Kyle Harrison
article

Go Big or Go Broke — The Middle Ground Is Gone (Ethan Kurzweil)

Ethan Kurzweil August 31, 2026 View original ↗

Go Big or Go Broke — The Middle Ground Is Gone (Ethan Kurzweil)

Author: Ethan Kurzweil (writes as a VC at Chemistry) URL: https://www.linkedin.com/pulse/go-big-broke-middle-ground-gone-ethan-kurzweil-4ocyc/ Published: August 31, 2026, as a LinkedIn article One-line: Once the industry has seen that trillion-dollar startups can be built on a venture timeline, it can’t unsee them. Capital at every stage concentrates into consensus mega-bets, and the “double or triple” outcome disappears.

Summary

The pink elephant. “Don’t think about a pink elephant” is the frame: founders and investors can no longer stop imagining the trillion-dollar startup.

Two immediate effects.

  • Funding reorients at every stage. Angels, seed, VC and growth investors pattern-match to their image of a “Trilicorn” and concentrate capital in the biggest swings. He cites Peter Walker’s Carta data showing higher valuations and more dollars flowing into fewer companies.
  • Founders adjust. They drop any aspiration short of going for broke and learn that more ambition attracts more funding. The ideas that work best are those “legible to capital,” a phrase from pieces by Will Manidis and Nikunj K.

What gets lost. Venture has become extremely consensus. But the biggest winners rarely look big at first: books by mail, online snowboard stores, graphics cards for gamers. Kurzweil sees fast, repeated progress in consensus areas (silicon, inference, drug design, foundation models, coding, physical AI, prediction markets) and neglect of out-of-favor ones (SaaS, infrastructure, consumer, commerce). Progress there turns linear and depends on stubborn contrarians working without attention, like OpenAI circa 2018, Palantir’s first decade or Cerebras.

How it ends. Consensus progress will eventually plateau, the neglected pioneers will be recognized, and the cycle will revert. Not soon, though: as long as IPOs, big acquisitions and rapid up-rounds keep coming, the two venture markets won’t converge. A footnote cites Nnamdi Iregbulem on the steady decline of the seed market.

Advice (from Chemistry). The two markets aren’t mutually exclusive, and there will be gems and fool’s gold in both. For founders: if you’re out of favor, plan to stay there longer than you think and make sure the business (and your psyche) can survive it; if you’re a consensus needle-mover, you can’t step off the train mid-journey.

Notable pushback in the comments

David Hornik (visible in the public comment thread, 2026-09-15; about 25 reactions) rejected the premise. In his view it’s absurd that a multi-billion-dollar outcome now counts as a mere double, and the problem is VCs who fund companies at billion-dollar valuations, not the companies. His prescription: fund interesting companies at rational valuations and build great businesses. He expects the trillion-dollar model to work only for a few lucky investors, while everyone chasing it loses a fortune, and warns it will be ugly when the music stops.

Full text

Archived privately against link rot, with the header image: ../attachments/go-big-or-go-broke/go-big-or-go-broke.md. Hornik’s comment is summarized above, not archived verbatim.

Connections

  • Ethan Kurzweil — author; writes as an investor at Chemistry.
  • Dear VCs, Just Give Up (Michael Dempsey) — the same month’s satirical treatment of the same concentration thesis.
  • David Hornik — the rational-valuations counterargument.
  • Hunter Walk — thanked as a reader of drafts.
  • Will Manidis and Peter Walker — sources for “legible to capital” and the Carta concentration data.
  • Non-Consensus, Contrarian Investing and Power Law — the dynamics at issue.
  • Seed Investing, Palantir and Venture Capital