Claims checked against evidence, page 6 of 29
Claims by Gary Marcus and Ed Zitron. Deadlines are judged as of 2026-10-29.
Claims checked against evidence (2859)
Each of these claims was checked against dated evidence.
Page 6 of 29: claims 501 to 600 of 2859, ordered by the date first said, newest first. To find a claim by its words, use search.
Assessment: UnresolvedSaid
Ed Zitron Observation
“Sidenote: In all honesty, I have no idea how Meta makes this make sense, as it plans to invest at least $125 billion in capex in 2026 and has, to this point, not shown any actual, real growth in its revenue from AI, and no, those increases in conversion don’t mean actual revenue.”
Our reading: Meta has not shown real revenue growth from AI despite planning at least $125B of 2026 capex.
No horizon Evidence as of
Limitation: Capex guidance is not realized expenditure or exclusively AI expenditure.
Source: AI Is Too Expensive
Assessment: UnresolvedSaid
Ed Zitron Causal claim
“The problem is simple: nobody actually knows how much AI is going to cost them in any given quarter. This means that the current token spend you’re seeing is entirely experimental, which is why organizations keep burning through their tokens so fast.”
Our reading: Nobody actually knows how much AI is going to cost them in any given quarter, so current token spend is entirely experimental, which is why organizations keep burning through their tokens so fast.
No horizon Evidence as of
Limitation: The budget-overrun account is reporting, not an independently inspected company ledger.
Source: AI Is Too Expensive
Assessment: UnresolvedSaid
Ed Zitron Causal claim
“Earlier in the week, carnival barker and Salesforce CEO Marc Benioff said his company would spend $300 million on Anthropic tokens in 2026, and as I discussed in my premium from Friday, unrestrained AI spending is inflating the revenues of Anthropic and OpenAI in a way that isn’t sustainable for anybody involved:”
Our reading: Unrestrained AI spending is inflating the revenues of Anthropic and OpenAI in a way that isn't sustainable for anybody involved.
No horizon Evidence as of
Limitation: The survey and spending panel do not provide matched customer budgets, invoices, returns and provider accounts.
Source: AI Is Too Expensive
Assessment: UnresolvedSaid
Ed Zitron Observation
“Okay, fine. Microsoft also has 20 million Microsoft 365 Copilot subscribers for an absolute maximum revenue of $7.2 billion…if every single one were paying $30 a month, which they are most assuredly not as Microsoft has been offering discounts on it for years.”
Our reading: Microsoft 365 Copilot's 20 million subscribers yield at most $7.2B a year, and actually less due to discounts.
No horizon Evidence as of
Limitation: The hypothetical calculation is verifiable; the stronger claim about the actual business remains unverified.
Source: AI Is Too Expensive
Assessment: UnresolvedSaid
Ed Zitron Observation
“Based on my reporting from last year, Microsoft made around $7.5 billion from OpenAI’s inference spend and $761 million from its revenue share in Fiscal Year 2025, a year when it invested (either spent or allocated) around $88.2 billion in capital expenditures.”
Our reading: In FY2025 Microsoft made ~$7.5B from OpenAI inference and $761M from revenue share while investing ~$88.2B in capex.
No horizon Evidence as of
Limitation: The underlying reporting documents were not available for inspection.
Source: AI Is Too Expensive
Assessment: UnresolvedSaid
Ed Zitron Observation
“I’d estimate it needs $5 billion a year for its compute deal xAI (so $20 billion over the total period) and an estimated $30 billion to cover its deal with CoreWeave. That brings us to a total of $380 billion.”
Our reading: By Zitron's partly estimated tally, Anthropic needs about $380 billion over the next four years: $330 billion in cloud obligations to Amazon, Google and Microsoft, an estimated $20 billion for its xAI compute deal, and an estimated $30 billion for its CoreWeave deal.
No horizon Evidence as of
Limitation: The reported Google and AWS durations do not establish how payments fall within the four-year window [e1].
Source: AI Is Too Expensive
Assessment: UnresolvedSaid
Ed Zitron Observation
“This means that around 30% of Microsoft’s capex ($87 billion) went to building OpenAI’s infrastructure.”
Our reading: Around 30% (~$87B) of Microsoft's capex since FY2023 went to building OpenAI's infrastructure.
No horizon Evidence as of
Limitation: The testimony is available here through news reporting rather than a primary court transcript.
Source: AI Is Too Expensive
Assessment: UnresolvedSaid
Ed Zitron Causal claim
“In some cases, the pressure to use AI for everything is turning companies’ software stacks into slop.”
Our reading: In some cases, pressure to use AI for everything is degrading companies' software stacks.
No horizon Evidence as of
Limitation: The accessible research summary does not expose company-level mandates and incident records.
Source: AI Is Too Expensive
Assessment: UnresolvedSaid
Ed Zitron Observation
“Based on Microsoft’s own disclosures, without Anthropic and OpenAI’s additions, RPO would have been effectively flat, as evidenced by the fact that in Q3FY26, remaining performance obligations sat at $627 billion.”
Our reading: Excluding OpenAI and Anthropic deals, Microsoft's RPO would have been effectively flat.
No horizon Evidence as of
Limitation: The calendar labels in the post are inconsistent with the named fiscal quarters.
Source: AI Is Too Expensive
Assessment: UnresolvedSaid
Ed Zitron Observation
“You can argue that “actually GPUs are profitable to run” (I disagree!), but for any of this to make sense, four things have to happen:”
Our reading: GPUs are not profitable to run.
No horizon Evidence as of
Limitation: The accounts do not isolate GPU-service revenue and all attributable costs [e1, e2].
Source: AI Is Too Expensive
Assessment: UnresolvedSaid
Ed Zitron Conditional forecast
“You are not recovering these investments without either a massive new product line that doesn’t exist today or three or four Anthropic or OpenAI-sized compute contracts.”
Our reading: Hyperscalers will not recover their AI investments without a massive new product line or three or four more Anthropic/OpenAI-sized compute contracts.
