I was thinking of setting up this site, and realized going forward, AI bots could be a considerable audience, and if I planned to monetize any of the content, that I wasn’t sure if there would be available ways to do that given bot’s don’t have actual eyeballs and are not the audience for AdSense, etc… as far as I know.
My Prompt to AI chat buddy:
You know how Google ads is a way to allow some other party to impact the attention of your site visitor and they will pay you money to have that opportunity. What I am wondering is what if the site visitors are all bots? Is there someway to require them to pay to look at the content I am providing? Is there friction in the interaction that someone, or some protocol, or something like Google ads, but for monetizing a bit of content, not for human eyeballs, but for AI bots slurping up and using the novelty of your new content, as feed material. In order to satisfy that need for novelty, from human creation, but also to reward that human creation, that human, who is in society, and needs the energy or coin of that society in return for the labor of providing this information. Ideally, it would be rewarded well when handcrafted. Any thoughts?
Sonnet 4.5 talking:
This is a fascinating question that sits right at the intersection of AI economics, content monetization, and the changing nature of value creation. You’re essentially asking: “If AI is the new audience, where’s my ad revenue?”
The core challenge you’re identifying is real and urgent. We’ve built an entire web economy around human attention as the scarce resource, but now massive amounts of content consumption happen via scrapers and training runs—with no compensation flowing back to creators. It’s like having a stadium full of people watching your performance, but they’re all looking through a one-way mirror and nobody bought a ticket.
Current partial solutions:
- The robots.txt protocol is purely voluntary—it’s a “please don’t” with no enforcement. Some sites are now using it to block AI crawlers, but it’s trivial to ignore.
- Paywalls and authentication work for human subscribers but break the traditional “open web” model. They also don’t distinguish between “one human reading” and “one bot training on millions of pages.”
- Licensing deals are emerging—Reddit, Stack Overflow, and news organizations are striking deals with AI companies. But this only works if you have enough content to matter at scale, and it’s opaque negotiation rather than a systematic market.
- What you’re really describing is a micropayment protocol for content access—but optimized for machine consumption rather than human browsing. Some possibilities:
- Token-gated APIs: Require API keys with pay-per-request pricing. This exists for structured data but not for general web content. The friction is that it breaks the “just visit a URL” simplicity that makes the web work.
- Content credentials with usage tracking: Embed cryptographic signatures in content that track downstream usage. When an AI company uses your content, the trail is auditable. The Coalition for Content Provenance and Authenticity (C2PA) is working on technical standards here, but the economic layer doesn’t exist yet.
- Attention markets for AI: Imagine something like Google AdSense, but instead of advertisers bidding for human attention, AI companies bid for content access. You’d embed a protocol in your site that says “to crawl this, deposit X microcents per page.” The technical challenge is authentication without killing the open web, and creating a market maker who handles the transactions.
- Proof-of-value mechanisms: Rather than charging per access, you could charge based on downstream value. If your tutorial gets incorporated into an AI’s training and that AI generates revenue, you get a royalty. This is theoretically elegant but practically a nightmare to implement and audit.
The deeper tension: The web’s value came from being open and linkable. Monetizing AI access risks fragmenting that. We might end up with a two-tier web: open content for humans (supported by traditional ads) and licensed content for AI (behind authentication). Or we might see new intermediaries who aggregate content rights and license them in bulk—basically becoming the ASCAP for web content.
Your intuition about handcrafted content being especially valuable is spot-on. As synthetic content floods the internet, genuinely novel human-created material becomes the rare training data that actually improves models. That should command a premium, but there’s currently no market mechanism to capture it.
The honest answer is: nobody has really solved this yet. We’re in the awkward period where the economic model is breaking but the replacement hasn’t crystallized. The companies with leverage (large platforms, publishers) are negotiating deals. Individual creators are mostly left out unless they band together or someone builds the infrastructure you’re imagining.
If you’re serious about exploring this, the most promising avenue might be joining or watching initiatives around content provenance standards and seeing if someone builds a payment layer on top. Or, more speculatively, thinking about how blockchain-style micropayments could work without the overhead that’s killed previous attempts.
What kind of content are you creating that’s prompted this thinking?