This doesn’t feel like news to me? A tech startup that has 18 months of runway is pretty good honestly. The story is the quantity of cash involved in that runway.
Edit: Startup might be the wrong term but Uber raised money every 18months at least for 10 years till it was finally profitable in 2023. My point is more that saying an unprofitable but massive company only has 18months of cash isn’t a new development. The new development is that the 18 months of cash is an order or two of magnitude more than prior companies.
I feel like they are signing many commitments way beyond that timeline.. some of them possible circular with the possibility to really dent the economy at scale. That’s the part that makes it news.
What does "tech startup" actually mean. Is there a concise, unambiguous definition
Putting aside the ambiguous term "tech", why not just use the term "company"
The term "company" can be defined concisely as, e.g., "an association of persons for the purpose of carrying on some enterprise or business". Note there is no need to refer to an example company
Is it possible to define "startup" without referring to an example startup
Where is the actual financial modelling? This is pure speculation?
I understand being bearish and frightened of AI but this accounts for absolutely NOTHING, and especially doesn't include any projections on potential ad revenue which is likely going to be huge given their DAU and what you can extrapolate their ARPU to be based on other big tech advertisers.
>Where is the actual financial modelling? This is pure speculation?
Every doom and gloom article about OpenAI is almost always speculation, with no actual evidence backing the claims. The issue is that people love a good "AI is going to fail" story, so it gets shot up to the front page. Unfortunately, some journalists now know that it can rake in clicks, so they will happily reduce their journalistic integrity to ride the wave.
One problem with OpenAi advertising is that users are already moving towards Gemeni, which isn't advertising.
Chatgpt is mostly worse than Gemeni too (arguably) and isn't nearly as rate limited. So they're already losing users and making their product a worse experiance than their competition.
Sure OpenAI will make some money from ads but will it be anything close to what it takes to quench the amount of money they're burning? It seems unlikely to me. They really need to be bought out by a sugar-momma who can afford to play this kind of game like MSFT.
OpenAI hit 800M weekly average users and communication to OpenAI investors from this week state:
> "Both our Weekly Active User (WAU) and Daily Active User (DAU) figures continue to produce all-time-highs (Jan 14 was the highest, Jan 13 was the second highest, etc.)"
This does not indicate that they're losing users, at all...
Ahh, maybe they're not losing users after all. I was thinking about market share as reported in several articles. I assumed if they were losing big lumps of market share they had to be losing users too, but I guess you can still grow even so.
How do we know that ad revenue will be huge? 80% of the questions that I ask can't be monetized because they're not about purchase intent. And even if they could, has OpenAI built an auction system to bid on keywords? How exactly will all this work and be streamlined in the next 18 months to the point that it could generate the revenue they need to keep with the ridiculous investment requirements in infrastructure?
The thing i keep coming back to is that an LLM backed query is so, so much more expensive than a typical web request. What kind of advertising is going to align in the value necessary to cover those costs, plus margin? Chatbots aren't YouTube, users aren't going to sit through 30 second ads, I don't think.
From what I understand of the advertising market, companies like Google and Facebook make bucketloads of ads primarily because they own so much of the vertical integration of ad markets. Meanwhile, the way OpenAI appears to be integrating ads makes it seem to me that they're positioned only to take the smallest slice of the pie--a place to hoist ads--which means they're revenue-per-user I would estimate to be a lot closer to, say, a newspaper website than the biggest of social media sites, or maybe along the lines of Twitter or Tumblr, which never posted spectacular profits.
OpenAI have signed something like $1.5tn worth of future spending deals as of the end of last year whilst making something like $13bn of _revenue_ for the year. There's no way that any of this can add up
"signed" $1.5T, or issued press releases that hint to $1.5T in synergistic, cross-collateralized theoretical future deals funded by market frenzy and investor inertia? i.e. how much of their own money has OpenAI committed?
> ad revenue which is likely going to be huge given their DAU and what you can extrapolate their ARPU to be based on other big tech advertisers.
Ad revenue doesn't come out of thin air. Unless budgets and TAM in the ad space increase (hint: they won't), the spend has to mostly come from cannibalizing META and Google. In that regard, I wish them luck - that will be a long and bloody battle. And both the established players can fight it longer than OAI because they have actually revenue streams and strong cash balances.
