[starfallg]

>Evaluation must be separate from dissemination.

Evaluation is inevitably linked to dissemination. We cannot evaluate what we have not received, yet we are now receiving too much to effectively evaluate.

And then there is the matter of bias - once an idea has been repeated enough times it takes hold, regardless of rationality.

>the dissemination of deliberate misinformation and lies, and the difficulty and _high cost_ of navigating this ecosystem.

Isn't this distinction arbitrary? To an observer, the intention of the messenger is opaque. Deliberate lies and innocent misinformation arrives together for us to evaluate. Dealing with it at the source introduces a variety of problems in our current model, and judging intent with regards to speech is problematic.