I’m troubling over the following dilemma:
On the one hand, serendipitous exposure to views unlike your own is good, because that increases the breadth of perspective that’s available to you. You become more cosmopolitan and tolerant.
On the other hand, exposure to views that are hateful, stupid, or evil can be bad, because this can be hurtful, misinforming, or disturbing. Broadly, content can harm.
So, suppose you are deciding what to expose yourself to, or others to, either directly or through the design of some information system.
This requires making a judgment about whether exposure to that perspective will be good or bad.
How is it possible to make that judgment without already having been exposed to it?
Put another way, filter bubbles are sometimes good and sometimes bad. How can you tell the difference, from within a bubble, about whether bridging to another bubble is worthwhile? How could you tell from outside of a bubble? Is there a way to derive this from the nature of bubbles in the abstract?