1 min read
In short
A WSJ headline promises to show why Big Tech's real AI spending may be $3tn above the usual estimates. But the available material carries no calculations, no methodology and no detail — and that matters more than the headline figure.
A figure of $3tn sounds like reason to reconsider the scale of the AI race. But for now it is not a conclusion, only a headline's promise: the available excerpt contains no explanation of what counts as spending or where such a difference comes from.
There is a practical signal here for the reader. Companies' official spending may not reflect the full cost of AI infrastructure — but without a methodology it is impossible to tell whether this concerns direct investment, associated costs or a different accounting system. Substituting a confident conclusion for the unknown would simply repeat the marketing pitch.
The limitation of the material is critical: the WSJ's calculations, source data, assessment period and list of companies are not available to us. So it cannot honestly be claimed that Big Tech really spent $3tn more on AI than stated. All we can do is note the article's claim and wait for the arguments.
When you see enormous estimates of AI spending, which matters more to you: the final sum, or a transparent methodology behind it? Source: Hacker News - Newest: ""AI" "LLM""