I was thinking exactly the same thing. It's the perfect analogy.

What do HN comments and AI Overviews have in common?

- All information went through a bunch of neurons at least once

- We don't know which information was even considered

- Might be completely false but presented with utmost confidence

- ...?