You mention truth as needing to be useful, and I think that's true. In a nutshell, if I were to propose an "epistemology of common sense", it would be this:
- All we truly know is the qualia of our perceptions.
- Our qualia only "matter" insofar as they are good or bad (e.g. pleasurable or painful).
- All other so-called "knowledge" is really just models that prove to be useful.
- Something is only "useful" if it pertains to something that matters, which is to say, whether it has a pleasurable or painful effect on our qualia.
- Therefore, a mental model is only useful insofar as it helps us make predictions that help us bring about desirable qualia.
I called that an epistemology of common sense. And I really want to emphasize the word "common". Because I think everything I said there would apply to any sentient being, not just humans.
And I think if we start from there, we can make a truly global truth community, which could have convergent, and therefore collaborative beliefs, more like math and science, and less like religion.
And I think we really need to focus on bringing the world closer to that point, or we will never be able to get along.
I've written extensively on this from a couple of different angles in these articles, if you care to take a look:
