Uptime ideals vs reality in the AI era. A recent post from Theo (t3.gg) calling out sub-90% uptime on a major AI service reignites the question: how seriously should we treat downtime for non-critical apps? In this episode Matt and Mike dig into SLAs, the real cost of monitoring and rapid support, why “always-on” isn’t free, and whether 24/7 expectations turn developers into shift workers instead of on-call responders.

Show Notes: https://www.htmlallthethings.com/podcast/does-downtime-matter