I have said it before but AI is not the internet, it won't even come close.
The internet doubled economic activity in the USA alone. It created new marketplaces to sell, new tools, online banking, online stock investing, money transfers, crypto, video games, software, accounting software, social media, advertising, streaming services, peer to peer sharing, music stores, cloud computing ...etc.
It created an entire world out of thin air. That is the internet and that is why it was so transformative. It transformed the USA from a manufacturing economy to a services economy.
So what is AI? Right now....it is just being used as the Internet +. If the Internet is the "Browser", AI is just the "add-on".
But the goal is AI is suppose to replace all human work in the entire world....but now the question is how? There is no proof in the concept and as we reach year 5 of consumer and corporate use and over 10 years of development, it isn't even close to the goals envisioned. The internet showed us webpages, online marketplaces and search engines by this time.
AI has shown us that it is mostly used as a better search engine...for consumer or for work. That alone is value, and gives the ability to effectively do tasks faster....but is the cost worth it? It does automate some white collar jobs mostly in coding. But they were never a "massive" workforce. Did you spend a trillion dollars to save $10 million?
For that efficiency is AI worth $10 Trillion dollars in infrastructure spending by 2030+? The internet when adjusted for inflation cost $1 Trillion dollars for the multi-year telecom network layout. Is AI creation worth 10x more than the internet creation in spending? This is what investors are trying to decide.
As for the tech bubble, the AI bubble will rationalize the market, a lot of virtual value will evaporate, but the infrastructure will stay. Difference is that an optic fiber is not obsolete after 4 years... We'll see what happen. But this correction is necessary to put logic back into this mess.
Agree. AI is powerful and helpful, but we literally don't know how much it costs to run. The hyperscalers won't tell anyone. From all we can gather it seems extremely expensive to run, so power hungry, and all the datacenters are being financed at extremely high interest rates. The internet might have had this issue but on a much smaller and cheaper scale. Like AI is hundreds of times more expensive to run than a simple website. AI will stay but probably only for workloads where it makes financial sense.
I'd add that the early internet also had easy and cheap options to access it for users.
AI has what looks like insane subsidisation to mask costs. Yes local models exist but you need expensive high cost hardware to get anything out of those.
You can run AI on your home computer, it's just not as good as the frontier models. But it's still pretty good. They are using mega-amounts of power due to the law of diminishing returns: a model that uses twice as much power is less than twice as good.
I think you have a strong argument here but you misunderstand what google and openai think they're building.
The AI companies believe that they can build something akin to god-like intelligence. The whole premise here is that being the first company to do so will let them extract a rent on people using it, and for them to keep all of the discoveries and research it does in medicine, computer science, electrical engineering, mechanical engineering, materials science, algorithms, physics, etc for themselves to make products so dramatically better that no one can compete and they absorb all profit in the market.
Now is this a long term business strategy? No. Is it what I think they want to do? Yes.
If you are going to have an analysis of the AI expenditure the end goal is not, we have enough robots and datacenters to automate all human labor. The end goal is, which person's capital owns a piece of god and how do I get there first to extract profits from my share of it.
I know this is kinda bullshit and handwavy, but I really think this is what people in San Francisco believe in and to think of this in an economy is kinda a misstep. If you believe it (super intelligence) can happen you have every incentive to put all of the capital you own into it as soon as possible. That's why the expenditure is high.
This is also why there is so much focus on mathematical proof and coding and reasoning. As others have said, the at home models are already basically good enough, so why focus on research with no economic value like mathematics? Because optimization for building god-like intelligence is the end goal.
It is less economics and more Eschatology that you have to work on here.
Not my own opinion just giving an explanation for why 50 trillion dollars in compute expenditure would be rational under certain beliefs.
Developed countries offloaded manual labor markets quite a while ago and that had massive economic impacts.
They switched largely to intellectual labor to sustain growing economies. For which companies are now actively looking for ways to offload intellectual labor onto computers.
It won't take a massive shift in job markets to see other, unrelated, areas affected by decreased labor value. Whether that's manual or intellectual labor. People shifting away from computer coding to ,,, whatever it is they will, will just decrease the value of work in those fields. Nothing good will come from this
The other thing is that if the main benefit of AI is savings through job replacement, what happens when there’s way fewer consumers to buy stuff? CEO like to talk about efficiency and cost savings, but the economy actually runs the best when people are working, making a decent living, and can spend because they aren’t worried about having to save a ton in case they lose their job and never get another one. Essentially they’ll turn everyone in their prime earning years into penny pinching pensioners.