No deadline stated Evidence as of
Only if: unless a massive new product line or three or four lab-sized compute contracts emerge
Source: AI Is Too Expensive
Assessment: SupportedSaid
Gary Marcus Observation
“The AI trial of the century has just ended with a procedural whimper rather a bang; the jury agreed that Musk was too late but never weighed in on the questions of whether OpenAI did was legitimate.”
Our reading: The Musk v. OpenAI trial ended on procedural grounds: the jury found Musk's claim too late and did not rule on whether OpenAI's conduct was legitimate.
No horizon Evidence as of
Limitation: The completed verdict form was not readable; the assessment uses readable court records for cross-checking and news reports for the quoted outcome evidence.
Assessment: SupportedSaid
Gary Marcus Observation
“The US has 1,200 AI bills (roughly 150 enacted into law), but nothing that feels like a coherent AI policy.”
Our reading: The US has 1,200 AI bills, roughly 150 of them enacted into law, but nothing that feels like a coherent AI policy.
No horizon Evidence as of
Limitation: The verdict depends on reading the abbreviated post with its linked context.
Source: US AI policy is a clumsy mess. Here’s what to do about it.
Assessment: UnresolvedSaid
Gary Marcus Forecast
“We can be absolutely sure that the task length “time horizon” for AI is not going to keep doubling until “time horizons will be 580 times the age of the universe” as Lisan al-Ghaib joked.”
Our reading: We can be absolutely sure that the AI task-length 'time horizon' will not keep doubling until time horizons reach 580 times the age of the universe.
No deadline stated Evidence as of
Limitation: The threshold lies far beyond the capabilities measured in the cited report.
Source: Misplaced panic over AI progress
Assessment: UnresolvedSaid
Ed Zitron Observation
“Every single one of these companies is talking about AI non-stop, and not one of them can show a profit.”
Our reading: Every one of the big tech companies under discussion is talking about AI non-stop, and not one of them can show a profit from it.
No horizon Evidence as of
Limitation: A literal claim that these corporations have no profit would be a different proposition from the contextual AI-specific reading.
Source: Am I Meant To Be Impressed?
Assessment: UnresolvedSaid
Ed Zitron Observation
“I also want to be clear about something else: I know, from documents viewed by this publication, that Microsoft has these line items fully itemized, and could share them if it wanted to, but intentionally chooses not to.”
Our reading: Microsoft has fully itemized AI revenue line items and intentionally chooses not to disclose them.
No horizon Evidence as of
Limitation: Public disclosures cannot establish the contents of unavailable internal documents.
Source: Am I Meant To Be Impressed?
Assessment: UnresolvedSaid
Ed Zitron Observation
“Anthropic and OpenAI make up the vast majority of all AI revenues and compute capacity. I estimate 70% of all revenues and capacity demand, if not higher.”
Our reading: Anthropic and OpenAI make up the vast majority of all AI revenues and compute capacity, which he estimates at 70% of all revenues and capacity demand, if not higher.
No horizon Evidence as of
Limitation: The compute counterevidence is explicitly retained as context because its measurement period differs.
Source: Am I Meant To Be Impressed?
Assessment: UnresolvedSaid
Ed Zitron Forecast
“OpenAI Represents 70%+ Of Microsoft’s AI Revenue and 80%+ Of Its AI GPU Compute Capacity, Creating The Illusion Of Growth That’s Dependent On A Company That Will Lose $25 Billion+ In 2026”
Our reading: OpenAI represents 70% or more of Microsoft's AI revenue and 80% or more of its AI GPU compute capacity, creating an illusion of growth that depends on a company that will lose $25 billion or more in 2026.
Not yet due Evidence as of
Limitation: One claim combines three financial or operational assertions with a qualitative interpretation.
Source: Am I Meant To Be Impressed?
Assessment: UnresolvedSaid
Ed Zitron Observation
“Outside of OpenAI and Anthropic, there’s barely any actual demand for AI services or AI compute at the scale necessary to substantiate a trillion or more in capital expenditures.”
Our reading: Outside OpenAI and Anthropic there is barely any AI demand at the scale needed to justify a trillion dollars in capex.
No horizon Evidence as of
Limitation: The demand evidence consists of company disclosures, not an independently reconciled market total.
Source: Am I Meant To Be Impressed?
Assessment: UnresolvedSaid
Ed Zitron Observation
“Today I can report, based on discussions with sources with direct knowledge of Azure revenue, that in Q2 FY2025, Microsoft brought in around $325.2 million in revenue via renting out GPUs and other AI infrastructure, and around $367 million in revenue from Microsoft 365 Copilot, or less than half of the $1.467 billion that OpenAI spent on inference.”
Our reading: In Q2 FY2025 Microsoft made about $325.2M from AI infrastructure rental and $367M from M365 Copilot, less than half of OpenAI's $1.467B inference spend.
No horizon Evidence as of
Limitation: Neither the revenue breakdown nor its category definitions were independently established.
Source: Am I Meant To Be Impressed?
Assessment: UnresolvedSaid
Ed Zitron Forecast
“By the end of 2027, big tech will have sunk $2 trillion into AI capex, with very little to show for it.”
Our reading: By the end of 2027 big tech will have spent $2 trillion on AI capex with very little to show for it.
Not yet due Evidence as of
Limitation: Total corporate capex cannot automatically be classified as AI capex.
Source: Am I Meant To Be Impressed?
Assessment: UnresolvedSaid
Ed Zitron Observation
“At the current rate of sales, it’s taking six months to install a quarter’s GPUs. At this point it’s obvious that there are warehouses of these things.”
Our reading: It takes about six months to install a quarter's GPU sales, so uninstalled GPUs are piling up in warehouses.
No horizon Evidence as of
Limitation: The evidence covers one operator's power-capacity additions, not industry GPU installation throughput.
Source: Am I Meant To Be Impressed?
Assessment: UnresolvedSaid
Ed Zitron Causal claim
“OpenAI and Anthropic both lose tens of billions of dollars a year (yes, Anthropic said it’ll lose $11 billion in a projection, and I believe they are being coy with their actual losses), which means that the majority of AI revenue and compute demand is dependent on whether Anthropic and OpenAI can continue to raise money.”
Our reading: OpenAI and Anthropic each lose tens of billions a year, so most AI revenue and compute demand depends on their ability to keep raising money.
