“AI data centers are extremely expensive.”
Our reading, the sentence that was checked:
AI data centers are extremely expensive.
Assessment: Supported Evidence as of
Where and when Ed Zitron wrote it
Where's Your Ed At, Dead Money, paragraph 141.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment wyea-dead-money-6054f8-a1
We recorded 50 claims from this post. See every claim from this post.
The wording above is the version this site observed on 2026-10-28. The source may have been edited after 2026-09-29. This page does not claim the wording was captured then.
What the claim commits to
- Kind
- Observation
- Deadline, as the speaker wrote it
- None stated. This site does not add one.
- Deadline state when assessed
- No horizon
- Deadline state on 2026-10-29
- No horizon (the release’s reference date)
From the speaker
- AI data centers
- extremely expensive
Our interpretation
- The ordinary reading is that large AI data centers require very large absolute capital investments: millions of dollars per MW and potentially billions at the 100-MW scale discussed in the post.
- This does not assert that every facility costs billions, that every project costs exactly $4–5 billion per 100 MW, or that high costs imply unprofitability.
- Use 2026 year-to-date cost information through September 29, the latest reportable cost-estimate period before publication.
- This is a qualitative industry-cost claim, not a financial result for a named company.
Counts as supported if: Contemporaneous construction and AI hardware benchmarks establish multimillion-dollar costs per MW and billion-dollar-scale capital requirements for large facilities.
Counts as contradicted if: Representative contemporary costs are materially below that scale, showing that the large AI facilities discussed do not require unusually large absolute capital investments.
Retrospective: these criteria were written after the outcome was public.
Current assessment
Was it true for the period described?
Outcome: Supported
The dated evidence matches what the claim committed to, under the criteria frozen before grading.
Scored on a plain reading. The speaker named no measure for this claim, so it was judged on the reading an informed reader would take, stated below under our interpretation. Both reviews had to find that reading fair.
- Evidence as of
- Deadline state when assessed
- No horizon
- Assessment recorded
- Policy version
- 2026-10-28.7
Rationale
Published development and AI fit-out benchmarks support the ordinary claim that large AI data centers require very large capital investments. Cushman & Wakefield's modern-facility benchmark scales to $1.76 billion at 100 MW before the GPU costs identified separately. JLL also reports substantial AI fit-out costs. These records establish the cost scale, not the post's exact project estimates or its conclusions about returns.
Limitations
- The adjective is qualitative; support applies to the stated ordinary reading.
- Benchmarks vary by market, facility design, and hardware.
- The 100-MW calculation is illustrative.
- JLL's fit-out number is an upper-end estimate, not an average.
Evidence (3 items)
Supports the claim: Cushman & Wakefield,
costs vary widely by market.
Why it matters here: The following sentence reports average greenfield costs of $17.6 million per MW for the most modern facilities. Scaling that benchmark to the post's 100-MW example yields $1.76 billion, supporting the ordinary claim of very large absolute capital requirements.
- Publisher
- Cushman & Wakefield
- Kind of source
- Independent measurement
- Arithmetic
- 17.6 million USD/MW × 100 MW = 1,760 million USD (inputs: Cushman & Wakefield modern-facility benchmark: USD 17.6 million per MW; Illustrative facility size from the post: 100 MW; USD 1.76 billion; illustrative scaling, not a project quotation.)
- Describes events of
- Public from
Supports the claim: JLL,
$25 million per MW
Why it matters here: JLL identifies AI technology fit-out as a separate, potentially substantial capital requirement. Its upper-end estimate supports the cost scale, without establishing a typical cost or a universal project budget.
- Publisher
- JLL
- Kind of source
- Independent measurement
- Describes events of
- Public from
Context only: RealtyWire,
excludes the chips and GPUs
Why it matters here: The report identifies September 3 as the guide's release date and clarifies the construction benchmark's scope. It helps prevent treating the development figure as an AI hardware-inclusive total.
- Publisher
- RealtyWire
- Kind of source
- News report
- Describes events of
- Public from
Where it was said (1)
Said once in the sources this register covers.
: Where's Your Ed At, Dead Money
AI data centers are extremely expensive.
- Where
- paragraph 141
- Published
- 2026-09-29
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment wyea-dead-money-6054f8-a1: Current. Outcome: Supported. Evidence as of 2026-10-28. Public date not recorded.
Recorded events
Assessment publishedBy this site
First assessment published after both reviews and the final-text check passed.