AI Is Slowing Down
We recorded 23 claims from this post. 17 were checked against evidence. 6 can’t be checked. 24 more are held. Read the post at its source.
Claims from this post (23)
In the order they appear in the post. Each quotation is the speaker’s words in this post. A label such as “Supported” or “Can’t be checked” is the claim’s current assessment; its page has the evidence and the reasons.
Assessment: Can’t be checkedSaid
Rhetoric
“Those excited about generative AI are either the victim or the perpetrator of a con centered around a technology to ingratiate at the highest cost possible.”
Our reading: Generative AI enthusiasm is the product of a con, and its enthusiasts are either its victims or its perpetrators.
No horizon
Assessment: UnresolvedSaid
Forecast
“These funds will be insufficient to cover Anthropic’s costs, as will Anthropic’s cash flow, meaning that it will have to raise at least another $200 billion in the next year.”
Our reading: Because the funds Anthropic raised in early 2026 and its cash flow will not cover its costs, Anthropic will have to raise at least another $200 billion within a year of 2026-06-08.
Not yet due Evidence as of
Limitation: No outcome evidence was gathered, because the deadline has not passed.
Assessment: UnresolvedSaid
Conditional forecast
“To pay for this compute, Anthropic must meet its projected revenue of $174 billion a year by 2029.”
Our reading: Anthropic must reach its projected $174 billion in annual revenue by 2029 to pay for its compute commitments.
Not yet due Evidence as of
Only if: to pay for its compute commitments
Assessment: UnresolvedSaid
Forecast
“These funds will be insufficient to cover Anthropic’s costs, as will Anthropic’s cash flow, meaning that it will have to raise at least another $200 billion in the next year.”
Our reading: Anthropic's $95 billion raised and its cash flow will not cover its costs, so it will have to raise at least another $200 billion within the next year.
Not yet due Evidence as of
Limitation: Management's liquidity outlook does not prove future financing needs or payment of costs.
Assessment: UnresolvedSaid
Forecast
“The $122 billion funding round from March will be insufficient to cover these costs, and it will require, at the very least, another $250 billion in funding by the end of the year.”
Our reading: OpenAI's $122 billion March round will not cover its costs, and it will need at least another $250 billion in funding by the end of 2026.
Not yet due Evidence as of
Limitation: OpenAI's complete payment schedules and cash-flow records were not available in the retrieved evidence.
Assessment: UnresolvedSaid
Conditional forecast
“When I say that, I mean it absolutely has to do that otherwise none of the data center capex makes sense, and neither Anthropic nor OpenAI can pay their commitments.”
Our reading: If AI revenue does not reach over $2 trillion a year by 2030, the data center capex will not make sense and neither Anthropic nor OpenAI will be able to pay their commitments.
Not yet due Evidence as of
Only if: if AI does not generate over $2 trillion in annual revenue by 2030
Assessment: UnresolvedSaid
Observation
“Between them, Anthropic and OpenAI represent the vast majority of all AI compute demand — at a minimum 70%, if not 80% to 90%.”
Our reading: Anthropic and OpenAI together account for at least 70%, and possibly 80-90%, of all AI compute demand.
No horizon Evidence as of
Limitation: The source leaves the demand measure and observation period unspecified.
Assessment: Can’t be checkedSaid
Forecast
“OpenAI expects to spend $50 billion on compute in 2026, and I wouldn’t be surprised if Anthropic spends anywhere from $30 billion to $50 billion.”
Our reading: Anthropic may spend $30 billion to $50 billion on compute in 2026.
No horizon
Assessment: UnresolvedSaid
Conditional forecast
“If we assume that half of that gets built, that’s still $875 billion in annual revenue that will be needed to keep these data centers from running out of money as their margins are atrocious and they’re all paid for with onerous debt.”
Our reading: Even if only half of the planned data center capacity gets built, it will need about $875 billion in annual revenue to avoid running out of money.
No deadline stated Evidence as of
Only if: if half of the planned 190GW gets built
Assessment: UnresolvedSaid
Forecast
“While Anthropic claims it will be profitable by then, I do not believe it will be, nor is it profitable at this point outside of financial engineering.”
Our reading: Anthropic will not be profitable by 2029, despite its claims.
Not yet due Evidence as of
Limitation: The full-year interpretation comes from the annual projections in the surrounding passage.