No horizon Evidence as of
Limitation: OpenAI evidence is reporting about financial records, not a directly accessible primary statement.
Source: Am I Meant To Be Impressed?
Assessment: UnresolvedSaid
Ed Zitron Observation
“While I only had half the fourth quarter’s numbers, I estimate that OpenAI’s annualized spend hit over $18.5 billion — or around $4.6 billion a quarter — by the end of the year, and that’s not accounting for things like Sora 2 or the launch of its Codex coding platform.”
Our reading: He estimates, from only half of Q4's numbers, that OpenAI's annualized Azure inference spend hit over $18.5 billion (around $4.6 billion a quarter) by the end of 2025, not counting things like Sora 2 or the Codex launch.
No horizon Evidence as of
Limitation: The contextual source reports audited statements; those statements were not independently inspected.
Source: Am I Meant To Be Impressed?
Assessment: UnresolvedSaid
Ed Zitron Observation
“Most of Amazon, Google and Microsoft’s capex is being driven into capacity mostly used by OpenAI and Anthropic, neither of whom have the money to pay without continual infusions of more capital.”
Our reading: Most of Amazon, Google and Microsoft's capex is going into capacity mostly used by OpenAI and Anthropic, neither of which has the money to pay without continual infusions of more capital.
No horizon Evidence as of
Limitation: The claim combines expenditure allocation, capacity use and financing dependence.
Source: Am I Meant To Be Impressed?
Assessment: UnresolvedSaid
Ed Zitron Observation
“These fucknuts have spent over a trillion dollars on AI and all they have to show for it is either nothing, vague statements about “everything lifting because of AI,” or pathetic revenues that only get worse the more you think about them.”
Our reading: Big tech has spent over $1 trillion on AI with only nothing, vague claims, or pathetic revenues to show for it.
No horizon Evidence as of
Limitation: The cumulative spending period is ambiguous.
Source: Am I Meant To Be Impressed?
Assessment: UnresolvedSaid
Ed Zitron Observation
“Outside of those two, the only other competitive AI businesses are those of Amazon, Microsoft and Google — two of which now have annualized AI revenues of around 6% of their capital expenditures so far.”
Our reading: Outside OpenAI and Anthropic, the only other competitive AI businesses are those of Amazon, Microsoft and Google, two of which now have annualized AI revenues of around 6% of their capital expenditures so far.
No horizon Evidence as of
Limitation: The packet does not establish cumulative capex starting dates, finance-lease treatment, or allocation between AI and other activities. [e1, e2]
Source: Am I Meant To Be Impressed?
Assessment: UnresolvedSaid
Ed Zitron Observation
“I’ve also confirmed that Microsoft’s “Fairwater” data centers — which constitute (when finished) “hundreds of thousands of GPUs” — are entirely reserved for OpenAI.”
Our reading: Microsoft's Fairwater data centers are entirely reserved for OpenAI.
No horizon Evidence as of
Limitation: The strongest counterevidence concerns intended use or later likely users.
Source: Am I Meant To Be Impressed?
Assessment: UnresolvedSaid
Ed Zitron Observation
“Microsoft desperately wants you to think that this is a diverse, booming revenue stream, when in fact it’s spent around $293 billion in four years to make — when you remove OpenAI — less than $3 billion a quarter in revenue, not profit.”
Our reading: Excluding OpenAI, Microsoft makes under $3 billion a quarter in AI revenue after spending about $293 billion in four years.
No horizon Evidence as of
Limitation: The claim combines measures with different periods and reporting bases.
Source: Am I Meant To Be Impressed?
Assessment: ContradictedSaid
Ed Zitron Observation
“It’s a pretty sweet deal for Google! Google pays Broadcom to develop TPUs, Anthropic pays Google to buy those TPUs once Broadcom builds them, Google installs those TPUs in a data center, and then Anthropic pays Google to rent them back.”
Our reading: Google and Anthropic run a circular arrangement where Anthropic buys Google TPUs and then rents them back from Google.
No horizon Evidence as of
Limitation: The detailed hardware purchase chain relies partly on reporting because the executed exhibit was unavailable through read_page.
Source: Am I Meant To Be Impressed?
Assessment: UnresolvedSaid
Ed Zitron Conditional forecast
“To be clear, I don’t think any of these companies die as a result of the bubble bursting”
Our reading: Meta, Microsoft, Google and Amazon will not die as a result of the AI bubble bursting.
No deadline stated Evidence as of
Only if: if the AI bubble bursts
Source: Am I Meant To Be Impressed?
Assessment: UnresolvedSaid
Ed Zitron Observation
“I’m gonna be honest, I think Anthropic is outright misleading its investors if it’s saying that it will only burn $11 billion in 2026 and 2027, per The Information”
Our reading: He thinks Anthropic is outright misleading its investors if it is saying it will only burn $11 billion in 2026 and 2027.
No horizon Evidence as of
Limitation: The evidence is a secondary account, not authenticated investor materials or internal forecasts. [e1]
Source: Am I Meant To Be Impressed?
Assessment: UnresolvedSaid
Ed Zitron Observation
“Based on my own reporting from direct Azure revenue numbers, this would make OpenAI’s $2.947 billion in inference spend in that quarter around 71% ($11.7bn) of Microsoft’s Q3 FY2025 AI revenue run rate.”
Our reading: Based on his own reporting from direct Azure revenue numbers, OpenAI's $2.947 billion in inference spend that quarter (about $11.7 billion annualized) would be around 71% of Microsoft's Q3 FY2025 AI revenue run rate.
No horizon Evidence as of
Limitation: Arithmetic consistency does not authenticate the spending figure.
Source: Am I Meant To Be Impressed?
Assessment: UnresolvedSaid
Ed Zitron Causal claim
“In fact, I’d argue that Anthropic is having to raise this money to pay up front for capacity that’s yet to be built.”
Our reading: He would argue that Anthropic is having to raise this money to pay up front for capacity that has yet to be built.
No horizon Evidence as of
Limitation: The disclosures do not match specific payment deadlines to construction milestones or trace fundraising proceeds to those obligations. [e1, e2]
Source: Am I Meant To Be Impressed?
