here’s the deal:
HN is a tiny forum which basically runs on a single machine with a single backup. There are surely only a few people monitoring it. It’s used by about ~1 million people a day, and nobody really “relies” on HN.
GitHub, Facebook, Google, AWS etc. are massive clusters of over 1000 servers spread across continents, with multiple failure modes and run by many DevOps. Many people and services rely on them being up. They’re used by billions each day.
Yet HN has more uptime than these behemoths, and way more uptime than e.g. Slack. It’s also faster, even though it’s only one server. HN went down for 2 (?) hours. HN also promptly posts why it went down. In this case because both the main and backup server failed.
Of course HN is definitely more reliable because it’s only one server. But it’s still impressive and shows that you can reach a large audience with a simple setup and few resources. Maybe if Google, Facebook, AWS, etc. found more straightforward solutions they too could have more uptime.