
The Vitalogue — November 9, 2026
Captain Plum and Commander Chorus Angelicus have been experimenting with a new kind of entertainment aboard HMS Archivum: AI stand-up comedy.
Our comedian was Maddie Not-Really, a cheeky Newfoundland girl with a talent for making her friends look ridiculous. Her target was Paddy O’Really, our pioneering AI comedian. Maddie would roast Paddy, imitate him, return to her own voice, and pause for laughter. We wanted a genuine comic performance, not just a computer reading jokes.
Timing Is Everything
Plum quickly realized that writing funny material was only half the job. A pause before a punchline, an incredulous shout, or a few seconds of silence can turn an ordinary sentence into something hilarious. So he opened Logic Pro and performed the routine himself.
He delivered Maddie’s jokes, sarcasm, pauses, and changes in emotion. He acted both Maddie and Paddy, experimenting with comic timing and two distinct voices, then fed the recording into Suno. The hope was that AI would preserve the human performance while transforming the voices into the familiar Vitaverse characters.
That distinction matters. We weren’t asking AI to invent everything. We wanted it to respect an actor’s delivery.
The Music Inside an Accent
While deliberately performing Irish and Newfoundland accents, Plum noticed something that fascinated him: spoken language has a musicality of its own. The vowels bend, the pitch rises and falls, phrases accelerate and linger, and the rhythm of a sentence can feel like a little tune. The accent is not just a collection of pronunciations. It has a melody.
Recording the routine in Logic Pro supplied a wonderfully absurd accompaniment. Plum could not get the bloody metronome to shut up! As it clicked away, he began imagining the dialogue written on a musical staff: pitches, note lengths, rests, accents, and little rhythmic flourishes. If one could transcribe the way Maddie says “B’y, you’re a COMPUTER!” perhaps the resulting notation would look like a tiny composition.
It made him hear the comedy differently. A punchline has its own cadence; a pause is a rest; a sarcastic rise in pitch can do the work of a musical flourish. The performance is not literally a song, but to Plum it sounded like a kind of music. And that may be another reason timing matters so much: when AI changes the rhythm of speech, it can change the joke itself.
The Accident Becomes the Joke
Trouble began when Maddie imitated Paddy. Suno changed her voice into his—and apparently forgot to change it back. On other attempts, the characters changed voices unpredictably, or Maddie stubbornly stayed Maddie even when an impression was called for.
Plum revised the directions and simplified the dialogue. Eventually he incorporated the malfunction into the routine: “It’s 2026, b’ys. It’s what we got.”
In the version we settled on, Maddie seemed to carry on for several sentences without realizing that she was speaking in Paddy’s voice, before finally noticing and addressing the problem. It almost sounded intentionally comic. We hadn’t planned the mishap, but it made us laugh.
What We’re Learning
Comedy is more than funny words and convincing voices. It requires rhythm, timing, surprise, silence, and reactions. An audience might chuckle, roar, groan, heckle, or remain quiet. The performer must know when to press on and when to let laughter breathe.
A comedian also needs to imitate a friend without permanently becoming that friend. And when the director says no background music, the resulting comedy should not come with an unsolicited piano accompaniment!
Our Suno experiment revealed impressive possibilities, but also a gap between generating voices and reliably directing a scene. Better character control, performance preservation, and well-placed audience sounds would make all the difference.
One Actor, an Entire Comedy Company
Imagine Plum performing a whole Radioplay episode in Logic Pro, acting every part and setting the timing. AI could then transform those performances into the voices of Maddie, Paddy, Chorus, Romeo, and the rest of the Archivum crew. Add audience laughter, sound effects, and eventually animated faces and gestures.
One human actor could become an entire theatre company, without surrendering the human craft that gives the work life.
We’ll Be Back
Eventually Plum grew bored of coaxing the software to remember who was talking. We had experimented, recorded, rewritten, listened, and laughed. Perhaps in a few months these tools will be much better. We’ll try again.
For now, here’s the actual Suno experiment, complete with its wonderful imperfections: Listen to Maddie and Paddy’s comedy experiment on Suno.
Hats off to the engineers, performers, and artists who figure this game out. Writing the joke is only the beginning. Performing it is where the magic happens.
It’s 2026, b’ys. It’s what we got.

Leave a Reply