In an ambitious new research project, a Melbourne academic is attempting to answer a question that has long puzzled technologists and comedians alike: can artificial intelligence ever learn to be genuinely funny? Dr Robert Walton, a dean's research fellow at the University of Melbourne's Faculty of Fine Arts and Music, has secured an Australian Research Council grant of approximately $500,000 to find out.
The Non-Verbal Approach to Machine Humour
Unlike common AI tools like ChatGPT, which often produce weak, formulaic jokes, Dr Walton's project takes a radically different path. He argues that the core of comedy isn't just about words, but about timing, reading the room, and physical connection with an audience. Therefore, his research will begin by focusing entirely on non-verbal communication.
"Robots are good at making people laugh… they are humorous because they break and they bump into things, and so we’re laughing at them," Walton observes. "However, when they try to do something funny on purpose, it ain’t so funny any more. We don’t laugh at them because we really, deep down, don’t believe that they can be funny."
To build this belief, he will work with an ensemble of roughly 10 robots. These won't be humanoid androids, but ground-based vehicles ranging in height from 40 centimetres to 2 metres. Their initial training will involve learning to be funny visually, through movement and interaction.
Teaching Robots to Sense a Room
The project's methodology involves equipping the machines with enhanced sensory capabilities, akin to human senses. They will be designed to detect movement, the tilt of a head, or the sound of laughter. "We’re giving these systems more senses… giving them ears, not just listening for words but for things like the gaps in between words, the rhythms of things," Walton explains.
He likens the process to teaching a baby, providing the AI with multiple ways to sense its environment and build a holistic understanding of social cues. The standup comedy format is seen as the perfect testing ground due to the clear, immediate feedback loop between performer and audience.
While the initial phase is wordless, Walton confirms that adding voices is a "potential" future step, depending on the project's progress.
The Double-Edged Sword of Robotic Humour
Walton is acutely aware of the broader implications and inherent tensions in his work. While the creative industries, including comedy, are grappling with the threat of AI to human jobs and originality, his goal is not to replace comedians.
Instead, the research aims to investigate whether robots can be taught believable comedy, thereby improving our understanding of human-robot interaction. This knowledge could reveal both the benefits and risks of machines using humour, which can be used to disarm or to manipulate.
"But while I’m looking into this work of building belief in comedy performance by machines, I’ve got this other eye on what does it mean, and how might this be used coercively?" he questions. The same techniques that could allow a robot to perfect comedic timing might also be used by a care robot to say the right thing at the right moment to cheer someone up.
The comedy world remains sceptical. At the Edinburgh comedy festival, Tina Fey stated AI is "unable to be funny." Australian comedian Tim Minchin suggested that humanity is interested in "the agency of their fellow human behind the art, struggling, striving, making choices and errors." He believes AI may replicate perfection but never our flaws, which are central to our humanity.
Susan Provan, director of the Melbourne Comedy Festival, echoes this, stating that enjoyment stems from "authentic human originality" and lived experience. "You’d be laughing at the robot stuffing up. That’s what would be funny," she concludes.
Dr Walton's groundbreaking experiment sits at the intersection of technology, art, and ethics. It seeks not just to program a punchline, but to decode the very essence of what makes us laugh, and to understand what happens when a machine tries to join the conversation.
