That’s my main concern with the bubble as well. The economy functions better with diversity. There are so many industries that could use funding outside of tech and some that could use funding with some new tech approaches.
Just the sheer size of investment into a single idea should worry people. Diversity is very important.
In the same way that bonds compete for investment dollars that would go into the stock market. The stock market has been going up only for AI companies, meaning all the money is flowing there.
If AI didn't have the hype, those marginal investment dollars driving up AI stock prices, they'd be flowing in to other uses of the capital, some of which I believe would be far more productive.
That presupposes we want home building to be driven by investors who are looking to make the highest ROI.
Isn't the healthier option that people who want homes get them built? You don't need a general investor in that loop, except if we're talking bout "mass housing" but even then, those are generally property developers and they are unlikely to use those dollars for non-property development.
Most people don't want to be their own builder. All of society is set up in the US to prevent people who want a home from building their own, from the permitting to the financing to the actual contracting. Except for the very wealthy.
A good chunk of people, perhaps most, want a house close to jobs and shops and schools and parks and museums and all those sorts of amenities.
Specialization into the very difficulty job of designing, permitting, and constructing housing makes sense. Not only is it a difficult job requiring tons of knowledge that a random person neither cares about nor needs to know about, specialized builders are far far far more efficient at building housing, and getting more housing completed for less money.
I didn't presuppose that housing should be built based on ROI, it just is that way, due to the realities of housing construction. I do think we need a public builder, modeled specifically after Finland's housing builder, as they have figured out some really great ways. Such a public builder has several economic advantages: provides a competitive floor for the private for-profit builders, and perhaps just as importantly a public builder can operate during the bottom of the economic cycle, employing construction workers through the downturn and providing much needed housing through the downturn thereby shortening the economic downturn.
I think Texas bucked the trend, because they did not try to remove profit from home building and in fact made it cheaper to build housing.
Homeowners and landlords profit from high housing prices, and in most economically productive metro areas local politics is controlled by homeowners and landlords. They therefore try to stop housing production, using sneaky and deceptive language like "taking the profits away from developers" which in turns mean that the landlords make even higher profits from unproductive scarcity.
That's just one of many factors of housing production, and in many places, high interest rates on loans, and high construction costs, are also barriers equivalent in cost to local permitting problems.
Subsidized construction loans, or lower interest rate mortgage loans only for the first owner of a new building might be one way to break out of the current environment. At least it's one of the very few levers that a government could have here, after fixing permitting to be fast, efficient, and cheap. Building a larger and more efficient construction workforce is hard and not easy for a government to stimulate, except through increasing the size of the construction market through other means, and then ensuring that there's a true market of builders rather than a monopoly.
What I'm trying to say is what is your concrete plan to incentivize investment AWAY from AI and towards housing? You can't tell people they're not allowed to invest in the thing that they think is gonna make them the most money.
Put differently, it is easy to say "there should be more investment in X and more in Y" but it takes concrete incentives to change that behavior.
Fro example, if you have a 401(K) or some other investments, should YOU be forced to redirect that into housing?
Yeah when I see someone's slop project get a billion from YC for a product which clearly is going to flop, it sickens me. Imagine how much you could do with that kind of money. Everyone is funding total trash just to get on the AI bandwagon.
There's a reason why every pre-existing startup has been slapping AI all over their products - it's basically the only way to get funding at the moment.
I think this has been a phenomenon since VC became a thing, there's always been trends and fads - CEOs have had to chase them even when it doesn't make sense for the product.
I think this is the most wide-scale one I can personally recall (although I presume dotcom boom was similar), but I remember through the 2010s every database startup having to chase the same ball-game. Everyone was doing NoSQL, then everyone was doing analytics etc. Of course now all those companies have AI-native, vector-embedding, RAG, MPCs slapped all over their products.
Anecdata but from my vantage this has been true for a decade, I imagine it is only more true today, but investors were nudging to add AI to products the entire time I’ve been in tech, and failure to do that was always flagged as a risk for funding rounds. Need to show “AI/ML in the pipeline” by the next funding round or x/y/z is out…
I’m asking because I see so many products I am interested in using, but I have no interest in connecting my Gmail, Slack, or other tools to it.
