Zero Marginal Cost of Excellence and The Median Creator
When speed and quality are part of the same equation
This is my hypothesis on how a median creator will need to evolve when AI creates pixel-perfect video, broadcast-grade audio, and print-ready prose.
The old content equation offered a single degree of freedom: choose speed or high-quality, never both. Generative AI erases that choice. Ideating, drafting, fact-checking, layout, even derivative design now clear in minutes, soon at a doctoral researcher’s quality, pushing marginal cost of excellence toward zero. Don’t agree? Remember the Ghiblification trend?
Abundance at the production layer triggers scarcity further downstream. Because the human attention is finite.
Feeds become high-definition torrents where every post is technically flawless. Newsletter inboxes saturate: model-written newsletters stack a day’s worth of polished prose by lunchtime.
This accelerates power-law stratification. Incumbents with large followings—Lenny Rachitsky, Ben Thompson—and creators who are already fluent in AI workflows—Varun Mayya for instance—collect an even bigger share of impressions. Their content feeds the loop; the loop widens their lead. For median creators, organic reach trends toward statistical noise.
When the open web no longer delivers reliable discovery, creators must migrate to terrain they can control. A community is no longer a growth hack; it is the last moat for median creators cut off from algorithmic reach.
A tight-knit community secures attention through three assets model can’t replace:
Presence Members recognise that a real person will challenge ideas, answer follow-ups, and remember past contributions.
Facilitation The creator brokers introductions, frames debates, and turns passive readers into collaborators—something automated feeds cannot orchestrate.
Peer-to-Peer Engagement Members come for the curator and stay for the members.
There’s a catch, though.
Because a creator’s time is finite, viable communities will be narrow for genuine exchange yet large enough to sustain paid access. Sustainable numbers may cluster around a Dunbar-like core of ~150 active contributors and perhaps 1500 lurkers. Retention holds when you run live cohorts, rotate facilitators, and layer tiered access so members graduate through trust levels.
I would expect nodes like “bootstrapped AI tools for indie game devs,” “carbon-accounting CFOs in climate fintech,” or “Bangalore D2C operations leads.” Creators will gatekeep these spaces and monetise them through content and curation—using the same AI tools that displaced them elsewhere.