“For the last few years, I have been warning you that the mania for LLMs would break eventually.”
Our reading:
The mania for LLMs will eventually break.
Assessment: Unresolved
What happened
Did the forecast come true?
Outcome: Unresolved
We checked, and the evidence so far doesn’t settle it.
Judged on a plain reading. They named no measure, so we used the reading a typical informed reader would take. It is spelled out under “How we judged it”.
Evidence up to . Checked .
Why
On 2025-08-20, Marcus wrote that the mania for LLMs "would break eventually." He gave no deadline. A Decrypt report available on 2025-08-20 quotes Sam Altman saying he thought investors were overexcited about AI (e3). That supports the idea that there was a mania. It does not show whether the mania later broke. Nvidia reported record revenue of $96.2 billion for the quarter ended July 26, 2026. That was up 106% from a year earlier. Data center revenue was $89.0 billion, up 117% (e1). On 2026-08-27, Nvidia's CEO said the AI infrastructure buildout was "at full steam" (e2). None of the evidence shows the sustained steep fall in AI spending that would settle this forecast early. The forecast has no deadline, so the lack of a collapse in the evidence does not count against it. It stays open.
Caveats
- The latest revenue figures in the evidence are for the quarter ended July 26, 2026. The evidence cannot show whether a collapse began between then and the 2026-10-28 evaluation date.
- The only evidence dated after that quarter is the Nvidia CEO's statement of 2026-08-27, which is the company's own view.
- The evidence has no data on hyperscaler AI capital spending, AI startup funding or valuations, or AI-linked stock indices.
- All of the market evidence comes from one company, Nvidia.
- 'Mania' and 'break' have no stated threshold.
Evidence (3 items)
Context only: StorageNewsletter (reproducing Nvidia press release),
Nvidia Corp. reported record revenue for the second quarter ended July 26, 2026, of $96.2 billion, up 18% from the previous quarter and up 106% from a year ago.
Why it matters: The main seller of AI chips more than doubled its revenue year over year into mid-2026. That shows the LLM investment boom had not broken by then. It does not refute an open-ended forecast.
- Publisher
- StorageNewsletter (reproducing Nvidia press release)
- Kind of source
- Company disclosure: the company describing itself, not an independent finding, unaudited
- Measures
- Nvidia, total revenue, USD billions, company reported, 2026-04-27 to 2026-07-26
- About events on
- Published
Context only: StorageNewsletter (reproducing Nvidia press release),
The AI infrastructure buildout is at full steam.
Why it matters: Nvidia's CEO described the AI buildout as still accelerating in August 2026. This is a company view, but it is consistent with the reported revenue.
- Publisher
- StorageNewsletter (reproducing Nvidia press release)
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- About events on
- Published
Context only: Decrypt,
“Are we in a phase where investors as a whole are overexcited about AI? My opinion is yes,” he said.
Why it matters: Even Altman said investors were overexcited around the time of the post. That backs the premise that there was a mania, but says nothing about whether it later broke.
- Where
- Final paragraphs
- Publisher
- Decrypt
- Kind of source
- News report
- About events on
- Published
How we judged it
- Kind of claim
- Forecast
- Deadline
- “eventually”
Where we had to interpret
- Plain reading: at some point, enthusiasm for LLMs would visibly collapse. That means a sharp, sustained fall in generative AI investment, AI chip and data center spending, and AI-linked valuations, not a short dip.
- He gave no deadline ('eventually'), so this is an open-horizon forecast. The lack of a collapse before the evaluation date does not count against him.
Right if: A dated, decisive collapse of the LLM boom: for example, AI data center or chip spending falling steeply year over year for several quarters, major AI valuations or funding rounds collapsing, and widespread reporting that the boom has ended.
Wrong if: With no deadline, the forecast cannot be shown false at a fixed date. Continued strong growth only shows the break has not happened yet.
We wrote this test after the outcome was known.
What would settle it
Dated company filings or reputable financial reporting showing a sustained collapse. For example, Nvidia data center revenue or combined hyperscaler AI capital spending falling more than 30% year over year for two or more quarters in a row, together with broad reporting that generative AI investment has collapsed. That would support the forecast. Because there is no deadline, no fixed date can contradict it.
We found 5 claims in this post. See them all.
Full record
Source and dates
Marcus on AI, Is this the moment when the Generative AI bubble finally deflates?, paragraph 1.
- Said
- We read the source
- live page, full text, public
- Evidence up to
- Checked
We read this wording on 2026-10-28. The post may have been edited after 2025-08-20.
Where it was said (1)
Said once in the posts we read.
: Marcus on AI, Is this the moment when the Generative AI bubble finally deflates?
For the last few years, I have been warning you that the mania for LLMs would break eventually.
- Where
- paragraph 1
- Published
- 2025-08-20
- We read it
- , live page
- Access
- full text, public
History
Versions
- Current: Unresolved. Evidence up to 2026-10-28.
What happened to this record
PublishedBy us
Published after both AI reviews and the final check passed.
Held backBy us
The final check failed: the write-up said more than the evidence showed.