Readership
The honest answer to “who reads a personal site in 2026” is: mostly machines. This page counts fetches by known AI crawlers, because when someone asks a model about me, what those crawlers read is what it answers from.
Distribution list — who receives copies of this document
| recipient | copies | reads most | first issued | last issued |
|---|---|---|---|---|
| Meta-ExternalAgent | 98 | /notes /patents /readership | 2026-07-18 | 2026-09-07 |
| GPTBot | 41 | / other /verified | 2026-06-12 | 2026-09-26 |
| ClaudeBot | 31 | /patents other / | 2026-07-11 | 2026-09-26 |
| Applebot | 13 | /notes / /verified | 2026-06-30 | 2026-09-03 |
| OAI-SearchBot | 11 | other / /readership | 2026-07-11 | 2026-08-16 |
| Amazonbot | 11 | other / | 2026-07-18 | 2026-08-16 |
| ChatGPT-User | 9 | other / /resume | 2026-07-11 | 2026-08-16 |
| CCBot | 5 | /readership /chat /verified | 2026-07-19 | 2026-07-26 |
| Bingbot | 4 | other /notes /verified | 2026-07-11 | 2026-07-26 |
| Claude-User | 3 | other / | 2026-07-18 | 2026-08-29 |
| PerplexityBot | 3 | other /chat | 2026-07-11 | 2026-07-19 |
| Googlebot | 2 | /patents other | 2026-07-11 | 2026-07-18 |
| Claude-SearchBot | 2 | other | 2026-07-18 | 2026-07-18 |
Method: edge middleware matches the request’s user-agent against a fixed list of known AI crawlers and increments a counter per crawler and page. A user-agent string is a claim, not proof — anyone can send one. No cookies, no fingerprinting, and humans are not counted. Humans are also welcome.