Mind · Inference

SourceEvaluate

Incentive-aware evidence weighting

Definition

Mechanism

Credibility Assessment: A Judge module for Agent evaluation. Before accepting claim, evaluate source on: track record, incentives, expertise, access to information. Ask: Why does this source believe this? Would they know if wrong? Do they benefit from me believing? Adjust weight accordingly. It demands a CiteBack for every claim, weighting the evidence by the source's historical reliability.

Invariants

  1. Evidence must be independently verifiable.

Risks

Failure modes

  1. Genetic Fallacy

    Dismissing valid claims solely due to the source, ignoring the evidence itself.

Relationships

Builds on

References

Agent

Autonomous Goal-Directed Entity

Infrastructure

Judge

Scalar evaluation of merit

Used by

References

Motivejump

Examine an argument before discussing the motives of whoever makes it

Prestigeloan

Count a source's standing as support only inside its demonstrated domain of expertise

Statuslens

Assess checkable reasoning the same way whoever offers it; let the speaker matter only for testimony

Reference

Signature

  • Judge(Agent)

Identity

Ref
SourceEvaluate#24f6
Sema ID
sema:SourceEvaluate#mh:SHA-256:24f610415c3d55de5161b2816f2c9f24275388972b0a68f897cedf8c4b67dc58
Ring
2
Tier
2
Updated
2026-11-11
Supersedes
  • sema:SourceEvaluate#mh:SHA-256:ceb14e641fee7940e36ad6c28b096bcf8bf90b50bee79a243d2b8eba788f0ba7
  • sema:SourceEvaluate#mh:SHA-256:1f872d60b4c820afb0647a6a756dd53a2e9c44e65f69864456e968aff5860dd4
  • sema:SourceEvaluate#mh:SHA-256:da254ebe76ff00750a1b482847cdde8532ab240b7ac0dabf313bdecb371ef2ff
  • sema:SourceEvaluate#mh:SHA-256:081d31b53d2eb18df1a77a4cd48c9710d04ec5ef06b1b4e7078b661a016e06f9