Assessment: UnresolvedSaid
Ed Zitron Forecast
“Let’s get specific: 2023, Microsoft, Google, Amazon, and Meta have spent a little over $850 billion in capex, mostly hoarding NVIDIA GPUs that will be near-to-completely obsolete by 2030.”
Our reading: The NVIDIA GPUs big tech bought with ~$850B capex since 2023 will be nearly completely obsolete by 2030.
Not yet due Evidence as of
Limitation: NVIDIA has a commercial interest in emphasizing hardware longevity.
Source: Am I Meant To Be Impressed?
Assessment: SupportedSaid
Gary Marcus Observation
“Additionally, in the new study 89.4% of agents showed drift relative to their goals after about 30 steps in their process, and 94% of agents with some form of memory-augmentation were vulnerable to poisoning attacks.”
Our reading: In the new study, 89.4% of agents drifted from goals after about 30 steps, and 94% of memory-augmented agents were vulnerable to poisoning.
No horizon Evidence as of
Limitation: The primary paper was not found. The assessment rests on news reports.
Assessment: SupportedSaid
Gary Marcus Observation
“A new study from researchers from Stanford, MIT CSAIL, Carnegie Mellon, ITU Copenhagen, and NVIDIA and Elloe AI Labs, examining 847 autonomous agent deployments drawn from healthcare, finance, customer service and code-generation, showed that 91% were vulnerable to subtle but dangerous tool-chaining attacks (see diagram below),”
Our reading: A new study found 91% of 847 autonomous agent deployments were vulnerable to tool-chaining attacks.
No horizon Evidence as of
Limitation: The primary paper was not found online. The assessment rests on news reports.
Assessment: SupportedSaid
Gary Marcus Observation
“The new paper very much amplifies what a team of AWS and Berkeley researchers showed in February, documenting similar vulnerabilities:”
Our reading: The new findings on agent vulnerabilities reinforce similar vulnerabilities documented by AWS and Berkeley researchers in February 2026.
No horizon Evidence as of
Limitation: Marcus does not name the February paper; it is identified as STAC by inference.
Assessment: SupportedSaid
Gary Marcus Observation
“It’s already splitting both parties.”
Our reading: Frustration with AI is already splitting both parties.
No horizon Evidence as of
Limitation: These examples do not establish party-wide realignment, formal fractures, or electoral consequences.
Source: The growing AI backlash
Assessment: UnresolvedSaid
Gary Marcus Observation
“Practically every day I see new signs that the backlash against AI is growing.”
Our reading: Public backlash against AI is growing, with new signs appearing nearly every day.
No horizon Evidence as of
Limitation: The result depends partly on how much weight is given to practically every day.
Source: The growing AI backlash
Assessment: UnresolvedSaid
Gary Marcus Causal claim
“Some of this is technical (LLMs aren’t reliable), some of it is political/economic (such as the utter lack of responsible regulation).”
Our reading: Some of generative AI's downsides have technical causes, such as LLMs being unreliable, and some have political or economic causes, such as the utter lack of responsible regulation.
No horizon Evidence as of
Limitation: One adjudicated incident cannot measure the general reliability of all LLMs.
Source: The growing AI backlash
Assessment: UnresolvedSaid
Gary Marcus Causal claim
“GenAI has been undermining secondary and college education, opening up mass surveillance, increasing disinformation, delusions, impersonation, phishing, and other forms of cybercrime, nonconsensual deep fake porn, bias in employment and other domains, and economic disparity, drowning the world in slop and unwanted, over-leveraged environment-damaging data centers that risk causing a recession.”
Our reading: Generative AI has been undermining secondary and college education, opening up mass surveillance, increasing disinformation, delusions, impersonation, phishing and other cybercrime, nonconsensual deepfake porn, bias in employment and other domains, and economic disparity, and drowning the world in slop and in unwanted, over-leveraged, environment-damaging data centers that risk causing a recession.
No horizon Evidence as of
Limitation: The search and verified evidence do not cover every listed harm adequately.
Source: The growing AI backlash
Assessment: SupportedSaid
Gary Marcus Observation
“The problem of course is that LLMs are mimics, and what they say isn’t always true.”
Our reading: LLMs are mimics and what they say is not always true.
No horizon Evidence as of
Limitation: The support for 'mimics' uses the ordinary behavioral reading specified in the contract.
Assessment: SupportedSaid
Gary Marcus Observation
“Realism re AI coding is knowing that next-word prediction gets us a surprisingly long way in writing code, but less far in making sure that code is robust.”
Our reading: Next-word prediction gets us a surprisingly long way in writing code, but less far in making sure that code is robust.
No horizon Evidence as of
Limitation: The April results come from a security vendor’s own research announcement; commercial interests warrant caution, and the packet provides no independent replication [e1].
Assessment: UnresolvedSaid
Gary Marcus Conditional forecast
“If the trial is cast as a referendum on Musk, rather than on OpenAI’s founders, he will lose.”
Our reading: If the trial is framed as a referendum on Musk rather than OpenAI's founders, Musk will lose the lawsuit against OpenAI.
No deadline stated Evidence as of
Only if: trial is cast as a referendum on Musk rather than OpenAI's founders
Assessment: UnresolvedSaid
Ed Zitron Observation
“As of April 27 2026, only two buildings are operational and generating revenue, and the third barely has any IT gear in it. I estimate the total cost of Stargate Abilene to be around $52.8 billion.”
Our reading: As of April 27, 2026, only two Stargate Abilene buildings are operational and generating revenue, the third barely has any IT gear, and the author estimates the total cost at around $52.8 billion.
No horizon Evidence as of
Limitation: The operational disclosures precede April 27 and do not establish billing on that date.
Source: AI's Economics Don't Make Sense
Assessment: UnresolvedSaid
Ed Zitron Observation
“This is the kind of thing that happens before a company starts dying. OpenAI’s growth is slowing at precisely the time it needs to accelerate. It needs to effectively 10x its current business by 2030 to make its obligations.”
Our reading: OpenAI's growth is slowing at precisely the time it needs to accelerate, since it needs to effectively 10x its current business by 2030 to meet its obligations, and this is the kind of thing that happens before a company starts dying.
