As AI-generated content supply grows, is content surfacing/distribution or raw content supply the binding constraint on what audiences see?

Harry Stebbings · published Sep 18, 2023

As AI makes content creation near-infinite, content discovery becomes the core problem for media

One long video can become a hundred short ones, expanding competition a hundredfold

58:39 20VC: Benchmark General Partner, Miles Grimshaw on The Five Pillars of Venture Capital, Why Data Can Be a Trap When Early-Stage Investing, Investing Lessons from Missing Figma and Plaid & The New Business Model for AI & Why Co-Pilot is an Incumbent Strate

Ethan Mollick · published Jul 31, 2024

The content-glut problem has already happened — attention is already scattered across a huge amount of content, before AI generation scales

Being a New York Times bestseller only takes around 6,000 hardcover copies in a week, showing how thin and dispersed audiences already are

59:54 20VC: Is More Compute the Answer to Model Performance | Why OpenAI Abandons Products, The Biggest Opportunities They Have Not Taken & Analysing Their Race for AGI | What Companies, AI Labs and Startups Get Wrong About AI with Ethan Mollick

Victor Riparbelli · published Jan 15, 2025

In five years there will be far more AI-made than human-made content, but the more important question is which content actually gets surfaced

Content abundance already exists — most YouTube videos get under ten views — so distribution and surfacing, not supply, is the binding constraint

Scope: the YouTube view statistic is offered as an approximation

41:47 20VC: Why Scaling Laws Will Not Continue | OpenAI vs Anthropic vs X.ai: Who Wins and Why | How Far Will Model Providers Go Into the Application Layer | The End State for Models: Many Specialised or Few Generalised with Victor Riparbelli @ Synthesia

Harry Stebbings · published May 22, 2024 · hedged

Because generative AI makes supply of content and outreach effectively unlimited, the value of that output may go to zero

An AI outbound rep can send a million messages a day, and podcast art or even podcast voices can be generated at will, so recipients get a thousand identical AI emails

Scope: stated as a personal worry

42:06 20VC: Box's Aaron Levie on Predictions for the Next Wave of AI: Will Foundation Models Be Commoditised | How the Business Model of SaaS Changes Forever | Startups vs Incumbents: Who Wins | App vs Infrastructure Layer: Where is the Value?

Harry Stebbings · published Jul 31, 2024

With an infinite supply of AI-generated content, the value of content falls and discovery becomes much harder — this is the main risk for creators

Infinite supply undermines both the price and the discoverability of any individual piece of content

Scope: Stated from his position as a content creator

59:41 20VC: Is More Compute the Answer to Model Performance | Why OpenAI Abandons Products, The Biggest Opportunities They Have Not Taken & Analysing Their Race for AGI | What Companies, AI Labs and Startups Get Wrong About AI with Ethan Mollick

Harry Stebbings · published Sep 12, 2026 · hedged

In a world of infinite AI-generated content supply, discovery becomes impossible and users get flooded

Scope: framed as a personal worry

18:57 20VC: 7 Predictions for How AI Changes the World: Labour, Engineering, Social Media, GrokBots Buying Cybercabs and more with Matteo Franceschetti, Co-Founder @ Eight Sleep

Aaron Levie · published May 22, 2024

Unlimited AI-generated supply will not drive value to zero; a power-law dynamic and creative destruction will sort winners from losers

The same objection could have been raised about tens of millions of iPhone apps twenty years ago; the world figures it out — things that don't work die off, people stop paying for outbound emails that fail, and what works takes off

42:38 Power law and creative destruction sort winners so value doesnt go to zero despite unlimited supply

Matteo Franceschetti · published Sep 12, 2026 · hedged

Even with infinite content, the best product still wins because personal agents will search out and recommend it — but companies will need two distinct strategies, one to reach humans and one to be found by agents

The volume of content will be too much for humans to process, so agents become the filter

19:05 Best product still wins via agent mediated discovery requiring dual human and agent strategies