Story Mode: AI masterclass on storytelling

AI is getting louder. But storytelling? It’s getting lost in the noise.

Uncrafted by a professional human in the loop, most AI content barely holds my attention for five seconds.

Because this is where AI still struggles:

• Rhythm. It doesn’t know when to slow down, when to lean on a line, or when to just pause and let something sink in. It can’t skillfully build layers, take it to a crescendo and then leave you with the feeling that you just witnessed or read something truly unique.

• Emotion. Not just the obvious kind. The breath before a reveal, the quiet shift in tone, the crack in the voice that tells you this matters. Emotion that reveals a persons history in a look, a characters past in a sigh.

• Subtext. Humans speak in layers. What we say is often not what we mean. If you just watch one episode of a great well written drama series, e.g. Mad Men, The Wire, every line is a rich cake with layer upon layer of meaning and depth.

But there are two AI generated experiences I had this month that reignited my passion for the art of story and that told me it’s not a question of if AI will get there, but when.

NotebookLM in Story Mode

NotebookLM simply takes any source material you can throw at it and becomes a tool to explore it in creative ways. What struck me most about it wasn’t the content it produced—it was the format it chose.

It didn’t just generate an outline or list of storytelling tips.

It made a podcast.

That choice matters.

Because podcasts are one of the few mediums where AI can start to feel less like a tool—and more like an experience. Spoken out loud, the content moves differently. It breathes. It lands in rhythm. It mimics a human exchange rather than the usual sounding board.

And that’s really key:
AI has always struggled to be emotionally resonant. But wrap it in a podcast, and suddenly it creates a story rich world.

The episode didn’t blow me away with insight—but it made me listen. It wasn’t flat or robotic. It had pacing. Tone. A sense of flow.

NotebookLM’s podcasts aren’t storytelling in the classic sense. But by using a podcast to frame its output, it hints at something bigger:

AI can move closer to story not by imitating plot—but by mimicking conversation.

Here is a clip from the first episode of StoryMode: An AI Generated Podcast (reply and let me know if you want the full episode and i’ll make it.)

If NotebookLM showed me how AI can speak with more rhythm and intention, Sesame showed me how it can start to feel like someone’s really there.

Great stories create a world.

And the better the storyteller, the less you notice the seams.

The illusion becomes complete.

You start to forget it’s a performance.

You feel like you’re there.

That’s what surprised me about interacting with Sesame’s voice AI (free demo here).

It wasn’t the realism on the surface—it was the small, finely tuned details that worked quietly underneath:

A pause timed just right.

A soft shift in tone at the edge of a sentence.

Those subtle nuances have a strange power. They pull you into a scene, and suddenly you’re not talking to an AI—you’re immersed in something. A space. A moment. A story.

As these tools improve—not just in what they say, but how they say it—they’re starting to cross that invisible line. The line between assistant and narrator. Between output and experience.

The more nuance they capture, the more immersive the story becomes.

And the more immersive it becomes, the more it starts to feel… real.