By Steven Morris
Publication Date: 2025-11-24 07:59:00
Comedians who rely on clever puns and witty headline writers may be a little calmer, at least for now, AI research shows.
Experts from universities in the UK and Italy have investigated whether large language models (LLMs) understand puns – and found them lacking.
The team from Cardiff University in South Wales and Ca’ Foscari University in Venice concluded that LLMs could recognize the structure of a pun but didn’t really understand the joke.
One example they tested was: “I used to be a comedian, but my life became a joke.” Replacing this with “I used to be a comedian, but my life became chaotic,” LLMs still tended to perceive the presence of a pun.
They also tried: “Long fairy tales have a tendency towards dragons.” When they replaced “dragon” with the synonym “extend” or even just a random word, LLMs seemed to think it was a pun.
Prof. Jose Camacho Collados from the Department of Computing and Computing at Cardiff University,…