No horizon Evidence as of
Limitation: The slowdown account is secondhand and lacks comparable growth-rate measurements.
Source: AI's Economics Don't Make Sense
Assessment: UnresolvedSaid
Ed Zitron Observation
“I really need to be clear that Oracle has no other path to making this revenue without OpenAI, and these projects are entirely financed and paid for using the projected cashflow of the data centers themselves.”
Our reading: Oracle has no other path to making this revenue without OpenAI, and these projects are entirely financed and paid for using the projected cashflow of the data centers themselves.
No horizon Evidence as of
Limitation: Complete project contracts, guarantees and repayment models are absent.
Source: AI's Economics Don't Make Sense
Assessment: UnresolvedSaid
Ed Zitron Observation
“This means that a large part — if not the majority of — AI compute revenue is dependent on a continual flow of venture capital and debt, both of which are only made possible by investors that still believe that generative AI will be the biggest, most hugest thing in the world.”
Our reading: A large part, perhaps most, of AI compute revenue depends on continued venture capital and debt.
No horizon Evidence as of
Limitation: The provider disclosure covers 2025 and is context only.
Source: AI's Economics Don't Make Sense
Assessment: UnresolvedSaid
Ed Zitron Forecast
“There has never been — and never will be — an economically-feasible way to offer services powered by LLMs without charging the actual token burn of each user, and in the process of deceiving said users, these companies have created products with illusory benefits and questionable return on investment.”
Our reading: There never has been and never will be an economically feasible way to offer LLM services without charging each user's actual token burn.
No deadline stated Evidence as of
Limitation: Private-company profitability statements may be credible without providing enough detail for this particular test.
Source: AI's Economics Don't Make Sense
Assessment: UnresolvedSaid
Ed Zitron Forecast
“I think it’s inevitable that the majority of AI subscriptions move to token-based billing, especially as both Anthropic and OpenAI have now done so with their enterprise customers.”
Our reading: The majority of AI subscriptions will inevitably move to token-based billing.
No deadline stated Evidence as of
Limitation: No representative denominator for AI subscription products is established.
Source: AI's Economics Don't Make Sense
Assessment: UnresolvedSaid
Ed Zitron Observation
“It’s also unclear whether the customer base for AI compute exists outside of OpenAI and Anthropic, whose demand accounts for 50% of AI data centers under construction, creating a massive systemic weakness if either of them lacks the money to pay.”
Our reading: OpenAI and Anthropic account for 50% of AI data centers under construction, a systemic weakness if either cannot pay, and demand outside them is unclear.
No horizon Evidence as of
Limitation: Known contracts do not reveal the whole construction pipeline.
Source: AI's Economics Don't Make Sense
Assessment: UnresolvedSaid
Ed Zitron Forecast
“I should also be clear that Oracle is mortgaging its entire fucking future on projects like Stargate Abilene, incurring billions of dollars of costs up front for a business that will take years to turn a profit even if OpenAI makes every single payment in a timely manner.”
Our reading: Oracle is staking its future on Stargate projects that will take years to turn a profit even if OpenAI pays on time.
No deadline stated Evidence as of
Limitation: Project accounts may be private.
Source: AI's Economics Don't Make Sense
Assessment: UnresolvedSaid
Ed Zitron Forecast
“The people that have so aggressively worked together to hold up this industry have only delayed its inevitable fall.”
Our reading: The AI industry will inevitably fall; its supporters have only delayed it.
No deadline stated Evidence as of
Limitation: Fall and the industry's boundaries are not precisely defined.
Source: AI's Economics Don't Make Sense
Assessment: UnresolvedSaid
Ed Zitron Observation
“It is unclear if that’s 200MW of critical IT capacity or the total available power at the Abilene campus, and in any case, this is only enough power for two buildings, which means that Oracle is most decidedly not “on schedule.””
Our reading: Oracle's Stargate Abilene is behind schedule, with only enough power for two buildings.
No horizon Evidence as of
Limitation: The 200MW disclosure does not specify critical IT load versus total facility power [e1].
Source: AI's Economics Don't Make Sense
Assessment: UnresolvedSaid
Ed Zitron Forecast
“I truly believe that LLM-based subscriptions are going to go away entirely, at least at the scale of any product that generates code, and in doing so, Amodei and Altman will wrap up their con, or at least believe that they have.”
Our reading: LLM-based subscriptions will go away entirely, at least for code-generating products.
No deadline stated Evidence as of
Limitation: Supplier-level changes cannot establish a market-wide outcome.
Source: AI's Economics Don't Make Sense
Assessment: UnresolvedSaid
Ed Zitron Observation
“Of the 114GW of data centers supposedly being built by the end of 2028, only 15.2GW is under construction in any way, shape, or form.”
Our reading: Of the 114GW of AI data centers announced for completion by end of 2028, only 15.2GW is under construction.
No horizon Evidence as of
Limitation: The evidence item provides no April project inventory reproducing both claimed totals [e1].
Source: AI's Economics Don't Make Sense
Assessment: UnresolvedSaid
Ed Zitron Observation
“Either way, Oracle is insufficiently-capitalized to finish Stargate Abilene. It will need at least another $150 billion to get this done, and that’s assuming that other partners pick up about $30 billion in costs.”
Our reading: Oracle is insufficiently capitalized to finish Stargate Abilene and will need at least another $150 billion to do so, assuming other partners pick up about $30 billion in costs.
No horizon Evidence as of
Limitation: The available financial baseline predates the post.
Source: AI's Economics Don't Make Sense
Assessment: UnresolvedSaid
Ed Zitron Causal claim
“And that day has finally arrived, because every single AI service you use subsidized compute, and every single service is losing money as a result:”
Our reading: Every single AI service you use subsidized compute, and every single one is losing money as a result.
No horizon Evidence as of
Limitation: The universal population is not precisely defined.
Source: AI's Economics Don't Make Sense
Assessment: UnresolvedSaid
Ed Zitron Conditional forecast
“And if OpenAI can’t pay for that compute, Oracle dies, because it’s taken on around $115 billion in debt just to build Stargate’s data centers, and needs another $150 billion to finish them:”
Our reading: If OpenAI cannot pay for its Stargate compute, Oracle will die, having taken on about $115 billion in debt and needing $150 billion more.
