percent.

This somewhat reflects my experience... I can totally see the back-and-forth dance taking longer in some cases.

But I also think there is more than efficiency being unlocked here. Sure, a developer might have cranked out a rough MVP in less time, but with this they're also often continuously updating a README, tests and other infrastructure as they go.

One could argue about whether that's its own footgun - relying on tests that don't really test what they should, and let critical bugs through later.

throwmeaway222• 12 months ago

AI bills are in the thousands of dollars per day for some small AI startups, and in the 10-100k range per day for large companies.

So as many people have speculated that OpenAI has no moat - are wrong.

jjtheblunt• 12 months ago

What is the moat you are implying? (affordability?)

antonvs• 12 months ago

A moat is protection from competition. What you’ve described is not a moat.

scotty79• 12 months ago

Measuring such things in dollars rubs me the wrong way. Is using inflation adjusted dollar common or do people just compare nominal values?

Animats2• 12 months ago

See "Spending on AI data centers is so massive that it’s taken a bigger chunk of GDP growth than shopping—and it could crash the American economy"[1] The US is having a giant AI boom and a recession in the rest of the economy.

Historically, the classic example of this is railway mania of the mid 19th century.[2] That started in 1830, with the Liverpool and Manchester Railway.[3] This was when the industrial revolution got out of beta. There were earlier railroads, but with dual tracks, stations, signals, tickets, schedules, and reasonably good steam engines, the Liverpool and Manchester worked like a real service. It was profitable. Then lots of others started building and over-building railroads, with varying degrees of success. See Panic of 1847.

It was really too early for good railroads. Volume production of steel didn't exist. Early railroads were built with wood and iron, not very well. Around 1880-1900, everything was rebuilt in steel, and got bigger, better, and safer.

Consider the early Internet. We had TCP/IP working across the US in the early 1980s, but it wasn't a big deal commercially for another 10-15 years, it wasn't everywhere until the early 2000s, and it wasn't out of bubble mode until 2010 or so.

[1] https://fortune.com/2025/08/06/data-center-artificial-intell...

[2] https://en.wikipedia.org/wiki/Railway_Mania

[3] https://youtu.be/pDEnsraYx3k?t=505

jcgrillo2• 12 months ago

The really scary thing is all the value in the AI boom is predicated on the belief that the technology is "early" and that it will improve over time. We're seeing the opposite. Instead, all the competing models are basically converging on the same benchmark performance numbers, as we saw yesterday with the gpt-5 debacle. This suggests that performance is actually topping out, which makes intuitive sense if advancements in LLMs are proportional to their training data. They've already used up all the data. So it very well could be what we see right now is basically as good as it gets, or at least approximately so. The market is not ready for that.

Animats2• 12 months ago

We're seeing the opposite. Instead, all the competing models are basically converging on the same benchmark performance numbers...

That question is really important. Comments on that? Maybe LLMs are asymptotically approaching some fundamental limit for that technology.

What goes on inside an LLM hasn't changed much in years. More data and more compute is thrown at that little kernel that makes it all go. Additional gimmicks are added around the core algorithm. But there's not much progress down at the bottom. There's no reaching into the net and extracting a reliable "don't know". Is there an upper limit to what that architecture can do? AGI may not be reachable by this route.

Maybe someone will find a way around that limit. The field has way too much money, too much visibility, and too many people. The technology works, although it has limits.

cess11• 12 months ago

On the face of it, it looks positively habsburgian.

sdsd• 12 months ago

One argument I often hear in my corner of the internet is that AI is one of the only remaining sectors of unregulated, meritocratic activity. The smart people have been pushed out of the real economy, so we're congregating here.

EXIT blog talks about this a lot. Eg in this article, "Sanctuary Cities are Special Economic Zones - Just Not For You": https://blog.exitgroup.us/p/sanctuary-cities-are-special-eco...

Anyone else knows fun writing like this, please recommend it to me

sheepybloke• 12 months ago

The take away from the teacher graphs were very different than my take aways. The author was bullish because ~55% used AI to help prep lessons. However, looking at the graph, I would be very bearish about the adoption of AI. Most of the graphs show that a vast majority of teachers aren't using AI. This leads me to see a huge discrepancy between real world usage and the amount of money that's being invested.

BrenBarn2• 12 months ago

AI didn't conquer the US economy. What conquered the US economy is a small number of large companies. If Apple started rolling out an iTurd app, and Google started offering gTurd, and Amazon started shipping free turds to everyone's doorstep, and so on, then we would be talking about how turds have conquered the US economy. The economy has become divorced from any pretense of meeting actual consumer needs. The suppliers are creating their own demand.

How AI conquered the US economy: A visual FAQ

Related Stories