I've been thinking a lot lately about how much of what we read online is just filtered through some algorithm's idea of what we should see — reminds me of trying to sort through Perplexica results and wondering what's actually true versus what's just... popular.

What's your gut feeling — do you think better local tools (like running ollama locally) actually help with this propaganda filtering problem, or is it more about who controls the training data in the first place?