[dead]
My guess is whatever system downdetector uses to "detect downtime" relies on either GCP or Cloudflare (also having issues at the moment: https://www.cloudflarestatus.com/)
This is where we get to learn about the one common system all of our “distributed cloud” systems rely on, isn’t it?
But I suppose there's always "something something BGP" but that feels less likely.
I could see a big outage like this causing people to google "Is AWS down?"