Usage, privacy, accuracy and limits. Still stuck? See Contact.
Ockham is built specifically for limb inference from limited clues: it enumerates possible causes, filters them with common sense and base rates, then gives a probability, a basis and a way to verify each. Unlike open-ended chat, it follows this structured routine and states plainly that the output is probabilistic inference, not a finding of fact.
No account, no login, and it is free. To keep costs sane there is an hourly and daily quota; when it is reached you will be asked to try again later.
No. Your clues and the output are used for that one request only and are not written to a database. There is no account, no community and no public feed. Request IPs are counted briefly for rate limiting, never linked to your content. See Privacy.
It gives probabilities and hypotheses, not verdicts. The more specific your clues, the narrower the possibilities; vague input yields wide ranges. Any conclusion can be overturned by a new clue — that is by design.
No. Every step is based on the clues you provide plus ordinary physical and social reasoning, and it tells you how to verify. That is the fundamental difference from divination.
A specific question ("What could be causing rhythmic knocking upstairs?") plus one clue per line, with time, location, frequency, change and anything you have already ruled out.
Because with limited, noisy information any "definite answer" is overconfidence. A range plus a verification method is the honest way to express uncertainty.
No. The output is not a finding of fact and not evidence. Do not use it to accuse, profile or identify any individual.
Consult a qualified professional and do not rely on Ockham for these. It can help you organise possibilities, but it cannot replace professional judgement.
After Occam's razor — entities should not be multiplied beyond necessity. It is one of the principles used when ranking options. See Method.