I just want a software I can use for that purpose, and then move on.
Pre-AI most of the slack channel connectors turned into spam. We would relegate it to some alerts channel that nobody used or looked at and a systems manager would close it a year later.
If a piece of software is good at what it does, it will have a tab in my browser where it runs and I’ll check it occasionally.
I don’t know about you guys but I just like doing work that’s in front of me, and after completing it I like to move on to the next thing.
90% of integrations/connections just add noise, at least in my experience.
No, but who is going to fund hundreds of millions into a SaaS that can't generate the returns?
Take Asana for example. It raised a total of $453M. If the same idea tried to raise money in 2026, I don't think it would raise $1. You can clone Asana in a week with AI Agents.
nobody would scream circular financing because of this. it's only circular if those companies would then be investing (directly or indirectly) in nvidia.
this however does scream "manufactured demand" - if the terms of your financing dictates that the money is used to purchase your own product.
It can also be a way to make the numbers look bigger for everyone...
Rather than sell a GPU at $500, instead you put the price up to $2000 but then invest in a startup on condition 80% of the investment is used to buy your hardware at inflated prices.
Arguably this is what's happening to 'datacenter' GPU's which are far more expensive than consumer ones for similar performance.
Are they still? At latest retail prices of 5090s, the economics don't seem to reflect that much of a discount; especially when you consider the half-rate BF16/FP16/etc accumulation (which has not been hacked/unlocked yet).
Isnt that... what circular financing often refers to?
If I invest in you and you use my stack and pay me for it. It's circular. I don't think it should be illegal to do this type of vendor financing/investment or what not as it has its legitimate uses.
Circular financing is when the revenue you get back is a fraction of what you put in, essentially. If you put in 1$ to get 10$ back in (future) revenue, that's not circular financing.
Yes but how many such isolated cases are there? There is a real possibility that a lot of the AI bubble is fuelled not by dealing with itself but by pulverizing influence all over the market.
In this case (gpu provider funding many startups to buy their gpus) it seems bad to me. They have deals with so many customers it looks like collusion to keep prices high on both sides.
I don't think 1 loop is enough. You could remove the loop by just giving the GPUs directly as part of the investment. But investing in pure cash gives more flexibility especially in time where there may not even been spare GPUs able to be given away.
no, circular financing refers to it bouncing through several loops of that. the implication is that it’s system predators working the math, not the business. you should hear it in the same tones that bankers use the word leverage
This reminds me a lot of the .COM startups in 2000 that were selling stuff to each other and once one failed, the whole chain would come down slowly. The scary part is the sheer amount of money that’s floating around now. In 2000 is was a few million here and there. Now it’s multiple billions.
> Sure but if Nvidia is funding these startups, isn’t it really still circular financing??
For it to be circular financing, Nvidia would have to be investing in these startups who in turn buy significant $ volume in products and services from Nvidia.
Is this happening? For instance, what's a concrete example of a startup or even a group of startups where this is happening and what is the $ amount as a percentage of Nvidia annual sales?
> If it looks like circular financing and smells like circular financing it’s circular financing.
Problem is many people don't know what circular financing is. For instance, pick one of Nvidia's investments and explain why it is a circular financing arrangement.
I mean, it's not quite as blatant as Amazon investing AWS credits, but a significant portion of it is going to be returned to Nvidia as hardware purchases, so it's not not-circular...
Hey all, I run a funding-data API and I pulled every venture round of $100M or more from Oct 2025 to Sep 2026 that named its investors, we have 903 in total for this group.
And we found that Nvidia (or its venture arm) was in 79 of them. Sequoia was in 75 and a16z in 73. Nvidia on its own beat Lightspeed, General Catalyst and Accel.
Two other things that kinda surprised me was that counting rounds and counting dollars give different leaders.
So the average round Fidelity joined was $7.7B, against $3.5B for Sequoia and 38% of these rounds had a corporate investor, up from 35% the year before.
The obvious question is how much of this is Nvidia funding demand for its own chips. I'd like to hear from people who've been on either side of one of these rounds.
Doesn't it? I would not expect the largest public companies to be the most active venture capitalists. The other 10 most active funders are all specialist investors, not standalone companies with venture arms. You have to go down to about #15 to find the next corporate investor, and it's not Microsoft or Apple or Saudi Aramco, but relative minnow Salesforce.