Altman saying they are going to spend a Trillion+ is (if anything) an anti signal to what the actual financial plan looks like. He is way out front as the hype man and booster. Most of what he says is wishful thinking or an outright lie.
Sometimes these "articles" are sent out as thinly veiled "press releases" prior to an new round of investment. Sometimes someone who thinks they are a "reporter" has what they think is an "exclusive" or a "hot take". Regardless, as someone who has spent all of his career in startups...this is...business as usual. Another round of funding/financing will commence. Open AI will be fine unless investors lose confidence in AI. We won't know how it will play out until it plays out. Media outlets reporting on this are playing off the AI bubble hype for clicks. (Yes, we are in a bubble. No, nobody knows when it will pop, nor how bad it will be, ad driven company just wants more ad revenue, nothing to see here, move along.)
The AI doom and gloom is so weird, and it's just turning into a bizarre echo chamber. AI is orders of magnitude more useful and transformative than Facebook was in 2005, and Meta is now one of the most valuable companies on the planet. Even if OpenAI has a down round or defaults on some loans, the technology has already proven to have dozens upon dozens of practical applications.
Disagree, no one's going to invite me to their kids birthday party via ChatGPT. It's innovation was in ads knowing so much about the people it targeted, and putting tracking pixels on every webpage with a Like button. Facebook was transformative for online surveillance
IMO LLMs will be equally transformative for online influence campaigns (aka ads + Cambridge analytica on steroids)
People are definitely going to be sending you AI generated birthday invite posters soon.
Oh and yeah, AI has already been shown to be more persuasive than the average human. It's only a matter of time before someone's paying to decide what it persuades you of
If only there were some way to avoid this persuasion by, I don't know, not using or relying on such controlled technology, or by not buying in to the hype of all the companies with vested interests in selling it
"Useful and transformative" doesn't mean "financially successful".
A single LLM provider might have been able to get great margins and capture a significant fraction of the total economic output of (currently e.g. junior grade software engineering), but collectively they're in an all-pay auction for the hardware to train models worth paying for, and at the same on questionable margins because they need to compete with each other on cost.
They can all go bankrupt, and leave behind only trained models that normal people won't be able to run for 5 years while consumer-grade stuff catches up. Or any single one of them might win, which may not be OpenAI. Any or all may get state subsidies (US, Chinese, European, whatever).
Paid/API LLM inference is profitable, though. For example, DeepSeek R1 had "a cost profit margin of 545%" [1] (ignoring free users and using a placeholder $2/hour figure H800 GPU, which seems ballpark of real to me due to Chinese electricity subsidies). Dario has said each Anthropic model is profitable over its lifetime. (And looking at ccusage stats and thinking Anthropic is losing thousands per Claude Code user is nonsense, API prices aren't their real costs. That's why opencode gives free access to GLM 4.7 and other models: it was far cheaper than they expected due to the excellent cache hit rates.) If anyone ran out of money they would stop spending on experiments/research and training runs and be profitable... until their models were obsolete. But it's impossible for everyone to go bankrupt.
That's more of "cloud compute makes money" than "AI makes money".
If the models stop being updated, consumer hardware catches up and we can all just run them locally in about 5 years (for PCs, 7-10 for phones), at which point who bothers paying for a hosted model?
| AI is orders of magnitude more useful and transformative than Facebook was in 2005
It better be, it's taken over 40000x the funding.
The question is not whether AI is useful, the question is whether it's useful enough relative to the capital expectations surrounding it. And those expectations are higher than anything the world has ever seen.
They're not arguing that AI sucks. Only that OpenAI has no hope of meeting it's financial obligations which seems pretty reasonable. And very on brand for Sam Altman. It seems pretty obvious at this point that model training is extremely expensive and affords very little moat. LLMs will continue to improve and gain adoption, but one or more companies will fall by the wayside regardless of their userbase. Google seems pretty clearly to be in pole position at this point as they have massive revenue, data, expertise and their own chips.
