What is High Agency?Definitions, examples, and how to develop it.

20 Examples of High Agency in the Wild

From whatishighagency.com ·

What does high agency actually look like in practice? Below are twenty cases, each tied to a primary or near-primary source. The point is not to celebrate the names (some of them are contested for unrelated reasons) but to make the pattern concrete: a constraint everyone treated as fixed turned out to be negotiable, and the person did the negotiating.

The companion essay What is High Agency? contains the definitions and origin of the term; Low Agency describes the opposite pattern; How to Build covers the practical drills. This page is the gallery.

1. Patrick Collison emails the Bay Area as a teenager

While still in rural Tipperary, Ireland, Patrick Collison wrote to prominent technologists — engineers, founders, professors — with specific questions and proposals. Most replied. By the time he and his brother John founded Stripe, they had a network most people don’t acquire by their thirties. The high-agency move was treating credentials and proximity as soft constraints, not hard ones.[1]

2. Elon Musk teaches himself rocketry

Before founding SpaceX, Musk concluded that the cost of getting to Mars was the binding constraint on space exploration. Rather than raise capital first, he read aerospace textbooks and held meetings until he could argue with engineers about whether a $100,000 part could be made for $5,000. The argument became the company.[2]

3. Apollo 13 — improvising the CO₂ scrubber

When the spacecraft’s air filter failed in 1970, Ed Smylie and the NASA Crew Systems Division designed a workaround using only materials aboard the capsule: duct tape, plastic bags, a sock, the cover of a flight manual. The high-agency move was constructing a solution from what was at hand rather than what was specified.[3]

4. Aaron Swartz publishes the Guerilla Open Access Manifesto

At twenty, Aaron Swartz wrote the Guerilla Open Access Manifesto, arguing that knowledge locked behind paywalls should be liberated and shared. He moved from critique to high-risk action. The case is morally and legally contested, but the disposition — refusing to treat the prevailing arrangement as fixed — is unmistakable.[4]

5. Drew Houston records a screencast for Hacker News

After Y Combinator initially passed on Dropbox, Houston filmed a casual demo and posted it on Hacker News, taking the product directly to the audience that mattered. The application got attention; the company got funded. The move was changing the venue of the pitch, not the pitch itself.[6]

6. Demis Hassabis talks chess with Peter Thiel

At a private drinks party, Hassabis approached Thiel and used a specific chess insight to hold his attention long enough to secure a meeting. That meeting led to early backing for DeepMind. The opening was unconventional, and it was the only one available to him.[5]

7. Stripe’s “speed of fruitful iteration”

In Stripe’s earliest days, Patrick and John Collison physically went to users to set them up, compressing the feedback loop. Patrick: “the main thing I think companies screw up at the pre-product-market-fit stage is sort of speed of iteration. Speed of fruitful iteration.”[1]

8. Airbnb sells election cereal

In 2008, near-bankrupt, Brian Chesky, Joe Gebbia, and Nathan Blecharczyk financed Airbnb by selling election-themed cereal — “Obama O’s” and “Cap’n McCain’s” — at the political conventions. They generated cash through an unrelated creative move rather than accept failure.[7]

9. Melanie Perkins through ~100 VC rejections

Before Canva, Perkins built Fusion Books as a wedge product, then raised money through what she counted as roughly a hundred rejections. Persistence alone is not the lesson; each rejection was treated as data, not as a final answer. Canva is now one of the most valuable software companies built in the last decade.[8]

10. Palmer Luckey’s garage Oculus prototypes

As a teenager, Palmer Luckey built head-mounted display prototypes in his parents’ garage and posted them to MTBS3D, a 3D-modding forum. He launched the Oculus Rift on Kickstarter — not the conventional path for hardware. The path didn’t exist for him; he made one.[9]

11. Alfréd Wetzler escapes Auschwitz

The most extreme case in this list. In April 1944, Slovak Jewish prisoner Alfréd Wetzler escaped Auschwitz with Rudolf Vrba and produced what became the Vrba–Wetzler Report — a detailed, first-hand account of the camp’s mass-murder operations. The report reached Allied governments and is credited with influencing the halt of deportations of Hungarian Jews later that year. George Mack uses Wetzler as the upper bound: not optimisation under comfortable constraints, but action under the worst constraints anyone has faced.[20]

