[Wump]

> spent his time on backups instead of solving millennium problems

The fundamental issue with this comment is that it assumes that solving millennium problems is obviously a high impact thing to do. It ignores the complexity and unpredictability of the world.

The TarSnap developer's story is a great example of this -- how "working on backups" led to impactful work that he probably didn't predict when he started it.

I often see people fall into the trap of moralizing about what "good" work is. Ironically, that prevents them from ultimately having the most impact that they could have.