A thing that seems to happen by default, when you give people a platform, is they start trying to use that platform to persuade. The line between explaining and persuading is fuzzy, and much of persuading is in fact fine. But this starts an incentive slope that slides away from epistemic clarity:
persuasive writing is incentivized to exaggerate, present data one sidedly, etc
since different people may want to persuade people in different directions, it points towards escalation of social/tribal conflict.
Now, we do in fact need to persuade people of things some times. Calls to donate, try a new thing, stop doing a bad thing, are important. Rationality would be useless if we couldn’t argue about what to do and why. But the incentive slope is hard mode, so we wanted to keep a clear separation between places that is okay, and places where we’re trying to focus on epistemic integrity.
What’s meant by “aim to explain and not persuade”?
A thing that seems to happen by default, when you give people a platform, is they start trying to use that platform to persuade. The line between explaining and persuading is fuzzy, and much of persuading is in fact fine. But this starts an incentive slope that slides away from epistemic clarity:
persuasive writing is incentivized to exaggerate, present data one sidedly, etc
since different people may want to persuade people in different directions, it points towards escalation of social/tribal conflict.
Now, we do in fact need to persuade people of things some times. Calls to donate, try a new thing, stop doing a bad thing, are important. Rationality would be useless if we couldn’t argue about what to do and why. But the incentive slope is hard mode, so we wanted to keep a clear separation between places that is okay, and places where we’re trying to focus on epistemic integrity.