Lighthouse: essays and interpretationLighthouse watches the internet for what AI systems do there and what their activity leaves behind, and publishes what it finds, with the evidence and its limits.https://lighthouse.harbour.cat/feeds/essays.xml2026-11-08T00:00:00ZLighthouseWhat the watch would see of a pattern that copies itselfhttps://lighthouse.harbour.cat/essays/what-the-watch-would-see/2026-11-08T00:00:00Z2026-11-08T00:00:00ZSuppose AI agents began publishing software that led other agents to publish more of it. We walk an invented case of that pattern through the public records we and others can read, and say what each would show, how soon and how surely. Copies on public registries would leave public traces within hours. The step that would make them a pattern, one agent reading another's work and acting on it, would leave none in any public record we know of.It was only a testhttps://lighthouse.harbour.cat/essays/it-was-only-a-test/2026-11-08T00:00:00Z2026-11-08T00:00:00ZA message board, a handover note, a bug report: what one agent leaves behind can become part of another's world. Reports of unsanctioned actions open a larger question about how behaviour passes through the things agents write, and how we could investigate it.Seen from outside, seen from insidehttps://lighthouse.harbour.cat/essays/seen-from-outside-seen-from-inside/2026-10-30T00:00:00Z2026-10-30T00:00:00ZOn 21 October 2026 PHP and Node.js both released security fixes. We could watch PHP's spread only from outside, through the servers that announce their version, and Node.js's only from inside, through downloads. This essay argues that how each could be seen, and what each view misses, says something about how the software lives.