Hyperscale Normalization
We recorded 12 claims from this post. 12 can’t be checked. 2 more are held. Read the post at its source.
Claims from this post (12)
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
Observation
“Sidenote: NVIDIA deploying capital at random to help counterparties raise debt, despite being legal, should immediately trigger memories of Enron, if only because their actions have the same intention of artificially inflating revenues. Real businesses do not need financial wizardry.”
Our reading: NVIDIA's investments helping counterparties raise debt are intended to artificially inflate its revenues, comparable to Enron.
No horizon
Assessment: Can’t be checkedSaid
Observation
“This is not the same government that could once afford to pump trillions of dollars into any kind of bailout without running the very real risk of destabilizing the US dollar.”
Our reading: The US government, with national debt now at $40 trillion, can no longer pump trillions of dollars into a bailout without running a very real risk of destabilizing the US dollar.
No horizon
Qualifier: with US national debt grown to $40 trillion
Assessment: Can’t be checkedSaid
Rhetoric
“The entire AI bubble — every bit of hype — is an attempt to rebrand old things as new.”
Our reading: All AI bubble hype is an attempt to rebrand old things as new.
No horizon
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“There is nothing rational about valuing Cognition at $26 billion, let alone the $40 billion one it’s allegedly raising at right now.”
Our reading: There is nothing rational about valuing Cognition at $26 billion, let alone at the $40 billion it is allegedly raising at now.
No horizon
Qualifier: $40 billion round is 'allegedly' being raised
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“I have no specific beef with Cognition, because the same can be said of Perplexity, Higgsfield, Harvey, or any number of other AI companies with triple-digit “annualized run rates” with double-digit billion valuations for businesses with questionable business models.”
Our reading: The same criticism made of Cognition applies to Perplexity, Higgsfield, Harvey and any number of other AI companies with triple-digit 'annualized run rates' and double-digit-billion valuations for businesses with questionable business models.
No horizon
Scope: AI companies with triple-digit annualized run rates and double-digit-billion valuations
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“There is nothing ‘anti-progress’ about opposing AI data centers, and nobody has a compelling explanation as to why we need more of them.”
Our reading: Nobody has a compelling explanation for why more AI data centers are needed, and there is nothing 'anti-progress' about opposing them.
No horizon
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“CoreWeave is a bad company that only exists because of the simple, systemic belief that If The Right Numbers Are Going Up, Everything Is Fine.”
Our reading: CoreWeave is a bad company that only exists because of the simple, systemic belief that if the right numbers are going up, everything is fine.
No horizon
Assessment: Can’t be checkedSaid
Observation
“Everything, even often in critical pieces, comes back to statements that reinforce the status quo, like “AI is, of course, transformative,” or “bubbles always form and leave value afterward,” even though that’s not really true at all when you look at history, and certainly isn’t true of this particular era.”
Our reading: The common claim that bubbles always form and leave value afterward is not really true historically, and certainly is not true of the current AI era.
No horizon
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“I’m sure there are automations that people have made with this stuff that help them — good for you! — but it will take a lot more than that to justify a trillion-plus dollars in investment or endless fucking prattling about how impressive and world-changing AI is.”
Our reading: Automations people have built with LLMs that help them are not enough; it will take a lot more to justify a trillion-plus dollars of investment or the endless talk about how impressive and world-changing AI is.
No horizon
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“The thing is, LLMs have improved in the last year, just not in a way that’s tremendously impressive to me given the amount of capital invested, or in a way that has manifested in a tangible product that can be described in a sentence.”
Our reading: LLMs have improved in the last year, but not in a way that is tremendously impressive to the author given the capital invested, nor in a way that has produced a tangible product describable in a sentence.
No horizon
Qualifier: given the amount of capital invested
Assessment: Can’t be checkedSaid
Causal claim
“Sidenote: While I don’t dispute that the AI industry has done something, the scale of its importance is entirely a result of the media’s puffery, subsidized subscriptions, venture capital and hyperscaler funding and endless debt.”
Our reading: The AI industry's apparent importance is entirely due to media hype, subsidized subscriptions, VC and hyperscaler funding and debt.
No horizon
Assessment: Can’t be checkedSaid
Forecast
“In fact, every time you read about some multi-billion dollar data center deal, know that it won’t be built for years, but NVIDIA will likely book the revenue immediately, because all it has to do is help raise the debt for the banks to send the money and the GPUs to be sold.”
Our reading: Multi-billion-dollar data center deals will not be built for years, but NVIDIA will likely book the associated revenue immediately.
No horizon
Qualifier: multi-billion dollar data center deals
Held from this post (2)
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: Once the venture-capital- and debt-subsidized spending of the big AI players is removed, there is very little real money in the system, because nobody other than Jensen Huang and the data center developers is making a profit from AI.
- The two reviews read the claim differently