0.7.7 Evidence and Verification

A Claim Becomes More Useful When You Can Support It

AI can generate factual claims quickly.

Before relying on an important claim, ask:

Verification means checking a claim against appropriate evidence.

The goal is not simply to find another page that repeats the same statement.

The goal is to determine whether the claim is supported by information that is relevant and trustworthy for the task.

Start by Identifying the Claim

Suppose an AI response says:

Before verifying it, separate the claim:

Activity diagram showing AI claim, then identify what must be true, then find appropriate evidence, then compare evidence with claim, then decide whether the claim is supported.
AIClaim.
  • New
  • Changed
  • Removed

The evidence controls the verification result.

The AI's confidence does not.

Verification Can Change What You Do With the Output

After verification, you might:

  • keep the original statement;
  • correct a factual detail;
  • remove an unsupported claim;
  • add a qualification;
  • reject the recommendation;
  • seek better evidence.

Verification is therefore part of decision-making, not a separate academic exercise.

The Main Habit

When an AI output contains an important factual claim:

  1. Identify the exact claim.
  2. Decide what kind of evidence could support it.
  3. Find evidence appropriate to that domain.
  4. Compare the evidence with the claim.
  5. State whether the claim is supported, contradicted, partly supported, or unresolved.

The next Learning Activity focuses on choosing a source with the right authority for that verification.