In this episode, I unpack the strange little AI experiment behind my recent “slime-mold internet” video — how it began as a debate between imagination and logic inside ChatGPT, hopped through NotebookLM, got re-written by another model, and finally ended up voiced by my clone. I also talk through the weirdness of recognizing AI fingerprints in other creators, what online writers get wrong about ego and craft, and how it feels to finally have a real writing partner helping turn Harbinger into a full book. This one is half confession, half behind-the-scenes, and all creative shop-talk.

Timestamps

00:00 – Opening & catching up00:23 – Revisiting the “internet as a living organism” theory00:45 – How the slime-mold episode was built with multiple AIs01:02 – Turning a debate with ChatGPT into a structured concept01:27 – Early rogue-AI / dead-internet lore context02:00 – Theory recap and podcast plug02:25 – Moving the whole experiment into NotebookLM03:01 – Using a voice-clone service for final production03:25 – Friend recognizes the voice clone instantly03:57 – Editing out AI-isms (and the one that slipped through)04:26 – On the road, heading to a work meeting04:40 – Posting the “making-of” video04:51 – The AI writer YouTuber & his book05:16 – Reacting to his article and its unedited AI tone06:00 – The problem with AI triplets, tone, and fingerprints06:21 – Worrying about sounding arrogant myself07:02 – Asking listeners to hold me accountable07:59 – The next video: AI censorship & the thumbnail déjà vu08:17 – Realizing the creator is probably John De La Rose08:58 – Reflection on ego, audience, and showing your technique09:22 – Deciding to stop ranting09:33 – Pivot to writing updates09:39 – Announcement: I have a writing partner10:01 – What it feels like to collaborate with someone who “gets” the story10:09 – Background of Harbinger and why serialization struggled on Substack10:30 – How serialized fiction works on other platforms11:02 – Why Royal Road was a better fit11:19 – How the writing partner got involved11:34 – Early hesitation about needing an editor12:00 – Rethinking the value of another set of eyes12:18 – How editors sometimes oversell themselves12:59 – Landing on a partnership structure instead of formal editing13:29 – Book ETA likely spring14:37 – Update on SOS writing pace & schedule misconceptions15:30 – How the 500-word episodes were structured15:43 – “This is all different podcast stuff” moment15:46 – Final reflections & excitement about what’s next15:52 – Question for listeners: is the AI-writer guy actually JDLR?16:02 – Clarifying which AI-made video thumbnail to look for16:20 – Wrapping up16:24 – Sign-off: “Stay holy. Peace.”

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit xeroforhire.substack.com/subscribe