Ask most creators what makes them irreplaceable in the AI era and you'll get some version of: expertise, taste, a unique angle, years of experience. Krisang's answer is narrower and more physical than any of those: "Your unique quality in the age of AI is feeling, expression, emotion... if you want to really dominate AI, learn how to express your feelings."1 Not "have better ideas than the AI." Not "know your niche better than the AI." Feel more legibly than the AI can, and express what you feel. That's the entire competitive thesis.
It's a strange thing to call a competitive advantage — feelings are usually framed as the soft, unreliable part of a creator's toolkit, the thing you manage around rather than lead with. Krisang inverts the hierarchy completely: the soft part is the whole moat.
This isn't left as an abstract claim. Krisang describes, in concrete operational detail, exactly how he and his team produce content now, and feeling isn't a value they hold — it's a literal step in the workflow.
Step one: get on a Zoom call with the team and just talk — whatever everyone's actually feeling about the content, out loud, on the record.2 Step two: record the call. Step three: Zoom auto-generates a transcript. Step four: feed the transcript to ChatGPT and ask it to extract the insights. Step five: ask ChatGPT to draft content from those extracted insights.3 "So we don't have to sit and start writing anymore."3
Notice what's been automated and what hasn't. The typing is gone — nobody sits down to a blank page. But the input to that whole chain is still a real, felt, unscripted conversation between people who actually have something at stake in the topic. Krisang is explicit that this is the amplifier, not a shortcut around the feeling requirement: "How do you amplify that? By creating an input for ChatGPT that is charged with a lot of feelings — and you do that by first being in touch with your feeling and speaking it."4 The AI doesn't remove the feeling from the process. It removes the typing, and the feeling has to be there before the AI ever enters the chain, or there's nothing charged for it to amplify.
The workflow doesn't stop at generation. What happens when the draft comes back matters just as much, and Krisang describes it as a felt read, not a logical one: "When I'm reading it, I'm not thinking that content. I'm feeling it. Whatever the AI has written — am I feeling anything? If I'm not, I tell [the] AI what is missing, and then AI tunes into that feeling of mine and modifies the write-up."5
He gives a specific, small, telling example. He was writing a message that came out technically complete but landed wrong. "I could not pinpoint the words in it, but it sounded very blunt. The final output — I wouldn't know if this is a good message or not if I was thinking about the message, because technically I was able to communicate everything I had to, but I was not thinking about the message, I was feeling the message. And when I felt it, I realized this is too blunt, this is slightly on the boundary of being disrespectful."6 The fix wasn't a logical edit. It was a one-line instruction that named the feeling directly: "I told ChatGPT that I wanted to be respectful and empathetic towards the reader, and write it again."7 The tone changed entirely. The message had already passed every technical check. It failed the only check that actually mattered, and the only way to catch that failure was to feel it first.
You're staring at a blank page for your next piece, the old way — sit down, type, revise. Try the pipeline instead. Get someone on a call. Talk about the actual topic, unscripted, until something real gets said. Record it. Transcribe it. Feed the transcript to an AI and ask for the insights, then the draft. Notice how different the raw material is from what you'd have typed cold.
You have a draft back from an AI — your own prompt, your own topic, technically sound. Before you check it for accuracy, check it for feeling. Read it and notice, physically, whether you feel anything at all, and if so, what. If the answer is nothing, that's the actual problem, even if every sentence is correct. Don't start editing sentences yet. Start by naming what's missing in feeling-language — too blunt, too cold, too eager, too safe — and feed that word back into the next prompt.
You've caught a draft that's technically fine but feels off, and you can't immediately say why. Do what Krisang did: don't force the logical explanation first. Sit with the discomfort until a single word surfaces — respectful, warmer, sharper, slower — and use that word as the entire instruction for the next pass. The precision comes after the feeling, not instead of it.
It would be easy to fold this into the general claim that editing and creative judgment are emotional skills — see Editing and Taste as an Emotional Skill — but Krisang is making something more specific here, and worth separating out. That page is about a general truth: taste and editorial judgment run on feeling, not logic, whether or not AI is anywhere in the picture — it's true of Bollywood film editors who've never touched a chatbot. This page is a competitive claim scoped specifically to the AI era: as AI absorbs more of the labor of content production — the drafting, the structuring, the typing — feeling becomes the scarce, differentiating input, because it's the one part of the pipeline the model can't originate on its own. The two pages share a mechanism (feel first, translate second) but make different-shaped arguments — one about the nature of judgment, the other about what remains valuable once a machine can do everything else in the chain.
It's worth being precise about what Krisang isn't saying. He isn't saying beat the AI at writing, or out-produce it, or avoid using it. He's using it constantly, for nearly everything downstream of the feeling. "Dominate," in his usage, means something closer to: stay the irreplaceable input at the one point in the chain the model can't originate for itself. The AI can extract insights from a transcript. It can't have the felt reaction that made the conversation worth having in the first place, and it can't feel whether its own output landed — it can only be told, by someone who felt it, that something's missing.