> AI is orders of magnitude more useful and transformative than Facebook was in 2005
This makes sense because Facebook was one year old in 2005 and OpenAI is 11 years old now. Eleven is just two ones so it’s basically the same thing as one so it is sensible to make that comparison
That wasn't the conclusion of the article (I'm assuming you're reacting to the headline and haven't read the article).
The author's prediction is that OpenAI will get bought.
I know it sounds kind of crazy, but there actually is precedent for that: Reid Hoffman sold his AI startup after he realized there was no possible way for an AI startup to compete with the Googles of the world with $100B in cash and giant free cash flow machines given how capital intensive it is to build AI.[1]
To me, it's not that outlandish to think that if OpenAI really does need to spend a ton of money to survive, they will probably have to either raise or get bought (or find that magic money tree by monetizing their business). Because right now, they don't have the cash flow to compete with Google.
Could obviously change, but I think that's where the author is coming from.
> Investors were briefly spooked last July when an M.I.T. study suggested that almost none of this is useful to businesses. Corporations had poured tens of billions of dollars into A.I., yet only one in 20 projects had succeeded, the study reported. But a Wharton study in October delivered the opposite verdict. After interviewing 801 leaders at U.S. companies, Wharton concluded that three-quarters of the businesses were getting a positive return on their A.I. investments.
MIT actually carried out a study. Wharton just asked some execs, who of course parroted the party line. Winter is coming.
This looks less like an AI failure and more like a compute economics problem. Frontier labs are chasing marginal model gains that require exponentially more GPUs, power, and capex, so burn rates explode even if demand grows. Centralized hyperscale data centers concentrate that risk on a few balance sheets. An alternative is treating AI as a distributed workload problem—using spot or decentralized GPU markets (io.net, Akash, etc.) to tap existing idle capacity instead of financing trillion-dollar builds. You trade enterprise SLAs for lower capex exposure, but structurally it changes the cost curve.
This doesn’t feel like news to me? A tech startup that has 18 months of runway is pretty good honestly. The story is the quantity of cash involved in that runway.
Edit: Startup might be the wrong term but Uber raised money every 18months at least for 10 years till it was finally profitable in 2023. My point is more that saying an unprofitable but massive company only has 18months of cash isn’t a new development. The new development is that the 18 months of cash is an order or two of magnitude more than prior companies.
I feel like they are signing many commitments way beyond that timeline.. some of them possible circular with the possibility to really dent the economy at scale. That’s the part that makes it news.
I don’t think OpenAI could be considered a startup anymore.
Does the term "startup" appear in any dictionaries before 1993
Post-1993, a Wikipedia search for "tech startup" redirects to the page for "startup company". Interestingly, it contains a reference to YC
The definition of "startup company" provided relies on a single reference to a 2013 Forbes article
https://web.archive.org/web/20131217064510if_/http://www.for...
What does "tech startup" actually mean. Is there a concise, unambiguous definition
Putting aside the ambiguous term "tech", why not just use the term "company"
The term "company" can be defined concisely as, e.g., "an association of persons for the purpose of carrying on some enterprise or business". Note there is no need to refer to an example company
Is it possible to define "startup" without referring to an example startup
Is OpenAI really a “tech startup” at this point?
Where is the actual financial modelling? This is pure speculation?
I understand being bearish and frightened of AI but this accounts for absolutely NOTHING, and especially doesn't include any projections on potential ad revenue which is likely going to be huge given their DAU and what you can extrapolate their ARPU to be based on other big tech advertisers.
>Where is the actual financial modelling? This is pure speculation?
Every doom and gloom article about OpenAI is almost always speculation, with no actual evidence backing the claims. The issue is that people love a good "AI is going to fail" story, so it gets shot up to the front page. Unfortunately, some journalists now know that it can rake in clicks, so they will happily reduce their journalistic integrity to ride the wave.
One problem with OpenAi advertising is that users are already moving towards Gemeni, which isn't advertising.
Chatgpt is mostly worse than Gemeni too (arguably) and isn't nearly as rate limited. So they're already losing users and making their product a worse experiance than their competition.
