AI-Generated Jokes Are Funnier Than Yours

A psychology study published by USC in PLOS ONE demonstrates that  AI-generated jokes are funnier than people-created ones.

The research team behind the study compared participants’ reactions to jokes written by ChatGPT 3.5 and others written by mere people. The goal was to determine if AI can outwit humans for a laugh. Participants rated the jokes without knowing their author.

Findings include:

• Nearly 70% of the participants rated ChatGPT-generated jokes as funnier than those crafted by humans
• Just over 25% of participants rated the human-crafted responses as funnier, while 5% rated jokes from both sources equally amusing.

Results were consistent across various demographics.

A second study challenged ChatGPT to develop new headlines in the satirical style of The Onion and had 200 participants rate the funniness of original Onion headlines and new ChatGPT-generated headlines. The verdict was that ChatGPT’s headlines were as funny as the original Onion headlines.

For the first study, both ChatGPT and 105 participants completed three tasks, including three prompts. In the first task, they crafted humorous new phrases for common acronyms. In a second task, ChatGPT and the human writers generated funny answers to fill-in-the-blank phrases such as, “A lesser-known room in the White House: _____.” They were also prompted to devise a “roast joke,” a funny slam inspired by an awkward fictional scenario.

One hundred and five writers wrote more than 945 jokes. ChatGPT generated 20 humorous answers for each prompt for a total of 180 jokes. A new group of participants then rated the funniness of the jokes.

In the second study, ChatGPT was given original headlines from The Onion and asked to generate new headlines in the same style. The team asked two hundred participants to rate the funniness of the original Onion headlines and the ChatGPT-generated ones without knowing the authorship.

The study originated with the entertainment world’s debate over writers’ concerns about using large language models (LLMs) like ChatGPT in entertainment production. Recent strikes by Hollywood writers and actors were partially due to the fear that the adoption of LLMs poses an existential threat to their respective professions, to art, and human creativity.

Click here to read the full study at PLOS ONE.