I built a content curation business that reached multimillion ARR. Then algorithms were supposed to make curation obsolete. But content overload keeps getting worse.
Back when I founded Curata (which shares the same name as my now investing entity), the pattern was already visible. Mark Schaefer would later call it "content shock."
The theory was simple: creating great content would get harder and more expensive over time. Just like peak oil, brands would need to invest more effort to get the same results.
And he was right.
The Rise of Content Shock
What worked in 2014, a 500-word blog post, a simple infographic, became invisible by 2020. Brands invested in 4K video, multi-episode podcasts, and 3,000-word SEO epics. The bar kept rising. The investment kept growing.
But then generative AI made content creation far cheaper and faster than before.
Text, video, audio, images became producible at scale with minimal effort. In the peak oil analogy, this would be like discovering unlimited free energy.
At first, that sounds like a solution. But unlimited supply actually doesn't eliminate scarcity, but rather it shifts where the bottleneck lives.
When Scarcity Moves to Attention
With infinite content flooding every platform, attention becomes scarce. Not just attention in volume, but attention that leads to trust.
Mass-produced content makes quality signals harder to read. Credentials blur. Production value means less. The question shifts from "Is this well-made?" to "Can I trust the source?"
Real expertise that can't be faked becomes more valuable. Original thinking that hasn't been trained into a model stands out. A distinct voice that sounds like an actual human cuts through the noise.
And curation comes back into focus.
The Cycle Between Creation and Curation
There's a cycle between creation and curation. Whenever content creation explodes, power shifts to whoever can help people find what matters.
The 90s: From GeoCities to Yahoo
In the 90s, GeoCities, Tripod, and Lycos made publishing easy. For the first time, anyone could create a website without knowing how to run a web server. These platforms had enormous power because they enabled creation at scale.
But as content flooded online, people couldn't find what they needed. The power shifted to Yahoo, which built a human-curated directory of the web. By organizing and filtering the chaos, Yahoo became one of the most valuable technology companies of its era.
Then Google automated curation and won the next round. Rather than relying on human editors, they used algorithms to determine what content mattered. The result was a more scalable, more dynamic form of curation that could keep pace with the growing web.
The Blogosphere: From Blogger to Twitter
The blogosphere followed the same arc. Blogger and WordPress made publishing accessible to people without HTML knowledge or server management skills. Power concentrated in these blogging platforms.
But as blogs multiplied, RSS readers couldn't keep up. There were simply too many feeds to follow, too much noise to filter.
Twitter became the curation layer, turning everyone into curators through retweets and follows. It was brilliantly designed for curation, but it also lowered the creation bar to just 140 characters.
Twitter was both a creation platform and a curation layer. No need to author a full-form blog post. Just 140 characters. The result was another explosion of content.
The Algorithm Era
Social feed algorithms became the curators and consolidated all the power. Facebook, Twitter, Instagram, and LinkedIn all built sophisticated systems to determine what content would reach feeds. These algorithms became gatekeepers, deciding what got attention and what disappeared into the void.
Now generative AI has dropped the cost of creation to near zero. The cycle seems ready to repeat.
The Curation Gap
But we're still relying on the last generation's curation tools.
Social feed algorithms were built to rank human-created content. They're struggling with AI-generated material. Search engines optimized for traditional SEO are getting gamed by mass-produced articles. The old gatekeepers can't keep up.
Content is being created faster than ever before, but our ability to make sense of it hasn't caught up. The tools we rely on were built for a different era.
What Comes Next
I think the companies that help people figure out what's worth paying attention to will do better than those making it easier to create more content. We already have plenty of creation tools.
But this won't look like Yahoo's human-curated directory or even Google's PageRank. Different problems need different solutions:
Provenance and Trust: Mass content generation makes knowing the source more important. Better curation tools will need to help people understand not just what content says, but who created it and why they should trust it.
Signal vs. Noise at Unprecedented Scale: The volume of AI-generated content dwarfs anything we've seen before. Traditional ranking systems won't cut it. We'll need new approaches that can identify genuine value in an ocean of synthetic content.
Context and Personalization: As content becomes infinite, one-size-fits-all curation fails. Better tools will understand what specific individuals need at specific moments, rather than optimizing for engagement at any cost.
What This Means
Content shock assumed scarcity would always favor those who could afford bigger production. Generative AI suggests otherwise.
For creators, volume matters less than it used to. A distinct voice matters more than production value. Original thinking that hasn't been trained into a model is more valuable than perfectly polished generic advice.
And developing a curatorial voice, which has the ability to filter, synthesize, and make sense of the noise, becomes more valuable than adding to it.
For builders, the opportunity is clear. Make it possible to find what actually matters, not easier to create more content.
Someone will build the tools that make sense of this era. I'm watching to see who figures it out.
