Silicon Valley has forgotten what normal people want (Lopatto, The Verge, April 20 2026)

"Silicon Valley has forgotten what normal people want" is a column-length essay (roughly 2,300 words) by Elizabeth Lopatto, a senior reporter at The Verge, published in the outlet's TLDR column on April 20, 2026 (Source: theverge.com). It argues that NFTs, the metaverse, VR/AR headsets, and LLMs share a structural failure to solve a consumer market problem, attributes that failure to a cultural pathology Lopatto names "Silicon Valley incuriosity," and contends that the only customer who can justify the cost of building LLMs is the US government.

Summary of argument

The essay builds a single thesis across three nested arguments. First, Silicon Valley's most-pushed consumer technologies of the last decade — NFTs, the metaverse, VR/AR headsets, and now LLMs — share a structural failure to solve a market problem and were instead optimized to enrich VCs and operators. Second, the underlying cause is a cultural pathology Lopatto names "Silicon Valley incuriosity," a self-conception that "if I don't know it, no one knows it" that, she argues, makes founders and VCs unable to recognize when their "discoveries" are well-trodden 101-level material elsewhere. Third, in AI specifically, "there is only really one customer for LLMs that can justify the massive cash incineration process that was required to build them: the US government." Lopatto holds that consumer adoption is real only "as long as they remain free," and that LLMs' durable use cases — cheating on schoolwork, looking up information as a Google Search substitute, enterprise data organization, and faster coding — are narrow and largely non-monetizable at the price the capital expenditure requires.

"Silicon Valley incuriosity"

Lopatto opens with anecdotes of technologists discovering, in conversation, things long known elsewhere. The opening example is an acquaintance announcing that LLMs reveal knowledge is structured into language, which Lopatto characterizes as a "naive, confused version of Structuralism; Saussure via a game of telephone." She extends the pattern to three further examples: Elon Musk marveling at the complexity of hands, which artists, surgeons, musicians, and neuroscientists already know as 101-level material (the cortical homunculus is encountered early in undergraduate neuroscience); Palmer Luckey claiming "no one has done a postmortem" on One Laptop Per Child when there is a whole book about it, The Charisma Machine; and Marc Andreessen announcing that Freud invented introspection, while adding as a bonus that Andreessen himself does not engage in it. The diagnosis Lopatto draws is that siloed founders cannot recognize that their "discoveries" are known elsewhere because they do not take for granted that other smart people have always existed.

From "identify a need" to "invent the future"

Lopatto traces a post-financial-crisis shift in Silicon Valley's self-conception, from "identify a need, fill it" to "invent the future and consumers' job is to go along." She offers the Steve Jobs post-return playbook as the earlier mode: the iMac for ease of use, the iPod to carry one's music, and the iPhone plus App Store for breadth of utility. NFTs, the metaverse, and the Oculus and Vision Pro headsets are presented as illustrations of the later failure. As she puts it, "At some point, our Silicon Valley overlords forgot that in order for their vision of the future to be adopted, people had to want it."

"The only customer for LLMs is the US government"

On AI specifically, Lopatto concedes that "AI is, admittedly, more useful — it's good for organizing large swaths of data," and notes that LLMs have widespread consumer adoption "as long as they remain free." She then asserts that "there is only really one customer for LLMs that can justify the massive cash incineration process that was required to build them: the US government." Because "there can only be a few winners on government contracts," she writes, "we are now treated to the spectacle of watching AI companies scramble. OpenAI is perhaps the funniest, because it is attempting to position itself as a consumer product."

She names two consumer-AI failure modes. On labor disruption, she argues that "the people who tell us that AI will dominate our future and take our jobs are the people who are hoping that will be true," casting the labor-disruption thesis as structurally motivated by founder self-interest — importance, billionaire status, and an inability to imagine other lives. On efficiency, she argues that "there are places in our lives where efficiency isn't desirable," using vacation planning as an illustration, citing the stock-market trading halt as a load-bearing inefficiency, and contending that AI music apps rest on the false premise that non-musicians want to make music.

Two worked examples she rejects anchor the AI section. On Sam Altman saying he needed ChatGPT to tell him how to raise a baby, Lopatto writes, "I presume what Altman actually did was hire a nanny," and adds that "a mandatory measles vaccine will do more for the survival of American children than anything OpenAI has accomplished." On Musk's vision of a humanoid robot servant, she writes, "I have a robot servant. Several, actually: a dishwasher, a washer for my clothing, and a dryer ... my microwave isn't much younger, and both of those things have been remarkable," setting the bar AI mobile robotics must clear at the value-for-money of 1990s white goods.

Key claims

The page assesses the essay's principal claims against other wiki coverage as follows.

The "as long as they remain free" claim is a forward-looking economic thesis that would be tested directly if OpenAI's $0 Apple-bundle partnership unwinds (an OpenAI v. Apple legal-action preparation is noted in the 2026-05-16 daily brief) and consumer ChatGPT moves to a meaningful price floor. The "only customer is the US government" claim is similarly falsifiable by tracking ARR, the enterprise-versus-consumer revenue mix, and the concentration of federal-contract revenue.

Provenance and context

The essay is an op-ed position rather than factual evidence; its claims are paraphrased here as arguments. It sits alongside Prepare for an AI jobs apocalypse (The Economist Leader, May 14 2026) as a same-period institutional-position piece, though it approaches AI skepticism from a product-failure rather than a labor-displacement angle. The Andreessen "no-introspection" thread extends an earlier Verge line of reporting on Andreessen as a "philosophical zombie." Lopatto connects her argument to the "pluto-populist" and AI-dread literature covered in Why Americans dread AI — Edward Luce (FT, May 12 2026) and AI Divides (Literacy / Occupational / Ethico-Philosophical), and to the AI-bubble debate covered in Even Silicon Valley Says That AI Is a Bubble — Lila Shroff (The Atlantic, March 2026), where Shroff defends the bubble and Lopatto attacks it from the consumer-product-failure angle.

Relationships

• supports: AI Bubble Debate (consumer-AI-has-no-business-model thesis); Enterprise AI Deployment Gap (LLM real value sits in enterprise, not consumer); AI Labor Disruption (labor-displacement-as-founder-wishful-thinking framing — a reframe, not a contradiction)
• contradicts: (Source: openai.com) + (Source: openai.com) (which position ChatGPT bundles as durable consumer-product moves — Lopatto's lens would call these scrambles for survival); the implicit consumer-AI revenue model in frontier-lab valuation rounds
• depends-on: AI Divides (Literacy / Occupational / Ethico-Philosophical) (pluto-populist + AI-dread literature she connects to); AI Bubble Debate
• related: Why Americans dread AI — Edward Luce (FT, May 12 2026) ("pluto-populist" framing — Lopatto-adjacent); Even Silicon Valley Says That AI Is a Bubble — Lila Shroff (The Atlantic, March 2026) (Shroff defends the bubble; Lopatto attacks it from the consumer-product-failure angle); the Andreessen "philosophical zombie" / no-introspection thread the Verge previously broke