Bear with me.... Leverage and Insurance. Leverage against whom? US govt.
Maybe VCs and Private Equity funds represents what is essentially oligarchs. They are oligarchs as institutions. Everyone knows this already. US has certain liking to captialism as long as it is represented trough institutions and not individuals. Controversial I know, but standard oil, billy boy gates and monopoly trials, musk and spacex being public companies. It is just America.
So, NVIDIA figured out an interesting play. NVIDIA is playing a risky game where US govt can anytime say "fuck you, I hate you now". But if they embroil themselves in enough foundational businesses across america saying fuck you to them is difficult. Startups are " think of the children" corporate America.
They have become part of the oligarch class, they have invested where politicians childrens are working or lets be clear founded startups. This is not curricular economy, it is nvidia buying insurance. If they were read blooded America it would have been leverage, but this is insurance. Saying fuck you to them is progressively getting harder.
Any way my legs are frozen that is enough conspiracy theories I can write from the stall.
People are downvoting you because they don't want to think of the implications of letting this system run for another 20-30 years. Look how bad its already gotten through monopolistic behavior and blatant government corruption. This thread is filled with evidence that oligarchs are literally buying up every successful startup but no one wants to recognize that having a few individuals control vast swaths of economic growth will lead us to ruin.
Everyone on HN too obsessed with getting their own bag to see the problem.
Evidence that large VC rounds are getting very concentrated on AI (specifically the kind where you need to buy lots of NVIDIA gear).
Is this leaving good companies unfunded because they're not sufficiently "AI"?
That’s my main concern with the bubble as well. The economy functions better with diversity. There are so many industries that could use funding outside of tech and some that could use funding with some new tech approaches.
Just the sheer size of investment into a single idea should worry people. Diversity is very important.
I think it's crowding out investment in things like housing. And lack of housing is economically killing the entire next generation of workers.
> I think it's crowding out investment in things like housing.
In what way is building homes competing for the same dollars as investing in startups or in listed companies?
In the same way that bonds compete for investment dollars that would go into the stock market. The stock market has been going up only for AI companies, meaning all the money is flowing there.
If AI didn't have the hype, those marginal investment dollars driving up AI stock prices, they'd be flowing in to other uses of the capital, some of which I believe would be far more productive.
That presupposes we want home building to be driven by investors who are looking to make the highest ROI.
Isn't the healthier option that people who want homes get them built? You don't need a general investor in that loop, except if we're talking bout "mass housing" but even then, those are generally property developers and they are unlikely to use those dollars for non-property development.
That's an is/ought fallacy.
Most people don't want to be their own builder. All of society is set up in the US to prevent people who want a home from building their own, from the permitting to the financing to the actual contracting. Except for the very wealthy.
A good chunk of people, perhaps most, want a house close to jobs and shops and schools and parks and museums and all those sorts of amenities.
Specialization into the very difficulty job of designing, permitting, and constructing housing makes sense. Not only is it a difficult job requiring tons of knowledge that a random person neither cares about nor needs to know about, specialized builders are far far far more efficient at building housing, and getting more housing completed for less money.
I didn't presuppose that housing should be built based on ROI, it just is that way, due to the realities of housing construction. I do think we need a public builder, modeled specifically after Finland's housing builder, as they have figured out some really great ways. Such a public builder has several economic advantages: provides a competitive floor for the private for-profit builders, and perhaps just as importantly a public builder can operate during the bottom of the economic cycle, employing construction workers through the downturn and providing much needed housing through the downturn thereby shortening the economic downturn.
the US state of Texas does break the mold of that unfortunately accurate overview, I am told..
I think Texas bucked the trend, because they did not try to remove profit from home building and in fact made it cheaper to build housing.
Homeowners and landlords profit from high housing prices, and in most economically productive metro areas local politics is controlled by homeowners and landlords. They therefore try to stop housing production, using sneaky and deceptive language like "taking the profits away from developers" which in turns mean that the landlords make even higher profits from unproductive scarcity.
