I don't want to be a content creator for llms

I think the harder question is whether people stop visiting because of LLMs, or because most content was interchangeable to begin with.

If someone publishes original research, a unique workflow, or a product people can actually use, I still think they'll come to the source.

I'm building an AI developer tool myself, and it's made me think less about "how do I stop AI from reading my content?" and more about "what value can only the original source provide?"

That is such a valid and timely concern. It feels like the "social contract" of the internet—where you provide value and get traffic/community in return—is being replaced by a one-way extraction model.
The Tailwind CSS example is a perfect case of "successful" documentation inadvertently making the source redundant, which is a scary prospect for any creator. It’s a tough spot to be in: wanting to be generous with humans while being protective against scrapers.
Have you looked into things like the GPTBot blocking protocols or the Spawning.ai "No AI" tools? They aren't perfect, but it's a growing movement for creators who want to reclaim their work.

I get the frustration, but I think there's a middle ground here. LLMs are a tool , they can scrape and summarize content, sure, but they don't replace the actual creation and curation part.

If someone's putting content behind a login to avoid being training data, that's totally fair.

But the issue isn't really the LLMs themselves, it's how they're being used and whether creators get compensated or credited. I'd rather see better attribution systems than just locking everything down.

The Tailwind example is interesting because it highlights a specific business model vulnerability: documentation-as-moat doesn't work anymore when LLMs can synthesize and serve that knowledge instantly.

But I think there's a nuance worth exploring. LLMs didn't kill the value of content — they killed the value of static, reference-style content. What still works:

• Opinions and perspectives — LLMs can summarize facts, but they can't replace a human's take on why something matters
• Real-time context — "Here's what I learned shipping X this week" has a shelf life that training data can't capture
• Community and interaction — The comments here are more valuable than the post alone because they're a live conversation

The irony is that putting content behind a login might protect it from crawlers, but it also removes it from the human-to-human discovery loop that makes content spread in the first place.

Maybe the answer isn't gating content, but shifting what kind of content we create. Less "how to do X" (LLMs own that now), more "here's my experience doing X and what surprised me."

I think the real shift is that content alone no longer has scarcity, context, intent, and interaction do.

Static explanations will get absorbed by LLMs. What doesn’t: live thinking, opinionated workflows, tools, and conversations that evolve.

Sharing “with humans” might mean designing content that expects participation rather than consumption otherwise machines will always win at mirroring it.