Assessment: UnresolvedSaid
Conditional forecast
“Both Anthropic and OpenAI need to be making $10 billion or more in monthly revenue by Q1 2028, or their growth rates aren’t going to support their compute commitments.”
Our reading: Unless Anthropic and OpenAI each reach $10 billion or more in monthly revenue by Q1 2028, their growth will not support their compute commitments.
Not yet due Evidence as of
Only if: to support their compute commitments
Assessment: UnresolvedSaid
Conditional forecast
“AI cannot, under any circumstances, slow down. In a year, Anthropic and OpenAI’s businesses have to be roughly twice the size they are today, and then double again basically every year until 2029 or 2030.”
Our reading: AI cannot slow down: Anthropic and OpenAI must roughly double in size within a year and keep doubling about every year until 2029 or 2030.
Not yet due Evidence as of
Only if: to sustain the AI industry's commitments
Assessment: UnresolvedSaid
Conditional forecast
“Also, if OpenAI can’t afford — or doesn’t want — its compute, Oracle will simply run out of money.”
Our reading: If OpenAI cannot afford or does not want its contracted compute, Oracle will run out of money.
No deadline stated Evidence as of
Only if: if OpenAI can't afford or doesn't want its compute
Assessment: UnresolvedSaid
Conditional forecast
“Sidenote: Larry Ellison has also got at least $21 billion in loans collateralized by his Oracle shares, and any doubts around Oracle’s ability to pay for its debts or OpenAI’s ability to pay Oracle for its compute will threaten massive margin calls.”
Our reading: Doubts about Oracle's ability to pay its debts, or about OpenAI's ability to pay Oracle, will threaten massive margin calls on Ellison's share-backed loans.
No deadline stated Evidence as of
Only if: if doubts arise about Oracle's debts or OpenAI's ability to pay Oracle
Assessment: UnresolvedSaid
Conditional forecast
“Anthropic needs to be making over $100 billion in two years or it can’t afford its commitments”
Our reading: Unless Anthropic is making over $100 billion within two years, it cannot afford its commitments.
Not yet due Evidence as of
Only if: to afford its commitments
Assessment: UnresolvedSaid
Observation
“Microsoft’s $37 billion in AI annual run rate — these fucking cowards never share actual AI revenues! — is predominantly made up of OpenAI’s compute, with the rest of it (maybe $8 billion in annual revenue at best) from Microsoft harassing its permanently-abused customer base into installing Copilot.”
Our reading: Microsoft's $37 billion AI annual run rate is mostly OpenAI's compute, with the rest, maybe $8 billion a year at best, coming from pushing customers to install Copilot.
No horizon Evidence as of
Limitation: The relevant composition was not established as publicly known by the evaluation date.
Assessment: Can’t be checkedSaid
Observation
“Sidenote: Amazon and Meta barely have AI stories.”
Our reading: Amazon and Meta barely have AI stories.
No horizon
Assessment: UnresolvedSaid
Observation
“These are signs that AI’s revenue growth is slowing, and it’s likely going to slow further, because we currently live in an era where Anthropic and OpenAI are straight-up abusing its clients, providing limited-to-no visibility into spend, per the Wall Street Journal:”
Our reading: Enterprise token spending caps are signs that AI revenue growth is slowing.
No horizon Evidence as of
Limitation: The speaker supplied neither a growth interval nor an industry accounting definition.
Assessment: Can’t be checkedSaid
Forecast
“These are signs that AI’s revenue growth is slowing, and it’s likely going to slow further, because we currently live in an era where Anthropic and OpenAI are straight-up abusing its clients, providing limited-to-no visibility into spend, per the Wall Street Journal:”
Our reading: AI revenue growth will likely slow further, because Anthropic and OpenAI give customers little or no visibility into what they are spending.
No horizon
Qualifier: because Anthropic and OpenAI give customers little visibility into spend
Assessment: UnresolvedSaid
Observation
“I’m sorry to be a little surly and dismissive, but the AI industry has burned over a trillion dollars and I’ve spent two years being told I’m a luddite and an ape for not celebrating it.”
Our reading: The AI industry has burned over a trillion dollars.
No horizon Evidence as of
Limitation: No comparable cumulative total or documented lower bound through June 8, 2026 was established.