The tech billionaires and the CEOs aren’t the same people imo.
Tech people simply think they will create and we will consume it. Apple hasn’t released a new product since the Apple Watch. Just add-ons.
The CEO people only see value in relation to how cheap it is. They jumped at the thought of replicating 100 workers with cheap Ai. But once Ai becomes too expensive they will leave it for the cheaper option
I’d challenge the idea that this is just internet+ and I think you know that. Every day I’m figuring out how to use the tool better to get my work done. The agent can write code and interact directly with APIs on my behalf, much faster than I can. So that means it’s being used for increasing work output in less time. And you’re also probably well aware that it can write a whole lot of content to varying degrees. For a lot of business content it’s at the point of good enough.
“write code and interact with APIs on my behalf” is basically what average webdev software engineers have been doing for the past two decades.
They started on the higher value tasks, and now with AI you can do the lower value tasks that weren’t worth paying even a mid-to-bottom rung software engineer to do.
That’s not world changing. That’s leftover crumbs.
Sure, it’s useful. But it’s mostly stuff that people with a little more computer and programming knowledge had already been automating for themselves for decades now.
Sally in HR being able to do a few more things on her computer without help from IT now is not at the same scale of world changing impact as the internet.
That's because it's close to free right now. How much would you be willing to pay for AI output? Like what is the max? Because we literally don't know what the real cost is.
Remember when DoorDash was cheap during covid? Everyone used it. It was cheap because they were operating at a loss. Now they make a profit but the prices are so expensive hardly anyone uses it.
Of course everyone is using AI now because it's literally free.
AI is going beyond just being profitable though. It's a national goal now - and even if investors aren't stepping in, government will step in with the funding.
The one benefit of AI I think is less appreciated is how much it will help smaller firms and developers.
A large firm has the capacity to hire 1000 low level workers to do the grunt work. They replace them with AI and it's not great for jobs. But a small firm can now for the same cost also have the same level of work force.
That will allow them to take on more jobs. Provide greater levels of competition.
Same for development. A small open source project just doesn't have the man power to really fix issues quickly and streamline processes. But now a lot of the boring grunt work is done for you and you can focus on features. Even stuff you might not be very good at you can try and if it kind of works but is janky you can make a more educated decision on finding someone more knowledgeable to do it better. Because you already know the worst version of it works.
But making small firms and developers more competitive is probably not in the best interests of these giants so I think we will see a lock down or change in access capabilities for these kinds of projects.
Imagine someone uses Gemini to make a YouTube or google maps competitor. Why would google allow that.
>Same for development. A small open source project just doesn't have the man power to really fix issues quickly and streamline processes. But now a lot of the boring grunt work is done for you and you can focus on features. Even stuff you might not be very good at you can try and if it kind of works but is janky you can make a more educated decision on finding someone more knowledgeable to do it better. Because you already know the worst version of it works.
Actually that is something I never thought off. I kind of hope that all the people who got laid off from AI, learn to utilize AI to sweep the product from under the Big Tech companies.
I have seen what photonshop and other free competitors did to photoshop and it is pretty amazing that for 80% of people who use photoshop, may never even notice a change....and the product is free. Same with people merging games together.
If the people who suffered from AI the most, are able to turn a small start up with limited manpower into a multi-billion dollar company with AI, then AI has shown real potential and ironically has cannibalized the value of what these large companies are creating.
Lol Ai is so much more than the “browser + add on.” I was a skeptic too, but Claude has been able to do real work and analysis for me at what I estimate is 1/12th the time.
Example: Ask a 4 person tech team to create a tableau dashboard measuring and tracking KPIs for my business - maybe a year and it will just be data visualization. Claude can knock this out and perfect it in about a month of vibe coding. Also the Claude version will be 100% interactive and present the data just how you want it and can pull from data and text sources (notes). Furthermore, if you don’t like something you can just ask it to fix it and in 40 seconds it’s fixed/rearranged/etc.
Another example: combine the graphics and cad drawings from 12+ reports into one interactive visual. May take a professional CAD engineer two weeks to get to 75% of the way there. Claude does it nearly perfectly within an hour of vibe coding.
