“Arguing that you don’t care about the right to privacy because you have nothing to hide is no different than saying you don’t care about free speech because you have nothing to say”

Snowden

Both could be filtered out in theory (which they should do if they were smart, because amount of training data matters way less than quality or training data), but filtering AI created slop is harder, especially if you just slightly modify it.

<5min Germany

Logseq, it’s a lot like Obsidian as it also has knowledge graphs, tags, is markdown-based and self-hostable but, in contrast to Obsidian, it’s fully open source