Nasdaq erases all 2026 gains as AI sell-off dampens Anthropic’s second day

Global markets are sliding even as Anthropic’s blockbuster IPO falters, intensifying doubts about whether big AI labs can ever grow into their trillion‑dollar valuations. Commenters question OpenAI and Anthropic’s business models, pointing to messy finances, looming competition from open and Chinese models, and the risk that AI commoditizes knowledge work faster than companies can capture value. Much of the debate centers on who will benefit from AI-driven productivity — investors, a handful of tech giants, or broader society — and whether current trajectories deepen inequality and economic fragility instead of creating widely shared gains.

IPO timing, OpenAI vs Anthropic

  • Commenters speculate OpenAI missed its best IPO window around GPT‑4 due to internal financial complexity, governance issues, and high valuation demands that banks wouldn’t meet.
  • Anthropic is seen as needing cash more urgently, pushing ahead despite a weak market; some call OpenAI’s delay a “generational whiff” if the window is now closing.
  • Several note that any frontier-lab IPO is itself a top‑signal: if you really believe you’re worth “trillions,” why sell to public markets now?

Anthropic IPO and stock drop

  • Anthropic’s stock fell sharply after listing (around −20%+ from the start of trading).
  • Some attribute this mainly to a broader market sell‑off and high‑beta behavior; others argue the scale of the drop and widespread paper losses suggest mispricing and “dumb money” exit liquidity.
  • Debate over whether IPO dips of this size are “normal”; some want historical context, others see this as typical IPO overpricing.

AI company valuations, moats, and business model

  • Skepticism that Anthropic or other labs can ever earn enough to justify current valuations; models are increasingly commoditized and open‑source competitors are strong.
  • Anthropic’s own TAM pitch (tens of trillions, ~35% of global GDP) is viewed as effectively “capturing all the money.”
  • Many argue moats based on proprietary models and compute are fragile: DeepSeek-style efficiency gains, Chinese open models, and potential new architectures could crush pricing power.
  • Others say long‑term value might come less from model quality and more from B2B integration, tooling, and lock‑in, similar to how M365 sells.

Macroeconomics, labor, and distribution

  • Large subthread on whether AI-driven automation destroys GDP or just reallocates spending.
  • One view: AI shrinks specific sectors but frees capital and consumers to spend elsewhere, raising long‑run productivity.
  • Counterview: displaced workers often do not find equally good jobs; demand shrinks, inequality deepens, and capitalism “eats itself” as labor income collapses.
  • Broken Window fallacy and Solow paradox are invoked; several stress that aggregate productivity gains can coexist with severe distributional harm.

Open vs closed AI, geopolitics, and regulation

  • Closed US labs are seen as trying to monopolize compute and infrastructure; Chinese open-weight models are framed as “wrench in the system,” commoditizing AI globally.
  • Speculation about bans on very large models, US–China competition, and whether regulation will entrench corporate “AI imperialism” or fail, leading to “cyberpunk anarchy.”

Meta / sentiment

  • Tone mixes excitement about AI capabilities with strong bubble, Ponzi, and “AI hype” accusations.
  • Some complain HN is skewed toward AI boosterism and Anthropic marketing, despite recent market turmoil.