Sure OpenAI will make some money from ads but will it be anything close to what it takes to quench the amount of money they're burning? It seems unlikely to me. They really need to be bought out by a sugar-momma who can afford to play this kind of game like MSFT.
Google is an ad company. I don't think Gemini can remain ad-free for too long
Just see what is happening regarding Youtube ads. It began in a small way, and it's now unbearable to watch without uBlock Origin
Google still has the Search money printer and can subsidize AI for longer.
OpenAI hit 800M weekly average users and communication to OpenAI investors from this week state:
> "Both our Weekly Active User (WAU) and Daily Active User (DAU) figures continue to produce all-time-highs (Jan 14 was the highest, Jan 13 was the second highest, etc.)"
This does not indicate that they're losing users, at all...
Ahh, maybe they're not losing users after all. I was thinking about market share as reported in several articles. I assumed if they were losing big lumps of market share they had to be losing users too, but I guess you can still grow even so.
https://www.bleepingcomputer.com/news/artificial-intelligenc...
How do we know that ad revenue will be huge? 80% of the questions that I ask can't be monetized because they're not about purchase intent. And even if they could, has OpenAI built an auction system to bid on keywords? How exactly will all this work and be streamlined in the next 18 months to the point that it could generate the revenue they need to keep with the ridiculous investment requirements in infrastructure?
The thing i keep coming back to is that an LLM backed query is so, so much more expensive than a typical web request. What kind of advertising is going to align in the value necessary to cover those costs, plus margin? Chatbots aren't YouTube, users aren't going to sit through 30 second ads, I don't think.
As the article pointed out, there is no moat. There’s no reason to think they’ll be more successful with ads than MySpace was versus Facebook.
You'll probably be able to vibe code filtering it out anyway.
From what I understand of the advertising market, companies like Google and Facebook make bucketloads of ads primarily because they own so much of the vertical integration of ad markets. Meanwhile, the way OpenAI appears to be integrating ads makes it seem to me that they're positioned only to take the smallest slice of the pie--a place to hoist ads--which means they're revenue-per-user I would estimate to be a lot closer to, say, a newspaper website than the biggest of social media sites, or maybe along the lines of Twitter or Tumblr, which never posted spectacular profits.
Even if they're at the ARPU of Twitter that's $18-20 (will take them a while to reach it imo). At 800M WAU that's up to $16B in additional revenue.
This is assuming they don't build the vertical stack at all which is unlikely given their highly competent product team.
What other layers are you referencing that OpenAI doesn't have?
OpenAI have signed something like $1.5tn worth of future spending deals as of the end of last year whilst making something like $13bn of _revenue_ for the year. There's no way that any of this can add up
"signed" $1.5T, or issued press releases that hint to $1.5T in synergistic, cross-collateralized theoretical future deals funded by market frenzy and investor inertia? i.e. how much of their own money has OpenAI committed?
Ads will certainly generate revenue but it won't miraculously help OpenAI generate profits.
> ad revenue which is likely going to be huge given their DAU and what you can extrapolate their ARPU to be based on other big tech advertisers.
Ad revenue doesn't come out of thin air. Unless budgets and TAM in the ad space increase (hint: they won't), the spend has to mostly come from cannibalizing META and Google. In that regard, I wish them luck - that will be a long and bloody battle. And both the established players can fight it longer than OAI because they have actually revenue streams and strong cash balances.
Altman saying they are going to spend a Trillion+ is (if anything) an anti signal to what the actual financial plan looks like. He is way out front as the hype man and booster. Most of what he says is wishful thinking or an outright lie.
It honestly doesn't matter.
Sometimes these "articles" are sent out as thinly veiled "press releases" prior to an new round of investment. Sometimes someone who thinks they are a "reporter" has what they think is an "exclusive" or a "hot take". Regardless, as someone who has spent all of his career in startups...this is...business as usual. Another round of funding/financing will commence. Open AI will be fine unless investors lose confidence in AI. We won't know how it will play out until it plays out. Media outlets reporting on this are playing off the AI bubble hype for clicks. (Yes, we are in a bubble. No, nobody knows when it will pop, nor how bad it will be, ad driven company just wants more ad revenue, nothing to see here, move along.)