That's just one of many factors of housing production, and in many places, high interest rates on loans, and high construction costs, are also barriers equivalent in cost to local permitting problems.
Subsidized construction loans, or lower interest rate mortgage loans only for the first owner of a new building might be one way to break out of the current environment. At least it's one of the very few levers that a government could have here, after fixing permitting to be fast, efficient, and cheap. Building a larger and more efficient construction workforce is hard and not easy for a government to stimulate, except through increasing the size of the construction market through other means, and then ensuring that there's a true market of builders rather than a monopoly.
The only way people who want homes get them built without an ROI to an investor is if they pay for it themselves.
Providing ROI to investors is a good thing if you want people to invest in something, because it means common self-interest will ensure it gets done.
Do you have a solution that would steer people who would build homes but who are investing in AI?
And to zoom out, aren't there other things that are EVEN MORE important to invest in than building homes?
People who want houses have to be willing to either pay more for them or pay higher mortgage rates, I don't know what else you want me to suggest?
Where else is investor ROI going to come from?
What I'm trying to say is what is your concrete plan to incentivize investment AWAY from AI and towards housing? You can't tell people they're not allowed to invest in the thing that they think is gonna make them the most money.
Put differently, it is easy to say "there should be more investment in X and more in Y" but it takes concrete incentives to change that behavior.
Fro example, if you have a 401(K) or some other investments, should YOU be forced to redirect that into housing?
> what is your concrete plan to incentivize investment AWAY from AI and towards housing?
I don't have one and don't think trying to nudge people this way is a good thing to do.
> You can't tell people they're not allowed to invest in the thing that they think is gonna make them the most money.
Exactly.
> "there should be more investment in X and more in Y"
I'm not the one saying that, you are!
Yeah when I see someone's slop project get a billion from YC for a product which clearly is going to flop, it sickens me. Imagine how much you could do with that kind of money. Everyone is funding total trash just to get on the AI bandwagon.
There's a reason why every pre-existing startup has been slapping AI all over their products - it's basically the only way to get funding at the moment.
I think this has been a phenomenon since VC became a thing, there's always been trends and fads - CEOs have had to chase them even when it doesn't make sense for the product.
I think this is the most wide-scale one I can personally recall (although I presume dotcom boom was similar), but I remember through the 2010s every database startup having to chase the same ball-game. Everyone was doing NoSQL, then everyone was doing analytics etc. Of course now all those companies have AI-native, vector-embedding, RAG, MPCs slapped all over their products.
Anecdata but from my vantage this has been true for a decade, I imagine it is only more true today, but investors were nudging to add AI to products the entire time I’ve been in tech, and failure to do that was always flagged as a risk for funding rounds. Need to show “AI/ML in the pipeline” by the next funding round or x/y/z is out…
I'm cancelling my blockchain toaster startup, anyone up for investing in my AI toaster (powered by Nvidia)?
I don't see how a pure SaaS company without any AI features can succeed in a big way in 2026.
Does everything need AI though?
I’m asking because I see so many products I am interested in using, but I have no interest in connecting my Gmail, Slack, or other tools to it.
I just want a software I can use for that purpose, and then move on.
Pre-AI most of the slack channel connectors turned into spam. We would relegate it to some alerts channel that nobody used or looked at and a systems manager would close it a year later.
If a piece of software is good at what it does, it will have a tab in my browser where it runs and I’ll check it occasionally.
I don’t know about you guys but I just like doing work that’s in front of me, and after completing it I like to move on to the next thing.
90% of integrations/connections just add noise, at least in my experience.
Take Asana for example. It raised a total of $453M. If the same idea tried to raise money in 2026, I don't think it would raise $1. You can clone Asana in a week with AI Agents.
That’s fair. I agree there.
I guess I meant more like apps for communication, or managing infrastructure.
I like having Slack like I said. I don’t like connecting 100s of things to Slack when 90% of them end up value-loss noise.
> Evidence that large VC rounds are getting very concentrated on AI (specifically the kind where you need to buy lots of NVIDIA gear).
The most recent $1B+ funding rounds:
Zipline - Nov 5 - $1.5B
HUMAIN - Nov 3 - $2.6B
Fal - Nov 3 - $1.3B
OpenAI - Oct 31 - $11B
Kalshi - Oct 29 - $1B
That's 3/5. Zipline is drone delivery and Kalshi is betting markets.