12. Cate Hall leaves law for poker, then biotech

As a young litigator, Cate Hall asked, in print, why anyone takes the conventional career steps when most are not enforced. She left, became a professional poker player, won a major tournament, then started a biotech company. The question that drove the move was, by her later account, the move itself.[19]

13. Vitalik Buterin proposes Ethereum

At nineteen, after identifying limitations in Bitcoin’s scripting language that the community wasn’t going to fix, Vitalik Buterin proposed an alternative protocol. The Ethereum whitepaper is what came of acting on the constraint rather than complaining about it. The new platform reshaped an industry.[10]

14. Jensen Huang ships DGX-1 before the market arrives

NVIDIA built and shipped its first AI supercomputer, DGX-1, in 2016 — before there was a clear customer base for it. Huang committed capital to a product the market had not yet validated. The market caught up later. The high-agency move was treating the timing of demand as something a company can pull forward.[11]

15. OpenAI ships ChatGPT on a two-week timeline

In late 2022, OpenAI’s leadership decided to release ChatGPT publicly on roughly a two-week schedule, against internal concerns about readiness. The release reshaped the AI industry. The move was deciding under uncertainty rather than waiting for certainty that wasn’t coming.[12]

16. Cursor enters a market everyone called crowded

Anysphere’s founders pivoted into AI coding tools believing the space was already competitive — and decided to go for it anyway. They built rapidly, out-iterated incumbents, and ended up with one of the fastest-growing developer tools in history. The high-agency move was acting on conviction in a space the consensus said was full.[13]

Rather than wait for Google to ship a chat-based search product, Srinivas built one — an answer engine that summarised the web with source footnotes. The product challenged the default search UI by inverting it. Perplexity now has a real share of the answer-engine category and forced incumbents to respond.[14]

18. David Holz launches Midjourney on Discord

Instead of building a standalone app, Holz launched Midjourney inside Discord, taking advantage of an existing community infrastructure. The high-agency move was using a venue that wasn’t designed for the product but worked better than anything custom would have, and shipping years before the “right” consumer surface area existed.[15]

19. ElevenLabs from a translation grievance

Mati Staniszewski and Piotr Dabkowski turned frustration with poor dubbing — a specific lived irritation — into a voice-AI company. The product solved the problem they personally had. The high-agency move was treating their own annoyance as a market signal.[16]

20. Andrej Karpathy starts Eureka Labs

After leadership roles at Tesla and OpenAI, Karpathy started Eureka Labs, an AI-and-education company, having treated education work as a “side quest” for years. The high-agency move was reweighting personal conviction over professional default. Prestige tracks were the constraint he chose to relax.[17]

See also

References

  1. Patrick Collison, Running Your Startup transcript.
  2. TIME, Elon Musk learning aerospace.
  3. NASA / Ed Smylie, Apollo 13 oral history, 1999.
  4. Aaron Swartz, Guerilla Open Access Manifesto, 2008.
  5. Rory Cellan-Jones, Demis Hassabis approaches Peter Thiel.
  6. Drew Houston, On starting and scaling Dropbox (YC Library).
  7. Fortune, Brian Chesky on the Obama O’s and Cap’n McCain’s cereal, April 2023.
  8. Yahoo Finance, Melanie Perkins on ~100 VC rejections.
  9. Palmer Luckey, Oculus Rift HMD prototype thread, MTBS3D.
  10. Vitalik Buterin, Ethereum whitepaper.
  11. Jensen Huang (via OfficeChai), On building DGX-1 before the market arrived.
  12. Futurism, OpenAI’s two-week ChatGPT release timeline.
  13. Fortune, Cursor / Anysphere founders pivot into AI coding, April 2026.
  14. Fortune, Perplexity CEO Aravind Srinivas.
  15. The Verge, David Holz interview on Midjourney via Discord, August 2022.
  16. Sequoia Capital, Mati Staniszewski (ElevenLabs) on Training Data podcast.
  17. Andrej Karpathy, X post announcing Eureka Labs, July 2024.
  18. Yahoo Finance, George Hotz / comma.ai working around regulators.
  19. Cate Hall, “Can You Just Do Things?”, Asterisk Magazine, 2026.
  20. George Mack, Thread on high agency thinking (X), 30 November 2018.