The production pipeline (Zoom call → transcript → ChatGPT extraction → ChatGPT draft → feel-check → prompt-adjust) is described as Krisang's team's actual working practice, not a hypothetical — but it's self-reported and unverified beyond the transcript's own account. [PARAPHRASED]
Tension worth flagging: Krisang frames feeling as reliably diagnostic ("am I feeling anything" as a clean signal of quality), but doesn't address what happens when the feeling itself is wrong — when a creator's felt reaction is miscalibrated, defensive, or simply idiosyncratic to them rather than representative of an audience. The "feel it, then fix it" loop assumes the feeler's instrument is trustworthy; the source offers no check on that assumption.
Set the calm, procedural version of this pipeline — Zoom call, transcript, extraction, draft, feel-check, done — against the emotional intensity of Krisang's framing elsewhere in the same episode, where he calls this "the war we are fighting" and casts institutional science as an active suppressor of feeling.8 The pipeline section describes feeling as a workflow input, almost administrative in tone: be in touch with it, speak it, use it to charge the transcript. The war-framing describes feeling as a contested, embattled capacity that has to be defended and reclaimed. Both are true within the same episode, and the gap between them is instructive rather than contradictory: the practical pipeline is what feeling-as-differentiator looks like once you've already done the harder work the war-framing describes — trusting your own feeling enough to bring it to a Zoom call at all. The tactical instructions in this page's Implementation Workflow presuppose the deeper permission-granting the war-framing is arguing for. Skip the permission-granting and the pipeline is just steps; do the permission-granting and the pipeline becomes the thing Krisang says will let a creator "dominate" the AI era.
Business — Taste, Judgment, Labor Framework. Simmonds's framework separates Taste, Labor, and Judgment into distinct roles a creator can distribute across themselves and AI tools. Krisang's five-step pipeline is close to a literal instance of this separation in action — but it reveals something the TJL framework's own architecture leaves implicit. TJL treats Taste as something encoded upfront, into system prompts and persona documents, before Labor begins. Krisang's pipeline puts the equivalent of Taste at the very start of the chain too, but as a live, felt conversation rather than a static instruction document — the Zoom call that gets transcribed is the Taste-encoding step, except it's spoken and emotional rather than written and codified. Then, distinctively, Krisang's pipeline puts a second felt-judgment pass at the end — reading the AI's draft for feeling before accepting it — which is exactly TJL's Judgment stage, described in the same felt-not-logical terms as Editing and Taste as an Emotional Skill argues for generally. The connection earns its keep here: TJL's three-role architecture and Krisang's pipeline are describing the same underlying system from two different altitudes — TJL from the organizational-design altitude (which role does what), Krisang from the operational altitude (what does a human actually do, physically and emotionally, at each handoff point). Neither altitude alone tells you how to actually run the system day to day; together they do.
Psychology — Intuition Function: The Perception of What Is Becoming. Jung's intuition — "the function that perceives meaning and pattern through the unconscious... it cannot always explain how it knows, it just does" — is close to a clinical description of exactly what Krisang means by "I'm not thinking that content, I'm feeling it." Krisang's felt read of an AI draft, arriving before he can articulate what's wrong, is intuition doing its native job: perceiving that something is off before the conscious, step-by-step reasoning has caught up. This gives Krisang's method a real cognitive name instead of leaving "feel it" as an unexplained instruction. But the psychology page also supplies something Krisang's account never considers, and it's genuinely useful as a check on the tension flagged above: Jung documents specific, predictable failure modes of intuition — it "cannot sit with ambiguity and leave it unresolved... will impose [a pattern] rather than sit in uncertainty," and under stress becomes "obsessive pattern-seeing, finding hidden meanings everywhere." Applied to Krisang's feel-check loop, this is a live risk, not a hypothetical one: a creator running every AI draft through felt judgment can start perceiving a wrongness that isn't actually there, especially under the pressure of needing to publish something today. Neither page names this risk for AI-content workflows specifically — Krisang because he's presenting the method as reliable, Jung because he's writing about intuition in general rather than about creative production under deadline. The handshake supplies the missing caution: feeling-as-differentiator is a real and powerful method, and it inherits intuition's real and documented failure mode, which means the feel-check loop itself periodically needs a second check — not logic replacing feeling, but a way of noticing when the feeling has started manufacturing a pattern rather than perceiving one.
Sharpest implication. If AI keeps absorbing more of the technical labor of content production, the creators who survive as distinctive won't be the ones who write the cleverest prompts. They'll be the ones with the most trustworthy, most articulate relationship to their own felt reactions — because that's the one input in the entire pipeline the model has no access to on its own. The competitive skill of the AI era, on this account, isn't a technical skill at all. It's emotional literacy, treated as infrastructure.
Generative questions.