Assessment: Can’t be checkedSaid
Causal claim
“I think that’s by design. AI labs want their costs to be high so that they can continue growing at ridiculous rates, all so that they can keep feeding money to their hyperscaler compute partners who then invest that money right back into them, creating further reasons to keep buying NVIDIA GPUs, so that NVIDIA can then invest that money back into either AI compute providers (who OpenAI and Anthropic pay) or the AI labs themselves.”
Our reading: He thinks it is by design that AI labs want their costs to be high, so they can keep growing fast and feeding money to hyperscaler compute partners, who invest it back into the labs and buy more NVIDIA GPUs, while NVIDIA in turn invests in compute providers or the labs.
No horizon
Qualifier: I think
Assessment: UnresolvedSaid
Causal claim
“Your endless death march of “do as much AI as possible or we’ll fire you” and forcing them to use these tools day-in-day-out has radicalized them against you.”
Our reading: Executive mandates forcing tech employees to use AI tools have turned those employees against their leadership.
No horizon Evidence as of
Limitation: Evidence of resistance is stronger than evidence identifying its cause or prevalence.
Assessment: Can’t be checkedSaid
Forecast
“One of my sources has come forward and brought me a story that will possibly burst the AI bubble.”
Our reading: A story he expects to publish within two weeks may burst the AI bubble.
No horizon
Qualifier: once his upcoming story is published
Held from this post (24)
These records did not pass every check, so nothing about them is published. Held is not a finding.
- The two reviews reached different assessments. Our reading, not assessed: If OpenAI or Anthropic sought profitability or sustainability (assuming either were possible), demand for AI compute would fall, reducing demand for Azure, Google Cloud, AWS, CoreWeave and Oracle Cloud Infrastructure, and in turn for NVIDIA GPUs.
- The two reviews read the claim differently
- The record does not meet the current rules. Our reading, not assessed: If Huang's per-gigawatt cost figures hold, the 190GW of planned data centers will cost $9.5 trillion to $15 trillion.
- The two reviews reached different assessments. Our reading, not assessed: 54% of NVIDIA's revenue comes from three clients, so NVIDIA's future largely depends on those three unnamed companies (likely Taiwanese ODMs building servers for Microsoft, Google and Meta) and on their counterparties' ability to keep raising debt almost indefinitely, as fewer firms can afford to buy thousands of $7.8 million Vera Rubin racks.
- The two reviews read the claim differently
- The two reviews read the claim differently
- The two reviews read the claim differently
- The criteria are ambiguous. Our reading, not assessed: Current AI revenues are either chaotically experimental or heavily subsidized, with labs giving away hundreds of dollars per user.
- The criteria are ambiguous. Our reading, not assessed: Anthropic is not currently profitable except through financial engineering.
- The criteria are ambiguous. Our reading, not assessed: Generative AI is inherently limited, impossibly expensive and economically unviable.
- The two reviews read the claim differently
- The two reviews read the claim differently
- The two reviews read the claim differently
- The two reviews read the claim differently
- The criteria are ambiguous. Our reading, not assessed: Justifying the AI buildout requires at least another $250 billion in annual compute demand, which likely means at least two more OpenAI- or Anthropic-scale companies.
- The criteria are ambiguous. Our reading, not assessed: Because OpenAI and Anthropic are racing toward IPOs and will have to look like real companies, both will inevitably start charging customers more, and will very likely move the vast majority of customers to token-based billing and either kill or vastly limit their subsidized subscriptions.
- The criteria are ambiguous. Our reading, not assessed: Outside of OpenAI and Anthropic, which lose billions or tens of billions a year, there is barely a few billion dollars of AI compute demand.
- The two reviews reached different assessments. Our reading, not assessed: Even if Anthropic and OpenAI spent $500 billion a year on compute, which they cannot afford, at least another $250 billion a year in compute revenue would be needed to justify the buildout.
- The criteria are ambiguous. Our reading, not assessed: Enterprise spending caps and doubts about cost are signs that AI is slowing down.
- The two reviews read the claim differently
- The criteria are ambiguous. Our reading, not assessed: AI is extremely expensive, and its costs for both providers and customers have only ever risen despite promises that they would fall.
- The two reviews read the claim differently
- The two reviews read the claim differently
- The criteria are ambiguous. Our reading, not assessed: Generative AI services cost too much to run, their makers have no path to profitability, and the cost and ROI of LLM tasks cannot be measured.