Hyperscalers are spending $725 billion on AI infrastructure in 2024, but the market is rejecting the output at scale. Fifty percent of US consumers actively prefer brands that don't use generative AI, according to Gartner research. Wikipedia banned AI-generated content with a 44-2 vote. Stack Overflow's new-question volume dropped 78 percent year over year. Google's AI Overviews, which surface AI-generated summaries at the top of search results, tanked click-through rates on the top search result by 58 percent.
The mismatch is stark. Infrastructure spending is accelerating in the exact segments where users are abandoning the technology. This creates a structural problem: massive capital allocation toward products consumers actively avoid.
The Wikipedia ban signals community-level resistance. Volunteer editors voted overwhelmingly to reject machine-generated contributions, treating them as lower quality than human curation. Stack Overflow's collapse in new questions suggests users are either finding answers elsewhere or losing motivation to contribute. The platform built on crowd-sourced programming knowledge is hemorrhaging participation.
Google's Overviews data is the most damaging metric. A 58 percent drop in top-result clicks means users ignore AI summaries and click deeper. They don't trust the output enough to stop searching. Google's own AI integration is cannibalizing its core search business.
The consumer preference data compounds the problem. When half of surveyed consumers explicitly prefer brands avoiding generative AI, the technology becomes a liability, not an asset. Marketing departments pitch AI as a cost-saving feature. Users hear "lower quality and effort."
This dynamic creates a capital trap. Companies cannot easily walk away from $725 billion in spending commitments. Infrastructure investments require long payoff horizons. But if demand keeps falling, utilization plummets and returns evaporate. The