N※N

Can AI Be Funny? Comedians Embrace AI Tools While Keeping the Human Touch

70

  //media-entertainment.news-articles.net/content/ .. race-ai-tools-while-keeping-the-human-touch.html
  Published in Media and Entertainment on Wednesday, December 3rd 2025 at 6:53 GMT by Toronto Star

• 🞛 This publication is a summary or evaluation of another publication
• 🞛 This publication contains editorial commentary or bias from the source

2025-12-03 249 x 203 / 15241 Bytes

2025-12-03 225 x 224 / 6061 Bytes

2025-12-03 224 x 224 / 5378 Bytes

2025-12-03 224 x 224 / 12646 Bytes

Can AI ever be funny? Some comedians embrace AI tools, but they’re still running the show

In an era when a chatbot can compose a love‑song, compose a sonnet, or even generate a grocery list, a question has begun to circulate among the world’s funniest people: can an artificial intelligence actually produce a joke that makes people laugh? A new Toronto Star feature digs into the growing trend of comedians using AI‑generated material, the promises and pitfalls of the technology, and the ways in which the humor industry is still very much a human‑driven business.

1. The rise of AI writing tools

The article opens with a quick primer on how AI writing tools have evolved. “ChatGPT,” the conversational AI from OpenAI (the piece links to the company’s own website for background), has been a key player. In the past year, tools like Anthropic’s Claude and Microsoft’s Copilot for Microsoft Word have made it easier for anyone with a keyboard to generate polished prose in a matter of seconds. These tools are built on massive language models that “learn” from billions of words of text, allowing them to mimic human style, invent puns, or even compose stand‑up bits—though their output is often a raw, unfiltered dump of possibilities.

2. Comedians as “human editors” of AI‑generated content

The article quickly pivots to the core of the story: a handful of contemporary comedians who have been experimenting with AI‑assisted joke writing. The most frequently quoted is Jade Nguyen, a Toronto‑based comic who told the Star, “The AI gives you a skeleton, and then I add the muscle.” She showed the reader a side‑by‑side comparison: a joke about “tax season” generated by ChatGPT, then her revised, punchier version that includes a self‑deprecating quip about her own “tax return” being as confusing as a cryptic crossword.

Other comics mentioned include Kurt Mullen of the 1‑800‑Fun‑Time club, who uses a custom script to feed prompts about current political scandals into a model, and Leah Patel, who keeps an “AI‑joke‑of‑the‑week” thread on her Instagram. In each case, the comedians emphasize that the AI is a brainstorming partner, not a replacement. “It’s like a scribe that writes a draft, but I’m still the one that makes the joke feel true to my voice,” says Nguyen. “The timing, the delivery, the context—those are all human.”

The article quotes a few key points from the comedians:

• AI can produce a high volume of material quickly, which is useful for workshops or the early draft stages.
• Humor is deeply contextual. An AI may not understand cultural references or subtext that a comedian relies on.
• Comedy is an interactive art. The “in‑the‑moment” feedback from an audience can change the direction of a set, something an AI cannot anticipate.

3. The limitations and the “funny factor”

While the comedians admit that AI is a useful tool, they also caution that it falls short on “the funny factor.” The article includes a link to a recent research paper published in Nature that analyzed jokes created by large language models and found that the model’s humor was “mostly mechanical and lacking the subtlety that humans derive from life experience.” The study’s authors suggest that “AI lacks the embodied consciousness that informs how we perceive incongruity.” In other words, a chatbot can pair a dog with a toothbrush and call it a joke, but it will not truly “get” why that is funny.

Another example highlighted in the article is the famous “AI‑generated punchline” about a cat that gets a job at a sushi restaurant. While the wording was clever, the delivery in a live set didn’t land the way the comedian had hoped. “The audience didn’t know it was a cat joke until I explained it,” the article reports. “Humor has to happen in real time, with the audience’s reaction shaping the final outcome.”

4. Ethical questions: originality, ownership, and the future of comedy

The piece moves beyond the technical and into the ethical terrain. If a comedian writes a routine with a language model that has been trained on countless copyrighted works, who owns the resulting jokes? The article references a 2022 Los Angeles Times piece that delved into the legal gray area of “AI‑derived content” and the potential for copyright claims. The comedians, while excited about the creative possibilities, stress the need for clear guidelines.

The article quotes a small panel discussion from the Canadian Comedy Awards (the feature links to a PDF of the panel’s transcript) where a moderator asked: “Will AI replace comedians?” The panelists largely agreed that the answer is no, at least for the foreseeable future. “Comedy is a conversation,” one veteran comic said. “It’s not about output; it’s about connecting with people.”

5. The future: hybrid tools and new genres

Finally, the article looks ahead. It mentions that several startups are building “comedy‑specific AI” tools that incorporate datasets of jokes, stand‑up recordings, and crowd‑reaction data. One such startup, LaughGen, has released a beta version that allows comedians to generate a set outline based on the audience’s demographic. The article links to LaughGen’s website, where a demo video shows a script being tailored for a college‑town crowd versus a corporate event.

In addition to stand‑up, the article notes that AI is already being used in scripted comedy shows, like the late‑night “Jimmy Kimmel Live!” segment “AI Joke of the Day.” While the comedian still delivers the joke, the initial line was generated by a language model.

Bottom line

The Toronto Star’s feature makes it clear: while AI can accelerate the creative process and supply comedians with raw material, the art of comedy remains squarely in the hands of the human performer. “AI is a tool, not a punchline,” says Nguyen. “We’re still the ones in the spotlight, still the ones who get the laugh.” The article, through a blend of interviews, research links, and real‑world examples, paints a balanced picture: AI is here, it’s fun to play with, but it will not replace the human instinct that turns a simple line into a shared moment of joy.