No deadline stated Evidence as of
Only if: if OpenAI can't pay for the compute
Source: AI's Economics Don't Make Sense
Assessment: UnresolvedSaid
Ed Zitron Observation
“As of April 27 2026, only two buildings are operational and generating revenue, and the third barely has any IT gear in it. I estimate the total cost of Stargate Abilene to be around $52.8 billion.”
Our reading: As of April 27, 2026, only two Stargate Abilene buildings are operational and generating revenue and the third barely has any IT gear in it; the author estimates the campus's total cost at around $52.8 billion.
No horizon Evidence as of
Limitation: The only evidence item concerns April 22, not April 27.
Assessment: UnresolvedSaid
Ed Zitron Observation
“On the back end, AI startups are annihilating cash, with up until recently Anthropic allowing you to burn upwards of $8 in compute for every dollar of your subscription.”
Our reading: Until recently Anthropic let subscribers burn more than $8 in compute per dollar of subscription, and AI startups are burning cash.
No horizon Evidence as of
Limitation: The allowance estimates come from independent reverse engineering rather than an audited company disclosure.
Assessment: UnresolvedSaid
Ed Zitron Observation
“I really need to be clear that Oracle has no other path to making this revenue without OpenAI, and these projects are entirely financed and paid for using the projected cashflow of the data centers themselves.”
Our reading: Oracle has no other path to making this Stargate revenue without OpenAI, and these projects are entirely financed and paid for using the data centers' own projected cash flow.
No horizon Evidence as of
Limitation: Funding sources and ultimate repayment sources are different questions.
Assessment: UnresolvedSaid
Ed Zitron Causal claim
“This means that a large part — if not the majority of — AI compute revenue is dependent on a continual flow of venture capital and debt, both of which are only made possible by investors that still believe that generative AI will be the biggest, most hugest thing in the world.”
Our reading: A large part, possibly a majority, of AI compute revenue depends on continued venture capital and debt.
No horizon Evidence as of
Limitation: The packet supplies no representative industry-wide revenue denominator.
Assessment: UnresolvedSaid
Ed Zitron Forecast
“This is a deeply sad era. The people that have so aggressively worked together to hold up this industry have only delayed its inevitable fall.”
Our reading: The AI industry's fall is inevitable; its supporters have only delayed it.
No deadline stated Evidence as of
Limitation: The speaker supplies no precise definition of collapse.
Assessment: UnresolvedSaid
Ed Zitron Forecast
“There has never been — and never will be — an economically-feasible way to offer services powered by LLMs without charging the actual token burn of each user, and in the process of deceiving said users, these companies have created products with illusory benefits and questionable return on investment.”
Our reading: There has never been, and never will be, an economically feasible way to offer LLM-powered services without charging each user's actual token burn, and in deceiving users these companies have created products with illusory benefits and questionable return on investment.
No deadline stated Evidence as of
Limitation: No admitted evidence pairs sustained service-level financial results with contemporaneous non-token billing terms.
Assessment: UnresolvedSaid
Ed Zitron Forecast
“I think it’s inevitable that the majority of AI subscriptions move to token-based billing, especially as both Anthropic and OpenAI have now done so with their enterprise customers.”
Our reading: The majority of AI subscriptions will inevitably move to token-based billing.
No deadline stated Evidence as of
Limitation: The result could depend on whether offerings, plans, providers, or subscribers are counted.
Assessment: UnresolvedSaid
Ed Zitron Conditional forecast
“At some point investors — even the braindead fuckwits who have been inflating the AI bubble — will begin to question mounting R&D costs (which is where AI token burn is usually hidden) when the company’s revenue growth fails to follow.”
Our reading: Investors will eventually question mounting R&D costs from AI token spending when revenue growth fails to follow.
No deadline stated Evidence as of
Only if: revenue growth fails to follow
Assessment: UnresolvedSaid
Ed Zitron Observation
“It is an incredibly unprofitable business, and even if it wasn’t, it would have to make so much more money than it currently does to pay Oracle on an ongoing basis.”
Our reading: OpenAI is incredibly unprofitable and would need far more revenue than it has to pay Oracle on an ongoing basis.
No horizon Evidence as of
Limitation: Financial evidence comes through news reporting rather than directly inspected statements.
Assessment: UnresolvedSaid
Ed Zitron Observation
“It’s also unclear whether the customer base for AI compute exists outside of OpenAI and Anthropic, whose demand accounts for 50% of AI data centers under construction, creating a massive systemic weakness if either of them lacks the money to pay.”
Our reading: OpenAI and Anthropic account for 50% of AI data centers under construction, a systemic weakness if either cannot pay.
No horizon Evidence as of
Limitation: The evidence lacks a matched April global construction inventory and customer allocation [e1].
Assessment: UnresolvedSaid
Ed Zitron Forecast
“I should also be clear that Oracle is mortgaging its entire fucking future on projects like Stargate Abilene, incurring billions of dollars of costs up front for a business that will take years to turn a profit even if OpenAI makes every single payment in a timely manner.”
Our reading: Oracle's Stargate business will take years to become profitable even if OpenAI pays on time.
No deadline stated Evidence as of
Limitation: Aggregate customer funding does not establish OpenAI's payment due dates and receipt dates [e1].
Assessment: UnresolvedSaid
Ed Zitron Forecast
“I truly believe that LLM-based subscriptions are going to go away entirely, at least at the scale of any product that generates code, and in doing so, Amodei and Altman will wrap up their con, or at least believe that they have.”
Our reading: LLM-based subscriptions will disappear entirely, at least for code-generating products.
No deadline stated Evidence as of
Limitation: The reviewed commercial records are not a census of every code-generating product.
Assessment: UnresolvedSaid
Ed Zitron Causal claim
“And that day has finally arrived, because every single AI service you use subsidized compute, and every single service is losing money as a result:”
Our reading: Every single AI service you use subsidized compute, and every single one is losing money as a result.
