Overview of It's come to this: Human certification in the age of AI slop
This NPR Indicator episode explores a new attempt to solve a growing problem in the age of generative AI: how to prove a creative work was made by a human. The story centers on novelist and standards expert Ned Hayes, who launched Human Intelligence, a certification system meant to function like an “organic” or “fair trade” label for art—signaling that a book, song, or other work is genuinely human-made.
The Core Idea: A “Seal of Humanness”
Ned Hayes’ inspiration came after a reader told him that a chatbot could probably write a novel as good as his. He saw the issue as bigger than just output quality: art matters because of the human process behind it.
What Human Intelligence tries to do
- Create a standardized human-authorship certification
- Give artists a recognizable H logo to display on verified work
- Help audiences and buyers distinguish human-created art from AI-generated content
- Potentially allow verified human work to command a higher price
How the Verification Works
Unlike basic AI detectors that look for stylistic clues, Human Intelligence tries to verify the creative trail itself.
The process includes:
- Identity verification
- A sworn attestation, under penalty of perjury, that the creator is human
- Submission of drafts, raw files, and work history
- Review of metadata from documents like Word or Google Docs
- Analysis of revision patterns to see whether the work shows signs of real human drafting and editing
For musicians, filmmakers, game designers, and writers, the company’s message is essentially: show us the drafts.
Why Hayes Thinks This Matters
Hayes’ background in internet standards shaped his thinking. He helped work on XML in the early web era, and he believes the AI era needs a similar common language or standard.
His larger concern
- AI-generated art may flood the market with “AI slop”
- Consumers may lose trust in creative work
- Artists could be forced to compete with mass-produced content from a small number of tech companies
- Human creators need a way to assert control over how their work is labeled and valued
Artists’ Reaction and the Real-World Test
At first, many artists dismissed the problem. But as AI-generated content became more visible, interest grew.
One notable early supporter
- Corin Tucker of Sleater-Kinney
- She agreed that artists need agency in the AI conversation
- She completed the certification process and received the human-verification logo
- For future releases, she could submit proofs like handwritten lyrics, early demos, and final mixes
Her support framed the issue as not just technical, but cultural and political: if artists don’t have a say, control may shift to Silicon Valley companies.
The Business Model and Limits
Human Intelligence charges artists a fee to verify works and maintain its database.
Key details
- Current fee: $20
- The service is still small and early
- It launched with only a handful of artists, but interest has grown
- The episode notes that other groups, like the Authors Guild, are also creating their own human-authored labels
Unresolved question
It’s unclear whether there will be one universal human-authorship standard or several competing certification systems.
The Episode’s Practical Experiment
The Indicator hosts decided to test the system themselves by submitting raw audio, drafts, and documents from their show.
What happened
- They sent in interviews and Google Docs
- Human Intelligence reviewed their revision history and materials
- After about a week, they passed the verification
The test reinforced the episode’s broader point: for now, human certification is less about perfect detection and more about creating trust, transparency, and a market signal.
Main Takeaways
- AI content is making it harder to know what’s human-made.
- New certification systems are emerging to label human-authored creative work.
- Human Intelligence uses drafts, metadata, and identity checks rather than simple AI detection.
- The system aims to give artists a trusted label and potentially a financial edge.
- The long-term future may involve multiple competing standards for human-made art.
- The cultural stakes are high: this is not just about technology, but about authorship, trust, and artistic value.