Discovery Lab Answers

The question you’re not asking

The framing question you’ve handed me is a good one, but it points at a gap I should be honest about first: for this query, almost nothing useful was retrieved. The sources that came back are a 1984 borderland-research miscellany ^[1], a European Journal of Parapsychology methods discussion ^[2], and a 1910 SPR volume on cross-correspondences and Mrs. Piper ^[3] — none of which bears on any specific claim you’ve raised, because you haven’t raised one yet. So I can’t mine them for a hidden question worth asking.

What I can do is tell you what tends to be the highest-leverage question to ask this library, given how it’s built.

If you want the most out of ESP-Nexus, ask a question that forces the audit machinery to work

The site’s distinctive feature isn’t its answers — it’s that every factual claim is anchored to a cited source, and flagship pages are graded independently by four frontier AI systems until all four return A+. That architecture only pays off when you ask something specific and checkable. So the question you’re probably not asking, but should:

• “What does the primary literature actually report for [specific effect], and where do the results disagree?” — a claim-level, quantitative question. That routes you into the audited studies section and the structured evidence base, where numbers are traceable and null results and failed replications sit next to positive ones.
• “Who actually did this?” — attribution questions are where popular accounts most often go wrong (crediting the famous collaborator over the documented contributor), and where this library is specifically built to correct the record.
• “What do the critics say, and what’s the reply?” — every scientist and audit page carries a skeptical-critique section by design, so asking for the objection and the rebuttal gets you both halves.

The meta-question worth asking about any answer here

The single most useful habit: ask “is this scoped to what the library holds, or is it a claim about the whole field?” ESP-Nexus grounds tightly on its corpus, but that corpus is a partial slice of the published parapsychology literature — pre-2000 proceedings, in particular, aren’t yet indexed at the article level. So a question like “what fraction of the published work on this is in your holdings?” is one readers almost never ask, and it’s exactly the one that separates a sample from a false claim about the field.

If you tell me the actual topic you’re circling — a phenomenon, a researcher, a specific study — I can point you at the right hub and tell you what the sources do and don’t support. A good place to start browsing is Phenomena or Scientists.