>Right now....it is just being used as the Internet +. If the Internet is the "Browser", AI is just the "add-on"....AI has shown us that it is mostly used as a better search engine
I'm not challenging your longterm hypothesis, but the S&P 500 is still up 4% over the past six months. Maybe there will be a global economic reckoning, but at least so far this isn't it.
Yes but the real question is less about how well the indexes have done and more about expectations of future performance. The trailing 6 month average is rapidly degrading as the run up from April-May is shed.
No, investment has been driving the US economy. Investment that would have been invested elsewhere within the economy had AI not been a thing. GDP growth was 3.4% annualized in Q3. Is that close to a recession? I don't think so.
So it's just that the labor market is less dynamic than it was between 2021 - 2023 from the post-Covid hiring sprees. In fact, by some metrics, such as prime age labor force participation, labour participation has risen significantly since Covid and remained mostly stable over the last 3 years. I mean, layoffs have been happening in tech and finance, that much is true, but not at the scale required to lead to an economic contraction.
>The NBER's traditional definition of a recession is that it is a significant decline in economic activity that is spread across the economy and that lasts more than a few months. The committee's view is that while each of the three criteria—depth, diffusion, and duration—needs to be met individually to some degree, extreme conditions revealed by one criterion may partially offset weaker indications from another. For example, in the case of the February 2020 peak in economic activity, we concluded that the drop in activity had been so great and so widely diffused throughout the economy that the downturn should be classified as a recession even if it proved to be quite brief.
I've had a problem with that definition for a few reasons: 1) we've had ongoing inflation since COVID for various reasons, including climate change, corporate greed and the decline of 3rd spaces. There's a lot of FOMO when you're not sure prices will be the same today as tomorrow. 2) someone above seemed to say that A.I. is pretty much keeping the economy afloat (as companies threaten widespread and long lasting unemployment no less) 3) Enshittification, requiring people to spend more money more often to get the same amount of quality they would have gotten 10 years ago
Also the labor market is by all means abysmal, even for the highly skilled, connected people the U.S. claims to prize. I'm willing to accept that we're willing to accept a slow death to quality of goods and an increase in corporate scamming. But the economy is so bad that you have pundits for respected publications (The Atlantic, WSJ) playing damage control by accusing people of lying about it.
I guess a recession isn't useful to describe relevant economic conditions for anyone under upper-middle class anymore.
I think I'll stick to trusting the NBER, who are one of the most reputable economic research organisations in the world, on their definition of recession.
> 1) we've had ongoing inflation since COVID for various reasons, including climate change, corporate greed and the decline of 3rd spaces.
You are listing irrelevant things. GDP is inflation adjusted. A recession is a technical term, very clearly defined by the NBER.
> 2) someone above seemed to say that A.I. is pretty much keeping the economy afloat
Investment, not AI. Money is being spent on investment - money which would still be out there had AI never existed. It's also still real investment, even if it is AI dominated.
> 3) Enshittification, requiring people to spend more money more often to get the same amount of quality they would have gotten 10 years ago
This is why you have quality adjustments in inflation baskets. (Again, generally not relevant to measuring a recession).
https://www.bls.gov/cpi/quality-adjustment/
It's highly detailed and regularly reviewed so that quality is controlled for in economic data. (And again, generally irrelevant to how we measure a recession).
>Also the labor market is by all means abysmal, even for the highly skilled, connected people the U.S. claims to prize.
Did you read the research I linked? Or even taken a look at prime age labor force participation rate:
https://fred.stlouisfed.org/series/lns11300060
>By combining these various dimensions of fluidity, Molloy et al. (2016) date the beginning of the downtrend to the early 1980s, possibly before. They, along with other researchers, have found that the decline [in fluidity] has been broad based across types of workers and sectors and that perhaps half of the general decline in fluidity can be attributed to demographic factors, notably the aging of the population and increases in levels of education.
>I guess a recession isn't useful to describe relevant economic conditions for anyone under upper-middle class anymore.
It's a useful metric to describe a recession. It was never ever intended to describe the economic conditions of a particular decile of the economy. What you said there is also just nonsense.
Unemployment numbers are falling because people have stopped looking for jobs plus this administration has been falsifying numbers on release with revisions months later.