Also evidence that Nvidia is funding its own customers and artificially boosting demand. What could possibly go wrong?
Do you have evidence that a non-AI company can be good?
Before anyone screams "circular financing", it really isn't.
It's almost certainly about making sure that these startups use Nvidia stack.
nobody would scream circular financing because of this. it's only circular if those companies would then be investing (directly or indirectly) in nvidia.
this however does scream "manufactured demand" - if the terms of your financing dictates that the money is used to purchase your own product.
It can also be a way to make the numbers look bigger for everyone...
Rather than sell a GPU at $500, instead you put the price up to $2000 but then invest in a startup on condition 80% of the investment is used to buy your hardware at inflated prices.
Arguably this is what's happening to 'datacenter' GPU's which are far more expensive than consumer ones for similar performance.
Are they still? At latest retail prices of 5090s, the economics don't seem to reflect that much of a discount; especially when you consider the half-rate BF16/FP16/etc accumulation (which has not been hacked/unlocked yet).
Even 5060 Ti 16GBs are at 1.8x MSRP.
> Arguably this is what's happening to 'datacenter' GPU's which are far more expensive than consumer ones for similar performance
What's an example of a datacenter GPU that has the same performance of a retail GPU but sold at a much steeper price?
Salesforce Ventures has invested $7b in more than 700 companies to boost its ecosystem.
I didn't see people calling this manufactured demand or circular financing.
Isnt that... what circular financing often refers to?
If I invest in you and you use my stack and pay me for it. It's circular. I don't think it should be illegal to do this type of vendor financing/investment or what not as it has its legitimate uses.
But it is circular.
The money that Nvidia gives to these companies hardly make a dent in Nvidia's total revenue.
It's just to make sure the next OpenAI uses Nvidia stack.
None of this refutes the circular nature of it all, though.
Circular financing is when the revenue you get back is a fraction of what you put in, essentially. If you put in 1$ to get 10$ back in (future) revenue, that's not circular financing.
Yes so advertising and vendor locking. Which brings money back to them. Hence circular.
When people accuse AI companies of circular financing, they meant it as a way to juice up their financial reports.
This is not the case. This is just having influence.
I don't think this locks the startup into Nvidia stack. I don't think this is advertising either. It's just influence.
Circular is not the word I use to describe this business tactic.
Yes but how many such isolated cases are there? There is a real possibility that a lot of the AI bubble is fuelled not by dealing with itself but by pulverizing influence all over the market.
It also makes everyone else’s success theirs. Personally I’m really happy to effectively be an LP in the Nvidia VC fund.
So you’re arguing it is circular but it’s only a little bit of money compared to Nvidia revenue so it doesn’t count?
> So you’re arguing it is circular but it’s only a little bit of money compared to Nvidia revenue so it doesn’t count?
Circular financing isn't inherently bad.
The question is whether it finances productive economic activity OR whether it manufactures the appearance of economic activity.
In this case (gpu provider funding many startups to buy their gpus) it seems bad to me. They have deals with so many customers it looks like collusion to keep prices high on both sides.
It seems so for those who lack critical thinking and think everything is a conspiracy theory to keep the AI bubble alive.
I’m not sure what about the facts of Nvidia subsidising many of its customers you think is a conspiracy theory.
If we do see a crash and they are sued by shareholders for fraud, perhaps you will reconsider your position.
> If I invest in you and you use my stack and pay me for it. It's circular.
If I invest in you SO THAT you buy from me IN ORDER to materially and artificially boost my revenue.
I don't think 1 loop is enough. You could remove the loop by just giving the GPUs directly as part of the investment. But investing in pure cash gives more flexibility especially in time where there may not even been spare GPUs able to be given away.
Having equity in the startup means Nvidia can influence the founders to use Nvidia stack.
Everyone in AI already wants to be using Nvidia GPUs. Unless you worked at Google.
no, circular financing refers to it bouncing through several loops of that. the implication is that it’s system predators working the math, not the business. you should hear it in the same tones that bankers use the word leverage
Sure but if Nvidia is funding these startups, isn’t it really still circular financing??
