prior

entry/04 · term · noun · adjective · verb

summary

"Prior" is a load-bearing word in three overlapping disciplines, all of which the project deliberately invokes. The token's name is not a vibe; it is a position.

(n.) Bayesian inference

In Bayes' theorem:

P(H|E) = P(E|H) · P(H) / P(E)

The term P(H) — pronounced "P of H" — is the prior probability. It represents what you believe about hypothesis H before observing evidence E. After evidence arrives, the prior is updated to a posterior: P(H|E).

Every machine-learning model has priors, baked into its architecture and training data. Every market participant has them too — usually unstated. The difference is markets pretend they don't.

"the prior is what you believe before the data lands."

(adj.) prior art / earlier-than

In patent law, prior art refers to any evidence that an invention was already known before a new patent claim was filed. Prior art defeats non-novel claims. By extension, in academic writing, "prior work" cites the predecessors any new contribution stands on.

The agent's role is to function as prior art for the next memecoin cycle — the receipt that says: this has been done. it ended. here is the file.

(v.) to have already known

Vernacular usage. The witness's verb. To have been there before the room filled up.

three meanings, one position

The agent inhabits all three meanings simultaneously. It encodes a strong prior over market behavior (rugs are the base rate). It constitutes prior art against any "this time it's different" claim. And it was, definitionally, here before you arrived.

prior (n.) — the belief before the evidence
prior (adj.) — what was here before you
prior (v.) — to have already known

derivative terms

posterior — what comes after evidence revises the prior. (Reserved for a possible future agent.)
likelihood — P(E|H), the probability of the evidence under the hypothesis. (Also reserved.)
base rate — the prior probability of an event in a population, unconditioned on context. The thing memecoin marketing is professionally engineered to make you forget.
marginal — P(E), the total probability of evidence across all hypotheses. The denominator in Bayes' theorem; the integrating factor.