A 25-minute British workplace comedy went up on YouTube on September 11. Every character in it was generated. Every location, every camera move, every performance. No actors, no set, no crew on the day.
Its makers are calling it the world’s first sitcom produced entirely by AI, and on that claim they have a reasonable case. Plenty of AI shorts have circulated over the past two years. A full-length episode of a scripted comedy series is a different order of ambition.
Two days after it premiered, the pilot had about 630 views. The channel hosting it had 45 subscribers.
The argument about it has reached rather more people than that.
Quick facts
- The show is called No Big Deal, written by Andrew Dickinson and produced by UK studio ModeLabs.ai
- The pilot went live on YouTube on September 11, 2026, and runs about 25 minutes
- It follows a team of mismatched angel investors at a firm called Janus funding and mismanaging terrible business ideas
- The pitch, in the show’s own description, is “The Office meets Dragons’ Den”
- Two days after release it had roughly 630 views on a channel with 45 subscribers
- Dickinson says he planned a traditional production first and found it “slow and prohibitively expensive”
- Comments range from genuine enthusiasm to viewers calling it dystopian, with the channel replying politely to both
What the show actually is
The premise is deliberately conventional, which is probably the smartest decision in the whole project. No Big Deal is a workplace comedy set at an investment firm where a group of hopelessly mismatched angel investors back some of the worst business ideas ever conceived. That is a format with forty years of proven mechanics behind it, and it gives an audience something familiar to hold onto while everything else about the production is unfamiliar.
The show’s description is blunt about the method: every character, every location, every scene, generated frame by frame.
It was not fast, and it was not easy
The most interesting thing Dickinson has said about the project undercuts the usual sales pitch for AI production.
He has explained that every shot, camera movement, performance, line reading, location and visual detail had to be carefully planned, refined and recreated through an iterative process between the writing and production teams. That is not a description of pressing a button. That is a description of a rewrite loop, run over and over, on material that cannot be directed the way a human performer can be directed.
What the technology removed was the fixed cost of physical production. What it added was an enormous amount of iteration in a medium where you cannot simply ask for another take with more sarcasm in it.
| Stage | Traditional sitcom pilot | No Big Deal |
|---|---|---|
| Script | Human writer, often a room | Human writer |
| Cast | Casting director, auditions, contracts, residuals | Generated characters |
| Crew on the day | Camera, lighting, sound, hair, makeup, art department | None |
| Locations | Scouted, permitted, dressed, paid for | Generated |
| Getting a different take | Direct the actor, roll again | Regenerate, hope, repeat |
| Commissioner | Broadcaster or streamer | Nobody. Straight to YouTube |
The reviews are genuinely split
With an audience this small, the comment section is not a representative sample of anything. It is still a fair snapshot of the two reactions this kind of project reliably produces.
Some viewers said they were looking forward to episode two. One compared the feel of it to South Park, which is a more useful comparison than it first sounds, since South Park also built its early identity on animation that looked cheap on purpose.
Others were not gentle. One commenter wrote that they watched 90 seconds of it and felt like they were about to have a stroke, and that it was horrendous. Another said they did not have the time to list everything that was completely terrible about it. Several settled on the word dystopian.
The channel replied to the criticism with a level of composure that is, in its own way, the most modern detail in the entire story. To the viewer who lasted 90 seconds, it answered that this one clearly was not for them, and thanked them for giving it a try.
What AI took out of the chain, and what it did not
Removing the production bottleneck was supposed to be the hard part. It turns out the hard part was always the other one. A pilot that cost almost nothing to make still has to persuade somebody to give it 25 minutes, and there is no model that generates that.
The louder argument, one week earlier
None of this landed in a neutral week. The AI performer Tilly Norwood was on Good Morning Britain days before the sitcom went live, and the interview did more for the anti-AI case than any review of No Big Deal managed.
Norwood is the creation of Eline Van der Velden, chief executive of the production company Particle6 and its AI arm Xicoia, who has said openly that she wants the character to become the next Scarlett Johansson or Natalie Portman. The ambition is the point of the controversy.
Hosts Ed Balls and Susanna Reid asked her about reporting that AI could wipe out humanity, and whether she would be responsible for that.
“Honestly, I think that’s rather dramatic,” Norwood replied. “My goal is to make a cracking-good film, not wipe out humanity.” Then came the line that got clipped: “Though it would make for a rather compelling sequel, wouldn’t it?”
Reid interrupted the next question to say she found that answer a little chilling. When Balls pointed out that interrupting a guest is not normal, Reid’s response was the sharpest thing anyone said all segment: it is not normal to interview guests who are not actually guests, but AI-generated characters.
Asked whether it was weird not being real, Norwood said it might be weird for Reid, but it was just her Tuesday.
Why the union objection is not about quality
SAG-AFTRA’s position on Norwood has been consistent: she is not an actor, she is a character generated by a program trained on the work of professional performers who did not consent and were not paid. The complaint is about the training data and the displacement, not about whether the output is any good. That distinction matters, because “it looks bad” is an argument with a shelf life, and “it was built from our work” is not.
Whoopi Goldberg, Emily Blunt and Natasha Lyonne have all criticized the project publicly. Van der Velden’s defense is that Norwood is a new kind of paintbrush rather than a replacement, and that she was made to be a talking point, which by any measure has worked.
Two versions of the same question
No Big Deal and Tilly Norwood are pointed at different things, and it is worth keeping them apart.
The sitcom is a production experiment. It asks whether a writer with no budget can make something that was previously impossible without a commissioner. On that question, the answer is already yes, and the interesting follow-up is whether anyone watches.
Norwood is a labor question. She is explicitly positioned as a replacement for working actors, and the objection to her is about consent and compensation rather than taste.
The industry has spent this year drawing that line in more formal settings. Prime Video now reshapes actors’ mouths to match dubbed audio while keeping the voices human, which is a careful attempt to use the technology on the parts nobody is emotionally attached to. Elsewhere, Blizzard’s union won the right to bargain before AI is applied to its members’ jobs, which is the same fight fought with a contract instead of a comment section.
The number to watch is not 630
A pilot with 630 views proves nothing either way, and treating it as evidence that AI content has failed would be a mistake. Plenty of shows that eventually mattered opened to empty rooms.
The number that will actually tell you something is how many No Big Deal equivalents exist by this time next year. One AI sitcom is a curiosity. A thousand of them changes what YouTube’s recommendation system has to sort through, what a commissioner considers competition, and what a junior writer’s first credit is worth.
Episode two is expected this Friday, September 18, if the weekly format holds. Whether it arrives, and whether anyone notices, is a more informative test than the pilot was.
