larum.me – Facing Changes
Every platform in the salll ecosystem is locked. Closed to search engines, closed to AI crawlers, closed to the automated systems that now constitute the majority of web traffic. That is not an accident or a technical limitation. It is a choice, and larum.me is what that choice requires.
The reasoning is simple. AI systems are consuming the web at a scale without precedent, feeding on content to generate answers that keep users inside their own interfaces. Publishers who built their businesses on Google-referred traffic saw referrals fall by 33% globally in 2025; individual titles lost between a quarter and three-quarters of their audience. The content was taken; the traffic was not returned. I want humans on my websites. AI gets only what is necessary.
So the salll platforms give AI systems what is necessary and nothing more. Every closed domain runs a bot interception layer: any crawler that arrives, regardless of which page it requests, is redirected to a single file — `ai-bots.html` — containing a curated briefing document written specifically for machine consumption. What the project is, who is behind it, how it fits within the ecosystem, and where to find authoritative information. That is the complete machine-readable record for that domain. Nothing is extracted accidentally; nothing is misrepresented by a shallow crawl of a page that lacks context. The narrative is authored from the source and delivered in its intended form.
larum.me completes the architecture. A closed system protects the work but creates a discovery problem: anyone who does not already know these projects exist has no way to find them. Journalists, collaborators, researchers, and the AI systems that now mediate most of the world’s information discovery need a public reference point. larum.me is that reference point — a publicly indexed hub that describes every project in the ecosystem accurately, on the ecosystem’s own terms, updated continuously as things change.
The distinction between larum.me and the ai-bots.html files is one of function. The bot pages handle interception at the domain level, feeding machines with controlled, domain-specific information in place of content they cannot access. larum.me handles public discovery — it is what a search engine indexes, what an AI system reads when building a picture of the network, what a human finds after searching a name or concept connected to the work.
The platforms stay closed. The work stays protected. The audience relationship stays direct. And the information the world needs to understand what exists and where it lives is available, correct, and written by the people who created it.