No horizon Evidence as of
Limitation: The evidence does not provide comparable service-level revenue and actual serving-compute expenses for January–March 2026 [e1, e2].
Assessment: UnresolvedSaid
Ed Zitron Conditional forecast
“And if OpenAI can’t pay for that compute, Oracle dies, because it’s taken on around $115 billion in debt just to build Stargate’s data centers, and needs another $150 billion to finish them:”
Our reading: If OpenAI cannot pay for its Stargate compute, Oracle will collapse given its ~$115 billion of Stargate debt.
No deadline stated Evidence as of
Only if: OpenAI can't pay for the compute
Assessment: UnresolvedSaid
Ed Zitron Observation
“These are astonishing numbers, made more so by the fact that there is no way you’re only spending $30 a day if you use any of Anthropic’s more-recent models.”
Our reading: Using Anthropic's recent models with Claude Code costs more than $30 per day.
No horizon Evidence as of
Limitation: The aggregate figures do not disclose model shares, sampling dates or underlying daily usage records [e1, e2].
Assessment: UnresolvedSaid
Ed Zitron Observation
“If we assume that the 15.2GW of data center capacity under construction (due by the end of 2028) has a PUE of around 1.35, that leaves us with roughly 11.2GW of critical IT. At $14 million per megawatt, that works out to around $156.8 billion in annual revenue in GPU rental revenue required to actually make these data centers worth it.”
Our reading: Assuming a PUE of about 1.35 and $14 million per megawatt, the 15.2GW of data center capacity under construction (due by the end of 2028) would require around $156.8 billion a year in GPU rental revenue to be worth it.
No horizon Evidence as of
Limitation: An arithmetic-only reading would receive a different assessment.
Assessment: UnresolvedSaid
Ed Zitron Causal claim
“while prices for some models might have come down, newer “reasoning” models burn way more tokens, which means the cost of inference has, somehow, gotten higher over time.”
Our reading: Because reasoning models consume far more tokens, the cost of inference has risen over time even if some per-model prices fell.
No horizon Evidence as of
Limitation: The main empirical source is a preprint.
Assessment: UnresolvedSaid
Ed Zitron Observation
“Anthropic and OpenAI both intend to burn tens of billions of dollars in the next few years, and neither of them have a path to profitability.”
Our reading: Neither OpenAI nor Anthropic has a path to profitability.
No horizon Evidence as of
Limitation: The evidence does not provide sufficiently complete operating plans and financial assumptions to test either company's route to profitability [e1, e2].
Assessment: UnresolvedSaid
Ed Zitron Observation
“Even Cursor, a company that briefly (before its pseudo-acquisition by Musk’s SpaceX) claimed it had positive gross margins, actually had negative 23% gross margins as of January, or negative 31% if you include the cost of non-paying users, which you fucking should if you actually care about your accounting.”
Our reading: Cursor had -23% gross margins as of January 2026, or -31% including non-paying users.
No horizon Evidence as of
Limitation: The accessible evidence is secondary reporting.
Assessment: UnresolvedSaid
Ed Zitron Observation
“Both assumptions were wrong, because the monthly subscription model does not make sense for any service connected to a Large Language Model.”
Our reading: The monthly subscription model does not make economic sense for any LLM-based service.
No horizon Evidence as of
Limitation: Economic viability is interpreted as sustainable coverage of delivery costs.
Assessment: UnresolvedSaid
Ed Zitron Observation
“While Copilot might not be “...the same product it was a year ago,” very little has changed about the underlying economic mismatch: that Microsoft was allowing users to burn more than their subscription costs in tokens every single month for three years.”
Our reading: For three years Microsoft let GitHub Copilot users burn more in tokens each month than their subscriptions cost.
No horizon Evidence as of
Limitation: The original WSJ article could not be read; its figures are represented through contemporaneous reporting.
Source: AI's Economics Don't Make Sense
Assessment: UnresolvedSaid
Ed Zitron Causal claim
“This is a blinking red fucking light, and in a sane market would send Oracle’s stock into a tailspin, because OpenAI’s ascent to over $280 billion in annual revenue is critical to Oracle’s ability to not run out of money.”
Our reading: Oracle avoiding running out of money depends on OpenAI growing to over $280 billion in annual revenue, so the CFO's worries are a serious warning sign.
No horizon Evidence as of
Limitation: OpenAI's detailed contracts and investor financial model were not obtained.
Source: AI's Economics Don't Make Sense
Assessment: MixedSaid
Ed Zitron Causal claim
“while prices for some models might have come down, newer “reasoning” models burn way more tokens, which means the cost of inference has, somehow, gotten higher over time.”
Our reading: Although prices for some models might have come down, newer reasoning models burn far more tokens, so the cost of inference has gotten higher over time.
No horizon Evidence as of
Limitation: The studies do not establish an industry-wide trend across production workloads or providers' internal inference costs [e8, e9].
Source: AI's Economics Don't Make Sense
Assessment: UnresolvedSaid
Ed Zitron Observation
“Anthropic and OpenAI both intend to burn tens of billions of dollars in the next few years, and neither of them have a path to profitability.”
Our reading: Anthropic and OpenAI both intend to burn tens of billions of dollars in the next few years, and neither has a path to profitability.
No horizon Evidence as of
Limitation: The cash-burn reporting relies on secondary accounts of private projections.
Source: AI's Economics Don't Make Sense
Assessment: UnresolvedSaid
Ed Zitron Observation
“On the back end, AI startups are annihilating cash, with up until recently Anthropic allowing you to burn upwards of $8 in compute for every dollar of your subscription. OpenAI allows you to do the same, though it’s hard to gauge by how much.”
Our reading: AI startups are annihilating cash: until recently Anthropic let subscribers burn upwards of $8 in compute for every dollar of their subscription, and OpenAI lets them do the same, though it is hard to gauge by how much.
No horizon Evidence as of
Limitation: The independent allowance analysis is not a provider-cost audit.
Source: AI's Economics Don't Make Sense
Assessment: UnresolvedSaid
Ed Zitron Observation
“Both assumptions were wrong, because the monthly subscription model does not make sense for any service connected to a Large Language Model.”
Our reading: The monthly subscription model does not make economic sense for any LLM-connected service.
