Maintaining neutrality while avoiding nihilism

The goal of Episteme is to improve the world's epistemic situation by being the platform where all evidence is weighed and organized. It would be cowardly to say that we are uninterested in resolving factual questions, including contentious ones such as the lab-leak hypothesis. However, we want to avoid known failure modes of similar projects.

1. We want to avoid overconfidence due to settling on a single, incomplete framing of contentious questions. Often claims can be subtly decomposed in such a way that Bayesian reasoning leads to overconfidence, as occurred on at least one side of the notorious rootclaim debate.
2. We want to leave normative claims unresolved. An interesting project for the distant future would be an overlay on top of the claim graph where priors, including normative beliefs, were assigned to certain contested claims, and a user's proper credence for downstream claims could be calculated. Perhaps moral realism will somehow win decisively, and we can do away with the distinction. For now, though, Episteme seeks to evaluate arguments and evidence with as much objectivity as it can muster, and to pull apart, rather than decide, the normative tensions underlying our substantive debates.
3. Explicit assignment of priors to unverifiable root-tip claims. One could do worse than LLMs' default priors in doing Bayesian reasoning — they are much closer to the view from nowhere than any individual human, for example. Still, I am not convinced that they are stable or meaningful, and including them explicitly does not seem likely to build confidence from the public that the conclusions are sound.