I haven't trusted unemployment numbers since 2022. I don't have many numbers, but a lot of growth has been concentrated in white collar jobs in the tech industry and maybe medical industry. Since late 2022, a lot of previously-secure tech companies like Google, Meta and Microsoft have been laying off their employees in waves. The job market is so bad that common advice is 1) don't leave your job unless you have one lined up and 2) nepotism is the best way to get a job. Hell, Microsoft just offered voluntary exit packages to 60000 of its employees!
Even this article here admits that not a lot of companies are hiring. Just ignore the bullshit about "Americans enjoying job security".
Was with you until this, which makes the comment feel like fantasy. There is no housing bubble. The country is just severely, severely supply-constrained from decades of underbuilding and obstructionism. The vast majority of homeowners have rates that we will probably not see for the rest of our lifetime, and will simply not sell.
I can't speak for him, but the concern arises when folks are forced to sell. Mass layoffs, need to move for a job, etc. We will find out what people do when we have significant forced relocation. Keep the house, attempt to rent while you rent? Take on two mortgages? what's the play? There's going to be some real losers.
Imagine thinking anything that has happened over a 1-2 hour span is enough to validate an opinion on whether or not we're seeing a correction or a pop.
This has got to be the strangest bot/user behavior i've seen on this sub.
It's just anecdotal, but this sort of commentary makes me even more worried. I commented yesterday about the the very similar sounding pro tech comments coming in on every story about a tech company's equity losses. To see this continue today in a different flavor is not great.
I'm a Boglehead. I don't usually follow the daily market changes. Because every day there's a new price and a lot of discussion about it and most of the time it amounts to nothing. The market going down a bit then stabilizing isn't anything especially interesting.
I’m not an options trader and never will be. But I’m majorly hedging out my long term plays with options on volatility. A lot of idiots are in the market ready to fire off sells
Questions EXISTENCE of returns. There is no evidence of all the supposed amazing amounts of money that AI has saved or earned. It's all circular financing and managers blowing smoke to justify their herd-following decisions.
How about the increases in cloud computing revenue amd profits that GOOG, AMZN. and MSFT are seeing? We can't say this is all AI specific spending, but this is what the data centers are being built for.
PrinnyFriend | 14 hours ago
I have said it before but AI is not the internet, it won't even come close.
The internet doubled economic activity in the USA alone. It created new marketplaces to sell, new tools, online banking, online stock investing, money transfers, crypto, video games, software, accounting software, social media, advertising, streaming services, peer to peer sharing, music stores, cloud computing ...etc.
It created an entire world out of thin air. That is the internet and that is why it was so transformative. It transformed the USA from a manufacturing economy to a services economy.
So what is AI? Right now....it is just being used as the Internet +. If the Internet is the "Browser", AI is just the "add-on".
But the goal is AI is suppose to replace all human work in the entire world....but now the question is how? There is no proof in the concept and as we reach year 5 of consumer and corporate use and over 10 years of development, it isn't even close to the goals envisioned. The internet showed us webpages, online marketplaces and search engines by this time.
AI has shown us that it is mostly used as a better search engine...for consumer or for work. That alone is value, and gives the ability to effectively do tasks faster....but is the cost worth it? It does automate some white collar jobs mostly in coding. But they were never a "massive" workforce. Did you spend a trillion dollars to save $10 million?
For that efficiency is AI worth $10 Trillion dollars in infrastructure spending by 2030+? The internet when adjusted for inflation cost $1 Trillion dollars for the multi-year telecom network layout. Is AI creation worth 10x more than the internet creation in spending? This is what investors are trying to decide.
ebfortin | 13 hours ago
As for the tech bubble, the AI bubble will rationalize the market, a lot of virtual value will evaporate, but the infrastructure will stay. Difference is that an optic fiber is not obsolete after 4 years... We'll see what happen. But this correction is necessary to put logic back into this mess.
Internationallegs | 12 hours ago
Agree. AI is powerful and helpful, but we literally don't know how much it costs to run. The hyperscalers won't tell anyone. From all we can gather it seems extremely expensive to run, so power hungry, and all the datacenters are being financed at extremely high interest rates. The internet might have had this issue but on a much smaller and cheaper scale. Like AI is hundreds of times more expensive to run than a simple website. AI will stay but probably only for workloads where it makes financial sense.
mayhemski123 | 11 hours ago
I'd add that the early internet also had easy and cheap options to access it for users.
AI has what looks like insane subsidisation to mask costs. Yes local models exist but you need expensive high cost hardware to get anything out of those.
PatchyWhiskers | 10 hours ago
You can run AI on your home computer, it's just not as good as the frontier models. But it's still pretty good. They are using mega-amounts of power due to the law of diminishing returns: a model that uses twice as much power is less than twice as good.
Icy_Cover_2472 | 10 hours ago
I think you have a strong argument here but you misunderstand what google and openai think they're building.
The AI companies believe that they can build something akin to god-like intelligence. The whole premise here is that being the first company to do so will let them extract a rent on people using it, and for them to keep all of the discoveries and research it does in medicine, computer science, electrical engineering, mechanical engineering, materials science, algorithms, physics, etc for themselves to make products so dramatically better that no one can compete and they absorb all profit in the market.
Now is this a long term business strategy? No. Is it what I think they want to do? Yes.
If you are going to have an analysis of the AI expenditure the end goal is not, we have enough robots and datacenters to automate all human labor. The end goal is, which person's capital owns a piece of god and how do I get there first to extract profits from my share of it.
I know this is kinda bullshit and handwavy, but I really think this is what people in San Francisco believe in and to think of this in an economy is kinda a misstep. If you believe it (super intelligence) can happen you have every incentive to put all of the capital you own into it as soon as possible. That's why the expenditure is high.
This is also why there is so much focus on mathematical proof and coding and reasoning. As others have said, the at home models are already basically good enough, so why focus on research with no economic value like mathematics? Because optimization for building god-like intelligence is the end goal.
It is less economics and more Eschatology that you have to work on here.
Not my own opinion just giving an explanation for why 50 trillion dollars in compute expenditure would be rational under certain beliefs.
TheTopNacho | 11 hours ago
Developed countries offloaded manual labor markets quite a while ago and that had massive economic impacts.
They switched largely to intellectual labor to sustain growing economies. For which companies are now actively looking for ways to offload intellectual labor onto computers.
It won't take a massive shift in job markets to see other, unrelated, areas affected by decreased labor value. Whether that's manual or intellectual labor. People shifting away from computer coding to ,,, whatever it is they will, will just decrease the value of work in those fields. Nothing good will come from this
PatchyWhiskers | 10 hours ago
Farm labour moved to factory labour moved to intellectual labor. Where next?
Mrsrightnyc | 10 hours ago
The other thing is that if the main benefit of AI is savings through job replacement, what happens when there’s way fewer consumers to buy stuff? CEO like to talk about efficiency and cost savings, but the economy actually runs the best when people are working, making a decent living, and can spend because they aren’t worried about having to save a ton in case they lose their job and never get another one. Essentially they’ll turn everyone in their prime earning years into penny pinching pensioners.
Content_Source_878 | 9 hours ago
The tech billionaires and the CEOs aren’t the same people imo.
Tech people simply think they will create and we will consume it. Apple hasn’t released a new product since the Apple Watch. Just add-ons.
The CEO people only see value in relation to how cheap it is. They jumped at the thought of replicating 100 workers with cheap Ai. But once Ai becomes too expensive they will leave it for the cheaper option
TheGRS | 13 hours ago
I’d challenge the idea that this is just internet+ and I think you know that. Every day I’m figuring out how to use the tool better to get my work done. The agent can write code and interact directly with APIs on my behalf, much faster than I can. So that means it’s being used for increasing work output in less time. And you’re also probably well aware that it can write a whole lot of content to varying degrees. For a lot of business content it’s at the point of good enough.
look | 11 hours ago
“write code and interact with APIs on my behalf” is basically what average webdev software engineers have been doing for the past two decades.
They started on the higher value tasks, and now with AI you can do the lower value tasks that weren’t worth paying even a mid-to-bottom rung software engineer to do.
That’s not world changing. That’s leftover crumbs.
TheGRS | 11 hours ago
Is your point that this isn’t a useful thing for the agent to do? Because yes doing these things was often tedious and time consuming.
look | 7 hours ago
Sure, it’s useful. But it’s mostly stuff that people with a little more computer and programming knowledge had already been automating for themselves for decades now.
Sally in HR being able to do a few more things on her computer without help from IT now is not at the same scale of world changing impact as the internet.
DarkSkyKnight | 9 hours ago
> That’s not world changing. That’s leftover crumbs.
🙄
I guess we have sunk so low that AI is larping as anti-AI redditors now
Internationallegs | 12 hours ago
That's because it's close to free right now. How much would you be willing to pay for AI output? Like what is the max? Because we literally don't know what the real cost is.
Remember when DoorDash was cheap during covid? Everyone used it. It was cheap because they were operating at a loss. Now they make a profit but the prices are so expensive hardly anyone uses it.
Of course everyone is using AI now because it's literally free.
mr-ron | 11 hours ago
Remember when the Internet cost money per minute? Remember when computers even been super basic ones for the home were outlandishly expensive?
TheGRS | 11 hours ago
It’s definitely subsidized, but open models for a lot of these tasks exist. The hardware may be expensive but it’s measurable.
Content_Source_878 | 9 hours ago
But the open models aren’t what these tech companies want us to use.
Like the next highest amount of data centers isn’t in China but they have a bigger population than the US.
The math ain’t mathing
AndTheOnlyOne | 11 hours ago
AI is going beyond just being profitable though. It's a national goal now - and even if investors aren't stepping in, government will step in with the funding.
britax12 | 9 hours ago
year 3 u mean?
fremeer | 8 hours ago
The one benefit of AI I think is less appreciated is how much it will help smaller firms and developers.
A large firm has the capacity to hire 1000 low level workers to do the grunt work. They replace them with AI and it's not great for jobs. But a small firm can now for the same cost also have the same level of work force.
That will allow them to take on more jobs. Provide greater levels of competition.
Same for development. A small open source project just doesn't have the man power to really fix issues quickly and streamline processes. But now a lot of the boring grunt work is done for you and you can focus on features. Even stuff you might not be very good at you can try and if it kind of works but is janky you can make a more educated decision on finding someone more knowledgeable to do it better. Because you already know the worst version of it works.
But making small firms and developers more competitive is probably not in the best interests of these giants so I think we will see a lock down or change in access capabilities for these kinds of projects.
Imagine someone uses Gemini to make a YouTube or google maps competitor. Why would google allow that.
PrinnyFriend | 7 hours ago
>Same for development. A small open source project just doesn't have the man power to really fix issues quickly and streamline processes. But now a lot of the boring grunt work is done for you and you can focus on features. Even stuff you might not be very good at you can try and if it kind of works but is janky you can make a more educated decision on finding someone more knowledgeable to do it better. Because you already know the worst version of it works.
Actually that is something I never thought off. I kind of hope that all the people who got laid off from AI, learn to utilize AI to sweep the product from under the Big Tech companies.
I have seen what photonshop and other free competitors did to photoshop and it is pretty amazing that for 80% of people who use photoshop, may never even notice a change....and the product is free. Same with people merging games together.
If the people who suffered from AI the most, are able to turn a small start up with limited manpower into a multi-billion dollar company with AI, then AI has shown real potential and ironically has cannibalized the value of what these large companies are creating.
anewleaf1234 | 8 hours ago
I would say the issue is both How and well if that happens, WTF.
If you remove all jobs from people how does the economy stay afloat?
Can can a person afford to buy a beer at a bar? How can someone afford day care?
Because there isn't ever an answer to that question. Just lots of false promises.
burnsniper | 9 hours ago
Lol Ai is so much more than the “browser + add on.” I was a skeptic too, but Claude has been able to do real work and analysis for me at what I estimate is 1/12th the time.
Example: Ask a 4 person tech team to create a tableau dashboard measuring and tracking KPIs for my business - maybe a year and it will just be data visualization. Claude can knock this out and perfect it in about a month of vibe coding. Also the Claude version will be 100% interactive and present the data just how you want it and can pull from data and text sources (notes). Furthermore, if you don’t like something you can just ask it to fix it and in 40 seconds it’s fixed/rearranged/etc.
Another example: combine the graphics and cad drawings from 12+ reports into one interactive visual. May take a professional CAD engineer two weeks to get to 75% of the way there. Claude does it nearly perfectly within an hour of vibe coding.
This is truly transformative work.
Beer_bongload | 11 hours ago
>Right now....it is just being used as the Internet +. If the Internet is the "Browser", AI is just the "add-on"....AI has shown us that it is mostly used as a better search engine
Thats a shit take and you know it. Be better.
nerdvegas79 | 9 hours ago
So, electricity was the real innovation then, and the internet was just electricity+ ?
Dumb take sorry.