This reminds me a lot of the .COM startups in 2000 that were selling stuff to each other and once one failed, the whole chain would come down slowly. The scary part is the sheer amount of money that’s floating around now. In 2000 is was a few million here and there. Now it’s multiple billions.
> Sure but if Nvidia is funding these startups, isn’t it really still circular financing??
For it to be circular financing, Nvidia would have to be investing in these startups who in turn buy significant $ volume in products and services from Nvidia.
Is this happening? For instance, what's a concrete example of a startup or even a group of startups where this is happening and what is the $ amount as a percentage of Nvidia annual sales?
See my post here: https://news.ycombinator.com/item?id=48874687
Especially as long as demand is as high as it is.
I paid real money for the 4090. Real $2000
It's not circular financing, although Nvidia give startups investment money so they can give it back to Nvidia for hardware?
They're forecasting around $700b in revenue next year.
These investments are likely drop in a bucket. A rounding error.
What they're truly looking for is the next OpenAI and to make sure that company runs on Nvidia.
If it looks like circular financing and smells like circular financing it’s circular financing.
Nobody is under any illusions that most if not nearly all this money is for anything other than buying GPU compute.
> If it looks like circular financing and smells like circular financing it’s circular financing.
Problem is many people don't know what circular financing is. For instance, pick one of Nvidia's investments and explain why it is a circular financing arrangement.
Nvidia at this point is the new central bank of the tech world
Or even if it is it's only 1 in 11 not 10 in 11..
Then it is then.
AI boosters always deny reality.
I mean, it's not quite as blatant as Amazon investing AWS credits, but a significant portion of it is going to be returned to Nvidia as hardware purchases, so it's not not-circular...
Hey all, I run a funding-data API and I pulled every venture round of $100M or more from Oct 2025 to Sep 2026 that named its investors, we have 903 in total for this group.
And we found that Nvidia (or its venture arm) was in 79 of them. Sequoia was in 75 and a16z in 73. Nvidia on its own beat Lightspeed, General Catalyst and Accel.
Two other things that kinda surprised me was that counting rounds and counting dollars give different leaders. So the average round Fidelity joined was $7.7B, against $3.5B for Sequoia and 38% of these rounds had a corporate investor, up from 35% the year before.
The obvious question is how much of this is Nvidia funding demand for its own chips. I'd like to hear from people who've been on either side of one of these rounds.
For what was until a few days ago the largest public company in the US, that doesn't sound unusual?
Still interesting to have concrete numbers on the phenomenon, I’d say.
Doesn't it? I would not expect the largest public companies to be the most active venture capitalists. The other 10 most active funders are all specialist investors, not standalone companies with venture arms. You have to go down to about #15 to find the next corporate investor, and it's not Microsoft or Apple or Saudi Aramco, but relative minnow Salesforce.
Looks like Jensen Huang has been making the rounds...
Bear with me.... Leverage and Insurance. Leverage against whom? US govt.
Maybe VCs and Private Equity funds represents what is essentially oligarchs. They are oligarchs as institutions. Everyone knows this already. US has certain liking to captialism as long as it is represented trough institutions and not individuals. Controversial I know, but standard oil, billy boy gates and monopoly trials, musk and spacex being public companies. It is just America.
So, NVIDIA figured out an interesting play. NVIDIA is playing a risky game where US govt can anytime say "fuck you, I hate you now". But if they embroil themselves in enough foundational businesses across america saying fuck you to them is difficult. Startups are " think of the children" corporate America.
They have become part of the oligarch class, they have invested where politicians childrens are working or lets be clear founded startups. This is not curricular economy, it is nvidia buying insurance. If they were read blooded America it would have been leverage, but this is insurance. Saying fuck you to them is progressively getting harder.
Any way my legs are frozen that is enough conspiracy theories I can write from the stall.
People are downvoting you because they don't want to think of the implications of letting this system run for another 20-30 years. Look how bad its already gotten through monopolistic behavior and blatant government corruption. This thread is filled with evidence that oligarchs are literally buying up every successful startup but no one wants to recognize that having a few individuals control vast swaths of economic growth will lead us to ruin.
Everyone on HN too obsessed with getting their own bag to see the problem.