> Another round of funding/financing will commence.
From where? There's not an infinite source of capital. There's already talk of junk bonds, CDS and private credit.
At some point the lenders look like the loan shark/death spiral finance options.
From an IPO.
I'm gonna silently hope that this means we'll suddenly have extremely cheap GPU and RAM selling out in 2027. Hardware prices have gotten out of hand.
The AI doom and gloom is so weird, and it's just turning into a bizarre echo chamber. AI is orders of magnitude more useful and transformative than Facebook was in 2005, and Meta is now one of the most valuable companies on the planet. Even if OpenAI has a down round or defaults on some loans, the technology has already proven to have dozens upon dozens of practical applications.
Disagree, no one's going to invite me to their kids birthday party via ChatGPT. It's innovation was in ads knowing so much about the people it targeted, and putting tracking pixels on every webpage with a Like button. Facebook was transformative for online surveillance
IMO LLMs will be equally transformative for online influence campaigns (aka ads + Cambridge analytica on steroids)
People are definitely going to be sending you AI generated birthday invite posters soon.
Oh and yeah, AI has already been shown to be more persuasive than the average human. It's only a matter of time before someone's paying to decide what it persuades you of
If only there were some way to avoid this persuasion by, I don't know, not using or relying on such controlled technology, or by not buying in to the hype of all the companies with vested interests in selling it
Agreed, just because something is useful for helpful doesn’t mean it’s easy to monetize.
"Useful and transformative" doesn't mean "financially successful".
A single LLM provider might have been able to get great margins and capture a significant fraction of the total economic output of (currently e.g. junior grade software engineering), but collectively they're in an all-pay auction for the hardware to train models worth paying for, and at the same on questionable margins because they need to compete with each other on cost.
They can all go bankrupt, and leave behind only trained models that normal people won't be able to run for 5 years while consumer-grade stuff catches up. Or any single one of them might win, which may not be OpenAI. Any or all may get state subsidies (US, Chinese, European, whatever).
All kinds of outcomes are possible.
Paid/API LLM inference is profitable, though. For example, DeepSeek R1 had "a cost profit margin of 545%" [1] (ignoring free users and using a placeholder $2/hour figure H800 GPU, which seems ballpark of real to me due to Chinese electricity subsidies). Dario has said each Anthropic model is profitable over its lifetime. (And looking at ccusage stats and thinking Anthropic is losing thousands per Claude Code user is nonsense, API prices aren't their real costs. That's why opencode gives free access to GLM 4.7 and other models: it was far cheaper than they expected due to the excellent cache hit rates.) If anyone ran out of money they would stop spending on experiments/research and training runs and be profitable... until their models were obsolete. But it's impossible for everyone to go bankrupt.
[1] https://github.com/deepseek-ai/open-infra-index/blob/main/20...
That's more of "cloud compute makes money" than "AI makes money".
If the models stop being updated, consumer hardware catches up and we can all just run them locally in about 5 years (for PCs, 7-10 for phones), at which point who bothers paying for a hosted model?
I don’t think the current industry can survive without both frontier training and inference.
Getting rid of frontier training will mean open source models will very quickly catch up. The great houses of AI need to continue training or die.
In any case, best of luck (not) to the first house to do so!
| AI is orders of magnitude more useful and transformative than Facebook was in 2005
It better be, it's taken over 40000x the funding.
The question is not whether AI is useful, the question is whether it's useful enough relative to the capital expectations surrounding it. And those expectations are higher than anything the world has ever seen.
They're not arguing that AI sucks. Only that OpenAI has no hope of meeting it's financial obligations which seems pretty reasonable. And very on brand for Sam Altman. It seems pretty obvious at this point that model training is extremely expensive and affords very little moat. LLMs will continue to improve and gain adoption, but one or more companies will fall by the wayside regardless of their userbase. Google seems pretty clearly to be in pole position at this point as they have massive revenue, data, expertise and their own chips.
> AI is orders of magnitude more useful and transformative than Facebook was in 2005
This makes sense because Facebook was one year old in 2005 and OpenAI is 11 years old now. Eleven is just two ones so it’s basically the same thing as one so it is sensible to make that comparison
Facebook hooked me up with 4 beautiful girlfriends. I don't think Chatgpt is going to do that any time soon.
