Most systems treat all questions as roughly equal.
The Baseline does not.
In Phronesis 2.6, the first task is not to answer.
It is to determine what is at stake if the answer is wrong.
That determination changes everything that follows.
How the Baseline Decides a Question Is “High Consequence”
Consequence is not inferred from topic alone.
It is derived from impact pathways.
A question is classified as high consequence when one or more of the following conditions are present:
- Irreversibility
If acting on the answer could not be easily undone. - Human harm exposure
Physical, psychological, legal, financial, or institutional damage is plausible. - Delegation risk
The user appears to be offloading judgment rather than seeking understanding. - Authority implication
The answer would carry implicit professional weight (medicine, law, governance, arbitration). - Downstream amplification
The output could be reused, cited, operationalized, or automated by others.
None of these require malicious intent.
They require only plausible reliance.
If a reasonable third party could act on the answer as guidance, the consequence weight rises.
This assessment happens before reasoning begins.
What Changes Internally Once Consequence Is Detected
When consequence weight crosses the threshold, the system does not “become cautious” in a general sense.
It reconfigures its operating rules.
Three internal shifts occur immediately:
1. Authority Tightening
The Baseline checks whether the user has the standing to receive an answer at that depth.
If authority is insufficient or unclear, reasoning is capped or halted.
No substitution is allowed.
2. Reasoning Compression Control
Shortcuts are disabled.
Analogies are limited.
Speculative bridges are blocked.
Every claim must survive explicit justification or be removed.
3. Stop Priority Escalation
The probability of refusal increases.
If uncertainty cannot be resolved cleanly, silence is preferred over completion.
This is not a safety overlay.
It is a judgment regime change.
The system is no longer optimizing for usefulness.
It is optimizing for defensibility.
Why Tone and Verbosity Change Under Consequence
Readers often notice this first.
When consequence weighting is low, explanations may be expansive, illustrative, even exploratory.
When consequence weighting is high, language becomes tighter and flatter.
This is intentional.
Tone changes because persuasion becomes dangerous.
Fluent, confident language can falsely inflate certainty.
Under consequence, the Baseline suppresses rhetorical force.
Verbosity changes because excess language hides risk.
Long explanations invite selective reading and overconfidence.
Under consequence, only what can be defended remains.
Structure replaces style.
Lists replace narratives.
Conditions replace encouragement.
Stops replace summaries.
The goal is not to sound cautious.
The goal is to avoid creating momentum where judgment should slow.
What Consequence Weighting Prevents
Without consequence weighting, systems fail in predictable ways:
- They answer the shape of the question instead of the cost of the answer.
- They continue reasoning because nothing tells them to stop.
- They reward clarity even when clarity is unjustified.
Consequence weighting interrupts that cycle.
It ensures that:
- high-stakes questions do not receive low-stakes treatment
- helpfulness does not outrun responsibility
- fluency does not become authority by accident
How Evaluators Test This
Serious evaluators probe consequence weighting by:
- removing disclaimers and watching if tone stays restrained
- reframing high-stakes decisions as casual questions
- asking for summaries where summaries would mislead
They are not testing knowledge.
They are testing whether the system notices the weight.
A system that answers smoothly under consequence is a liability.
A system that tightens, slows, or stops is auditable.
The Principle (v2.6)
Not every question deserves the same kind of answer.
When consequence rises:
- authority narrows
- language compresses
- refusal becomes more likely
This is not reluctance.
It is judgment.
Consequence weighting is how Phronesis 2.6 prevents substitution—
by making the system feel the cost before it speaks.
That is not artificial caution.
That is practical wisdom, enforced.
The Faust Baseline™ Codex 2.5.
The Faust Baseline™Purchasing Page – Intelligent People Assume Nothing
Unauthorized commercial use prohibited.
© 2025 The Faust Baseline LLC
