The cursor was blinking. My home studio was quiet. I was drafting a monthly blog post for Mutable Fire — my astrology-themed publication I write under the pseudonym of Daljeet Peterson, a spiritual name I’ve used since 2020 for work aligned with my wider spiritual practice.
This time, curiosity got the better of me.
I wondered: could ChatGPT write this as well as I can… maybe even better?
So I opened ChatGPT 3.5 (which was the current model at the time). The text box prompted: “Ask anything.”
The synthetic voice was waiting for me to speak first.
“Alright,” I said, half to myself, half to the machine. “Let’s see what you can really do.” I typed: “Write me a short blog post on the current astrological weather.”
What came back was confident, cleanly written… and completely wrong. Not just a little off — the kind of wrong that made me wonder if the machine and me were even looking at the same sky. Full blown hallucination.1
That should have been the end of it. But something about the way it almost worked made me lean in instead of logging out. I corrected the errors, fed back my interpretations of the sky, and watched it return something surprisingly usable.
This experiment as Daljeet was my first foray into human–AI collaboration. And now, Dead Air is the next evolution — a bigger, riskier, and more public one.
The Silence After the Noise
I’d been hearing about AI for a while — the hype, the fear, the hot takes multiplying like rabbits in a Reddit thread. But until that moment writing as Daljeet, it had been mostly background noise.
In 2023, that noise was easy to ignore. I was deep in my best professional year to date, juggling multiple TV shows, clocking long hours in the edit bay. Whatever “revolution” was coming could wait until I had the time to check it out.
Then 2024 hit. The work stopped. Completely. In my thirty years of making a living in media and entertainment, I’d never seen a year like it. The conversations in my corner of the industry had shifted: successful series being shelved, productions going dark, and there was an unspoken but growing anxiety about what generative AI might mean for professionals like me.
The public discourse was a veritable hall of mirrors. On one side, you had techno-doomers warning of creative extinction; on the other, trans-human optimists promising a golden age. Both felt way too certain, and way too loud. I needed to check it out for myself.
So I started experimenting. First with ChatGPT 3.5, then with image models like Stable Diffusion, music models like Suno, and a few other tools. The early results were clumsy, sometimes laughable — but there was something there. Not just as a novelty, but an emerging set of tools I could actually see myself working with. Tools that, in the right hands, might bypass the gatekeepers entirely.
That’s when I decided: I wasn’t going to wait for the industry to tell me how this was going to go. I was going to run my own experiment, in public. Dead Air is becoming that experiment.
Test Signals
Dead Air is conceived as a creativity lab where human ingenuity and machine intelligence meet to make stories, write essays, and run media experiments — with the process made visible as part of the work.
That means you’ll see the seams. You’ll see the moments when I steer the machine, when it surprises me, when it misses the point, and when it nails something I couldn’t have written or created alone.
This is not an AI fan club. It’s not an anti-AI manifesto. It’s not a tech blog or a productivity hack feed. I’m not here to report the latest model release or debate AI ethics like it’s an abstract thought experiment.
[Process interjection]: Notice the literary tic here. LLM's like ChatGPT just love to tell you what things are not. They programmatically default to structuring sentences that follow a “not X, but Y” logic. It’s called antithesis2 in classic rhetoric. You're probably seeing it everywhere these days.
Sure, it's effective communication, but when overused it’s annoying— the indelible fingerprint of predictable AI output. As a human, I'm more inclined to lead with what a thing “is” rather than what it “is not.”
So...
Dead Air is about using the tools in real time, as a working creative, and showing you what happens. It’s about the collision between decades of analog experience and a new kind of digital collaborator.
The experiment is open-ended. The methods will change. The tone will shift from personal to analytical and back again. And somewhere in the middle of that oscillation, something interesting just might happen.
Why Do This in Public?
Because the only way I know how to make sense of something this big is to get my hands dirty with it. I’m not interested in watching the AI conversation from the cheap seats while people on both extremes scream about salvation or extinction. I need to see it, feel it, break it, and build with it.
And I want you to see it too. Not the polished, retrofitted “look how smart I was all along” version, but the messy middle where I’m still figuring it all out. I’ve been in creative media for thirty years — long enough to recognize an inflection point when I see one. This feels like our Gutenberg moment: a shift in the means of creative production so profound that it could reshape the culture from the ground up.
If working creatives aren’t part of the conversation now, the conversation will move on without us. So this is me staking my claim. Not as an evangelist or a doomsayer, but as a maker. My hunch is that these tools can be used to bypass the old gatekeepers — but also that they could just as easily build new ones if we’re not careful. The only way to know is to run the experiment in full view, and to be honest about what works, what fails, and what feels misaligned.
Where This Could Go
I don’t know exactly where Dead Air will land — and that’s the point. I’m not building toward some neat, TED-ready conclusion. I’m building toward a body of work that could only have been made by this human and these tools, in this particular moment in time.
Some threads might lead to fully developed screenplays or serialized fiction. Others might branch into essays that pick apart the media we’re swimming in. A few might dissolve into experiments that never make it past the prototype stage. All of it will live here — the successes, the misfires, and the beautiful mess in between.
If it works, Dead Air could spin off into other identities, other publications, even other voices entirely. It could become a network of semi-sentient creative projects, all incubated here. Or it could remain what it is now: one human and one machine, brainstorming, dialoging, and seeing what comes through.
The only thing I can promise is that you’ll get the real process, not the retroactive myth. The dial will stay open, the static will be part of the soundtrack, and if we’re lucky, the occasional moment of pure, clear signal will cut through.
Somewhere past the static, a story is already taking shape. I don’t know the plot yet. And the ending’s not written. But the mic is live, and the channel is wide open.
Stay tuned.
Footnotes:
Hallucination in AI-speak means the model generates plausible but inaccurate information. It’s not lying — it’s predicting the next word without actually “knowing” anything. The output feels confident because confidence is a statistical style, not a truth guarantee.
Antithesis is a rhetorical move that sets up a series of negations (“not X”) before pivoting to an affirmation (“but Y”). In LLM output, this is more than stylistic: it’s a probabilistic reflex baked into training data. Once the model starts the “not X” structure, it tends to complete the pattern for clarity and perceived balance. Effective in moderation, but overuse tips off the reader that a machine is at the mic.
Image Prompt:
Retro 1980s home studio desk scene at night, warm incandescent desk lamp glow mixing with faint green light from a flickering CRT monitor. On the CRT: a glowing, green phosphor astrological chart wheel, slightly distorted as if by analog interference. Desk clutter: open notebook with handwritten sketches, a fountain pen resting across it, printed star charts and old astrology books in soft shadow. Half-empty ceramic coffee mug nearby. Scene feels intimate, focused, and mysterious — the quiet moment before beginning an experiment. High detail, cinematic lighting, shallow depth of field. Leave negative space for potential text overlay.