Spenraw | 15 hours ago
I don't see how the entire us economy isn't insanely fucked to the point of collapse?
Japanese Yen just strengthened today as well, the yen carry trade with this AI bubble finally getting ready to pop all with the oil crisis
Best case just us fucked but looks like going to take entire us economy with it
TrueBirch | 15 hours ago
I'm not challenging your longterm hypothesis, but the S&P 500 is still up 4% over the past six months. Maybe there will be a global economic reckoning, but at least so far this isn't it.
tymtt | 15 hours ago
Yes but the real question is less about how well the indexes have done and more about expectations of future performance. The trailing 6 month average is rapidly degrading as the run up from April-May is shed.
Pugsly007 | 15 hours ago
AI has been driving the US economy. If this carries longer we’ll be in a recession soon if we aren’t already. Then the bubble in housing hits.
defaultedebt | 15 hours ago
No, investment has been driving the US economy. Investment that would have been invested elsewhere within the economy had AI not been a thing. GDP growth was 3.4% annualized in Q3. Is that close to a recession? I don't think so.
FinsOfADolph | 14 hours ago
Stupid question. Why aren't a lot of US companies hiring if we're not in a recession? Also, what's your definition of a recession?
defaultedebt | 14 hours ago
>Why aren't a lot of US companies hiring if we're not in a recession?
Well, they're also not firing.
https://www.clevelandfed.org/publications/economic-commentary/2026/ec-202617-the-low-hire-low-fire-labor-market
So it's just that the labor market is less dynamic than it was between 2021 - 2023 from the post-Covid hiring sprees. In fact, by some metrics, such as prime age labor force participation, labour participation has risen significantly since Covid and remained mostly stable over the last 3 years. I mean, layoffs have been happening in tech and finance, that much is true, but not at the scale required to lead to an economic contraction.
> Also, what's your definition of a recession?
https://www.nber.org/research/business-cycle-dating/business-cycle-dating-procedure-frequently-asked-questions
NBER's definition, which is:
>The NBER's traditional definition of a recession is that it is a significant decline in economic activity that is spread across the economy and that lasts more than a few months. The committee's view is that while each of the three criteria—depth, diffusion, and duration—needs to be met individually to some degree, extreme conditions revealed by one criterion may partially offset weaker indications from another. For example, in the case of the February 2020 peak in economic activity, we concluded that the drop in activity had been so great and so widely diffused throughout the economy that the downturn should be classified as a recession even if it proved to be quite brief.
FinsOfADolph | 14 hours ago
I've had a problem with that definition for a few reasons: 1) we've had ongoing inflation since COVID for various reasons, including climate change, corporate greed and the decline of 3rd spaces. There's a lot of FOMO when you're not sure prices will be the same today as tomorrow. 2) someone above seemed to say that A.I. is pretty much keeping the economy afloat (as companies threaten widespread and long lasting unemployment no less) 3) Enshittification, requiring people to spend more money more often to get the same amount of quality they would have gotten 10 years ago
Also the labor market is by all means abysmal, even for the highly skilled, connected people the U.S. claims to prize. I'm willing to accept that we're willing to accept a slow death to quality of goods and an increase in corporate scamming. But the economy is so bad that you have pundits for respected publications (The Atlantic, WSJ) playing damage control by accusing people of lying about it.
I guess a recession isn't useful to describe relevant economic conditions for anyone under upper-middle class anymore.
defaultedebt | 13 hours ago
I think I'll stick to trusting the NBER, who are one of the most reputable economic research organisations in the world, on their definition of recession.
> 1) we've had ongoing inflation since COVID for various reasons, including climate change, corporate greed and the decline of 3rd spaces.
You are listing irrelevant things. GDP is inflation adjusted. A recession is a technical term, very clearly defined by the NBER.
> 2) someone above seemed to say that A.I. is pretty much keeping the economy afloat
Investment, not AI. Money is being spent on investment - money which would still be out there had AI never existed. It's also still real investment, even if it is AI dominated.
> 3) Enshittification, requiring people to spend more money more often to get the same amount of quality they would have gotten 10 years ago
This is why you have quality adjustments in inflation baskets. (Again, generally not relevant to measuring a recession).
https://www.bls.gov/cpi/quality-adjustment/
It's highly detailed and regularly reviewed so that quality is controlled for in economic data. (And again, generally irrelevant to how we measure a recession).