No horizon Evidence as of
Limitation: The quarter and meaning of economic viability require interpretation.
Source: AI's Economics Don't Make Sense
Assessment: UnresolvedSaid
Ed Zitron Observation
“Eager math-knowers in the audience will also realize that, if we assume a $5-a-month subscription cost, even if OpenAI succeeds in what would be the single-largest user acquisition campaign in history, it would still be $155 million short.”
Our reading: Assuming a $5-a-month subscription price, even if OpenAI succeeds in growing ChatGPT Go to its projected subscriber count (which he calls what would be the single-largest user acquisition campaign in history), the new Go revenue would still be $155 million short of the ChatGPT Plus revenue lost.
No horizon Evidence as of
Limitation: The monthly gap is not a realized annual revenue loss or a measure of profitability.
Assessment: UnresolvedSaid
Ed Zitron Forecast
“I also just did a piece about how OpenAI will kill Oracle.”
Our reading: OpenAI will kill Oracle.
No deadline stated Evidence as of
Limitation: The corporate-failure threshold and principal-cause requirement are evaluator interpretations of kill.
Assessment: SupportedSaid
Gary Marcus Observation
“But that’s exactly why we still need software engineers — and why what Amodei recently said about software engineers disappearing is so absurd.”
Our reading: Software engineers are still needed, and Amodei's claim that they will disappear is absurd.
No horizon Evidence as of
Limitation: The decisive occupational count is secondary evidence.
Source: Dario Amodei, hype, AI safety, and the explosion of vibe-coded AI disasters
Assessment: SupportedSaid
Gary Marcus Observation
“AI coding tools are genuinely revolutionizing the software industry, but more than almost anything I have ever seen, they fall into the category of things you maybe shouldn’t try at home without significant prior experience:”
Our reading: AI coding tools are genuinely revolutionizing the software industry.
No horizon Evidence as of
Limitation: The reporting period precedes the post by several months.
Source: Dario Amodei, hype, AI safety, and the explosion of vibe-coded AI disasters
Assessment: UnresolvedSaid
Gary Marcus Forecast
“And I am sure we will see many more catastrophic failures soon enough, involving things like privacy, security and data loss. Already there have been many.”
Our reading: Many more catastrophic vibe-coding failures involving privacy, security and data loss will occur soon.
No deadline stated Evidence as of
Limitation: Public incident reporting is incomplete.
Source: Dario Amodei, hype, AI safety, and the explosion of vibe-coded AI disasters
Assessment: UnresolvedSaid
Gary Marcus Observation
“Here is a further example of my last point that ChatGPT’s exciting new image system has trouble with functional understanding, sent to me by a reader:”
Our reading: ChatGPT's exciting new image system has trouble with functional understanding.
No horizon Evidence as of
Limitation: The assessment concerns observable functional depictions, not whether the model possesses human-like understanding.
Assessment: UnresolvedSaid
Gary Marcus Observation
“But if you look closely, there are several errors, and those errors are revealing.”
Our reading: ChatGPT's new image engine's labeled bike diagram contains several revealing labeling errors.
No horizon Evidence as of
Limitation: This is an evidence-access limitation, not a finding that the labels are correct.
Assessment: UnresolvedSaid
Gary Marcus Observation
“But anybody knowledgeable about bikes (racers, mechanics, designers etc) would immediately see numerous problems.”
Our reading: ChatGPT's tandem bike image contains numerous problems that bike experts would immediately spot.
No horizon Evidence as of
Limitation: The packet contains no inspectable copy of the tandem image or qualified independent assessment.
Assessment: UnresolvedSaid
Ed Zitron Observation
“While the documents refer to moving “all” GitHub Copilot users to token-based billing, it’s unclear at this time how Microsoft will be handling individual Pro or Pro+ subscribers.”
Our reading: Although documents say all GitHub Copilot users move to token-based billing, it is unclear how individual Pro and Pro+ subscribers will be handled.
No horizon Evidence as of
Limitation: This assessment does not dispute the author's account of his own knowledge.
Source: [Updated] Exclusive: Microsoft Moving All GitHub Copilot Subscribers To Token-Based Billing In June
Assessment: UnresolvedSaid
Ed Zitron Observation
“Documents viewed by Where’s Your Ed At shed additional light on Microsoft’s transition to token-based billing for GitHub Copilot, as the company grapples with spiraling costs of AI compute.”
Our reading: Microsoft is transitioning GitHub Copilot to token-based billing at a time when it is grappling with spiraling AI compute costs.
No horizon Evidence as of
Limitation: An unspecified accounting basis prevents a definitive financial score.
Source: [Updated] Exclusive: Microsoft Moving All GitHub Copilot Subscribers To Token-Based Billing In June
Assessment: SupportedSaid
Gary Marcus Observation
“That’s still true – and it very much applies in medicine.”
Our reading: Large language models are still frequently wrong yet confident, producing authoritative bullshit that shouldn't be trusted, and this applies strongly in medicine.
No horizon Evidence as of
Limitation: Support depends on the contextual evidence-period reading.
Assessment: SupportedSaid
Gary Marcus Observation
“All of this – the hallucinations, mistakes, and overconfidence – is entirely typical of LLMs, and entirely problematic in medicine.”
Our reading: Hallucinations, mistakes and overconfidence are entirely typical of LLMs and entirely problematic in medicine.
No horizon Evidence as of
Limitation: The decisive audit uses misinformation-oriented prompts and older model versions.
Assessment: SupportedSaid
Gary Marcus Observation
“Because a large fraction of the population has begun to turn to chatbots for medical advice.”
Our reading: A large fraction of the population has begun using chatbots for medical advice.
No horizon Evidence as of
Limitation: The U.S.-adult population is an interpretation, not an explicit commitment by Marcus.
Assessment: SupportedSaid
Gary Marcus Causal claim
“we should not take the apparent “confidence” of large language models — itself an illusion of how they are trained — to mean”
Our reading: The apparent confidence of LLMs is an illusion that comes from how they are trained.
No horizon Evidence as of
Limitation: The result depends on reading 'illusion' as misleading certainty rather than complete absence of information.