What a useless article. OpenAI will obviously do many things before "running out of cash" -
1/ Implement more aggressive advertising 2/ Stop training new models 3/ Raise more funding
That wasn't the conclusion of the article (I'm assuming you're reacting to the headline and haven't read the article).
The author's prediction is that OpenAI will get bought.
I know it sounds kind of crazy, but there actually is precedent for that: Reid Hoffman sold his AI startup after he realized there was no possible way for an AI startup to compete with the Googles of the world with $100B in cash and giant free cash flow machines given how capital intensive it is to build AI.[1]
To me, it's not that outlandish to think that if OpenAI really does need to spend a ton of money to survive, they will probably have to either raise or get bought (or find that magic money tree by monetizing their business). Because right now, they don't have the cash flow to compete with Google.
Could obviously change, but I think that's where the author is coming from.
[1]https://www.eesel.ai/blog/inflection-ai
Advertising doesn’t mean free money. Aggressive advertising will almost certainly drive away users.
4) Sell all of those RAM chips they bought
They can't... not in an affordable way anyway. Those are integrated into other hardware they have. They don't buy the ram sticks others can use.
The RAM doesn't exist. What they bought were RAM futures.
You can sell this, although the value would plummet if they did.
Very good outcome.
> 2/ Stop training new models
Google et al won't stop training new models.
And it’s just a restatement of the original, which isn’t any good either. Really mediocre.
Primary source: https://archive.is/Pf1M6
> Investors were briefly spooked last July when an M.I.T. study suggested that almost none of this is useful to businesses. Corporations had poured tens of billions of dollars into A.I., yet only one in 20 projects had succeeded, the study reported. But a Wharton study in October delivered the opposite verdict. After interviewing 801 leaders at U.S. companies, Wharton concluded that three-quarters of the businesses were getting a positive return on their A.I. investments.
MIT actually carried out a study. Wharton just asked some execs, who of course parroted the party line. Winter is coming.
This looks less like an AI failure and more like a compute economics problem. Frontier labs are chasing marginal model gains that require exponentially more GPUs, power, and capex, so burn rates explode even if demand grows. Centralized hyperscale data centers concentrate that risk on a few balance sheets. An alternative is treating AI as a distributed workload problem—using spot or decentralized GPU markets (io.net, Akash, etc.) to tap existing idle capacity instead of financing trillion-dollar builds. You trade enterprise SLAs for lower capex exposure, but structurally it changes the cost curve.
I don't know why I am expecting that to happen earlier with the rate they are burning throgh the dollars by the billions...
No, they can start selling overpriced RAMs; they might even sell it to Nvidia and buy back GPUs.
Given that OpenAI’s strongest skill seems to be fundraising, I could see them transforming into a financial institution.
OpenFI. Still got that jing.
OpenFLI. You read it here first.
OpenFTX. Are you in?
OpenClosed, that'll show'em
Original link: https://www.nytimes.com/2026/01/13/opinion/openai-ai-bubble-...
This is pretty normal for a fast-growing startup, although OpenAI may be the largest to ever do it.
The billions and billions of dollars standing in a lonnnnng queue begging to get in the door at OpenAI suggests Bruno is a little over his skis.
Is there anything we can do to help them reach this goal sooner?
Strategies such as this would accelerate said endeavor:
https://news.ycombinator.com/item?id=46577464
Crash the PC market, & then disappear down the drain. Fascinating business plan.
Question: Why aren't we seeing similar reports for Anthropic?
Amazon can buy them when it is time
Lower profile and they're not making as many crazy deals but they're also losing incredible amounts of money.
I read somewhere (404 media? ) that they spent more on AWS than they got in revenue for 2025.
After getting $8bn in investment from Amazon... (as of nov 2024)
https://old.reddit.com/r/stocks/comments/1qf6z4v/anthropic_3...
They haven't turned on the ads yet
Also Musk is trying to sue them for over $100 billion
https://news.ycombinator.com/item?id=46662986
When reached for comment, Altman responded “Hold my beer”