>Also the labor market is by all means abysmal, even for the highly skilled, connected people the U.S. claims to prize.
Did you read the research I linked? Or even taken a look at prime age labor force participation rate:
https://fred.stlouisfed.org/series/lns11300060
>By combining these various dimensions of fluidity, Molloy et al. (2016) date the beginning of the downtrend to the early 1980s, possibly before. They, along with other researchers, have found that the decline [in fluidity] has been broad based across types of workers and sectors and that perhaps half of the general decline in fluidity can be attributed to demographic factors, notably the aging of the population and increases in levels of education.
>I guess a recession isn't useful to describe relevant economic conditions for anyone under upper-middle class anymore.
It's a useful metric to describe a recession. It was never ever intended to describe the economic conditions of a particular decile of the economy. What you said there is also just nonsense.
Savings_Painting1588 | 14 hours ago
What? Are US firms not… hiring? I thought US unemployment was low/falling?
Zlifbar | 14 hours ago
Unemployment numbers are falling because people have stopped looking for jobs plus this administration has been falsifying numbers on release with revisions months later.
defaultedebt | 13 hours ago
So we're just openly lying now? Great.
FinsOfADolph | 14 hours ago
I haven't trusted unemployment numbers since 2022. I don't have many numbers, but a lot of growth has been concentrated in white collar jobs in the tech industry and maybe medical industry. Since late 2022, a lot of previously-secure tech companies like Google, Meta and Microsoft have been laying off their employees in waves. The job market is so bad that common advice is 1) don't leave your job unless you have one lined up and 2) nepotism is the best way to get a job. Hell, Microsoft just offered voluntary exit packages to 60000 of its employees!
Even this article here admits that not a lot of companies are hiring. Just ignore the bullshit about "Americans enjoying job security".
events_occur | 13 hours ago
> Then the bubble in housing hits.
Was with you until this, which makes the comment feel like fantasy. There is no housing bubble. The country is just severely, severely supply-constrained from decades of underbuilding and obstructionism. The vast majority of homeowners have rates that we will probably not see for the rest of our lifetime, and will simply not sell.
twenty_three_three | 11 hours ago
I can't speak for him, but the concern arises when folks are forced to sell. Mass layoffs, need to move for a job, etc. We will find out what people do when we have significant forced relocation. Keep the house, attempt to rent while you rent? Take on two mortgages? what's the play? There's going to be some real losers.
Spenraw | 14 hours ago
This just looks like a correction not a bubble pop
TrueBirch | 14 hours ago
Considering the market has gained a tiny bit since I posted my comment, I'm inclined to agree.
I do think a lot of stocks are probably overpriced. But that doesn't mean every red day is TEOTWAWKI.
frenchiefanatique | 12 hours ago
Imagine thinking anything that has happened over a 1-2 hour span is enough to validate an opinion on whether or not we're seeing a correction or a pop.
twenty_three_three | 11 hours ago
This has got to be the strangest bot/user behavior i've seen on this sub.
It's just anecdotal, but this sort of commentary makes me even more worried. I commented yesterday about the the very similar sounding pro tech comments coming in on every story about a tech company's equity losses. To see this continue today in a different flavor is not great.
TrueBirch | 7 hours ago
I'm a Boglehead. I don't usually follow the daily market changes. Because every day there's a new price and a lot of discussion about it and most of the time it amounts to nothing. The market going down a bit then stabilizing isn't anything especially interesting.
twenty_three_three | 6 hours ago
Why did you feel to comment today, here if you dont usually do this?
TimeGrownOld | 13 hours ago
I'm just going to stay invested in crypto
That way when it crashes I'll have only myself to blame
FloppingGazelle | 12 hours ago
I’m not an options trader and never will be. But I’m majorly hedging out my long term plays with options on volatility. A lot of idiots are in the market ready to fire off sells
Zlifbar | 14 hours ago
Questions EXISTENCE of returns. There is no evidence of all the supposed amazing amounts of money that AI has saved or earned. It's all circular financing and managers blowing smoke to justify their herd-following decisions.
DrTreeMan | 13 hours ago
How about the increases in cloud computing revenue amd profits that GOOG, AMZN. and MSFT are seeing? We can't say this is all AI specific spending, but this is what the data centers are being built for.
astral-dwarf | 12 hours ago
Revenue from anthropic and OpenAI?