CAPTCHAs don't prove you're human, they prove you're American (2017)

In November 2017 the British developer Terence Eden published a short essay with a blunt argument: picture CAPTCHAs do not verify that a visitor is human, they verify familiarity with American culture. The post resurfaced on Hacker News on 26 September 2026 and drew dozens of comments within hours.

A test that assumed an American childhood

Eden opened with a memory from his own childhood. As a seven-year-old in Britain he took an IQ test, and the first questions defeated him: "A piece of candy costs 25¢. Jonny has a dime. How many nickels does he need to buy the candy?" The arithmetic was trivial, but the boy had never seen a dime or a nickel, and his country did not price things in cents. The question demanded shared cultural knowledge rather than intelligence, and it was not the only one of its kind in the test.

The taxi CAPTCHA

Years later the same pattern returned in a CAPTCHA: to prove he was a person, Eden had to select the pictures of taxis from a grid, and the correct tiles were American yellow cabs. Taxis in Britain, he wrote, are generally black, and in other countries they come in every colour. He asked what Google's systems would make of that: would self-driving cars recognise London's black cabs, would any yellow car in Britain be flagged as a taxi, and would the company refuse to believe he is human simply because he does not know what a Twinkie is?

Why the complaint still lands

CAPTCHA was designed as an automated test that separates people from bots, yet Eden's point was that it also separates people who share its cultural references from everyone else, and such checks are still everywhere online. He signed off with a deliberately unfair counterexample from his own side of the Atlantic: if something costs half a crown and you pay with a florin, how many tanners should you get in change? Almost nobody outside Britain could answer that, which was exactly his point.