A growing backlash is forcing a reckoning with whether the tech industry's most celebrated investment cycles actually solved anything for ordinary people, or simply enriched those already closest to the capital.
The discourse has a familiar rhythm by now. A technology emerges, venture capital floods in, thought leaders declare a paradigm has shifted, and consumer adoption either plateaus or collapses entirely. What is different in 2026 is that the backlash has become articulate. Across X and Reddit, a coordinated skepticism is targeting the three defining investment themes of the post-pandemic era, NFTs, the metaverse, and generative AI, with a single damning thesis: Silicon Valley has been solving problems that only Silicon Valley has.
Meta is the most expensive cautionary tale in the argument. When Mark Zuckerberg rebranded the company in late 2021, the metaverse pitch was civilizational in scope. By Q4 2023, Reality Labs had posted losses exceeding $13.7 billion in a single quarter, with user adoption figures that never came close to justifying the rhetoric. The company has since pivoted sharply into AI infrastructure, but the metaverse chapter reads as a textbook case of an executive class convincing itself, and its shareholders, that aspiration was the same thing as demand.
NFTs followed a structurally identical arc. Market volume peaked above $25 billion in 2021, driven by a speculative frenzy that conflated digital ownership with cultural relevance. By late 2023, Coinbase was publicly distancing itself from the technology. The collapse was not simply a correction; it was a signal that the underlying value proposition had never been tested against the needs of anyone outside the collector economy.
Generative AI deserves its own analysis because the stakes are genuinely different. NVIDIA briefly crossed a $2 trillion valuation in 2024, and OpenAI has been valued at roughly $80 billion, figures that reflect real infrastructure demand, not pure speculation. The problem is not the technology itself but the gap between enterprise deployment and consumer reality. Data from late 2025 indicated that nearly half of early users abandoned novel AI tools within weeks, citing hallucinations and poor workflow integration as the primary reasons.
Consumer hardware has been the most visible failure surface. The Rabbit r1 and Humane AI Pin arrived with substantial press, were reviewed as buggy and incomplete, and effectively became symbols of an industry shipping ambition ahead of function. The pattern is instructive because it suggests the problem is not capability but prioritization. Engineers capable of building large language models apparently found it difficult to build a product that reliably does what it says on the box.
What makes the current backlash worth taking seriously is where it appears to be redirecting capital. Venture investors are increasingly signaling a preference for hard tech, defense, manufacturing, and energy infrastructure, sectors where the feedback loop between product and utility is short and measurable. Consumer applications built on hype cycles are facing longer scrutiny periods before funding closes. That is a structural shift, not a sentiment blip.
The historical parallel being invoked most often is the dot-com collapse of 1999 to 2000. The comparison holds in one important respect: infrastructure survived while speculative ventures evaporated. Cisco and Oracle outlasted Pets.com. The question for this cycle is which AI companies are laying pipe and which ones are selling the dream of pipe. The answer will take another two or three years to become fully legible.
There is also a class dimension that the online discourse is surfacing with unusual clarity. Wall Street's AI-fueled stock euphoria is running in parallel with genuine economic anxiety on Main Street, where consumers are managing inflation, housing costs, and stagnant wages. An industry that responds to that environment with spatial computing headsets and AI-generated art has a positioning problem that no amount of thought leadership can paper over. The next phase of technology growth that actually reaches ordinary people will almost certainly be defined by affordability and friction reduction, not by the ambitions of a conference keynote.
Also read: A Chinese robot just ran a half-marathon faster than any human ever has • Stanford report reveals China has overtaken the United States in AI model production while the flow of global researchers into America slows to a halt • Apple and Google are still promoting nudify apps to children three months after the problem was first exposed
