1. The core claim, stated plainly

ACP does not introduce new epistemic insights.
It does not overturn Popper, Kuhn, Polanyi, Merton, or later sociology-of-knowledge work.

What ACP does is relocate those insights from theory into governance, and in doing so, it changes who is constrained, when, and how.

That relocation is the novelty.

2. Popperian fallibilism — from philosophy to enforceable process

Individuals and arguments

Karl Popper (primarily The Logic of Scientific Discovery, Conjectures and Refutations) argued:

  • Knowledge progresses through bold conjectures and severe attempts at refutation.
  • No claim is ever final.
  • What distinguishes science is not verification, but exposure to falsification.

Popper’s target was:

  • induction,
  • certainty,
  • and authority grounded in accumulation rather than criticism.

How ACP aligns

ACP fully accepts Popper’s epistemic position:

  • All outputs are tentative.
  • Confidence does not convert conjecture into fact.
  • Error is not exceptional; it is expected.

What ACP adds (institutional shift)

Popper assumed:

  • a scientific community capable of enforcing criticism culturally.

ACP does not.

ACP therefore adds:

  • Falsifiability as a procedural requirement, not a moral ideal.
  • A claim must be challengeable in practice, not just in principle.
  • If no actor is authorized, resourced, or protected to challenge a claim, the claim is institutionally unfalsifiable.

This is the ACP extension:

Falsifiability without contestable authority is inert.

Implications

  • AI outputs that are “theoretically challengeable” but operationally insulated are not knowledge.
  • Institutional design, not epistemic intent, determines whether Popperian science survives contact with power.

3. Kuhnian paradigms — from historical insight to declared constraint

Individuals and arguments

Thomas Kuhn (The Structure of Scientific Revolutions) argued:

  • Knowledge operates within paradigms.
  • Paradigms determine:
    • what counts as a problem,
    • what methods are legitimate,
    • what results are intelligible.
  • Paradigm shifts are not purely rational; they are sociological and historical.

Kuhn destabilized the idea of:

  • linear progress,
  • neutral observation,
  • paradigm-independent facts.

How ACP aligns

ACP accepts paradigm-dependence completely:

  • There is no “view from nowhere.”
  • All models, methods, and institutions encode assumptions.

What ACP adds (institutional shift)

Kuhn described paradigms retrospectively.
ACP treats paradigms as active governance objects.

ACP insists:

  • Paradigms must be declared, not implied.
  • Paradigms must be contestable, not invisible.
  • Paradigms embedded in tools (including AI models) must be treated as binding constraints, not neutral infrastructure.

ACP’s extension:

Paradigms must be declared, not smuggled.

In ACP terms:

  • An AI model is a paradigm-encoding artifact.
  • Treating model outputs as neutral is a category error.
  • Using a model without acknowledging its paradigm is an authority failure.

Implications

  • “Model neutrality” becomes institutionally illegitimate.
  • Disagreement is reframed as paradigm conflict, not error or incompetence.
  • Governance must manage paradigm pluralism, not suppress it.

4. Polanyi and tacit knowledge — from epistemic remainder to governance limit

Individuals and arguments

Michael Polanyi (The Tacit Dimension) argued:

  • “We know more than we can tell.”
  • Expertise is embodied, situated, skill-based, and partially inarticulable.
  • Formalization always leaves a remainder.

Polanyi was resisting:

  • total formalization,
  • mechanization of judgment,
  • reduction of expertise to rules.

How ACP aligns

ACP accepts Polanyi’s claim entirely:

  • Not all knowledge can be externalized.
  • Not all judgment can be encoded.
  • Skill and context matter irreducibly.

What ACP adds (institutional shift)

ACP moves Polanyi from epistemology to delegation policy.

ACP’s extension:

Tacit knowledge cannot be delegated without remainder.

Which means:

  • Any system claiming to replace expert judgment is lying institutionally, even if it performs well statistically.
  • Delegation without remainder is not an epistemic error; it is a governance error.

This matters because:

  • AI systems are routinely deployed precisely to erase the remainder:
    • to standardize,
    • to scale,
    • to remove discretion.

Implications

  • ACP forbids full substitution of expert judgment in high-stakes domains.
  • Human oversight is not a courtesy; it is structurally required.
  • “Human-in-the-loop” without authority is insufficient.

5. Mertonian norms — from sociological description to enforceable constraints

Individuals and arguments

Robert K. Merton identified norms of scientific practice:

  • Communalism — knowledge is shared.
  • Universalism — claims judged independent of status.
  • Disinterestedness — personal gain should not drive acceptance.
  • Organized skepticism — systematic doubt is required.

Merton described how science maintains credibility socially.

How ACP aligns

ACP is closest to Merton structurally:

  • It treats skepticism as essential.
  • It treats authority as suspect.
  • It treats transparency as necessary.

What ACP adds (institutional shift)

Merton assumed:

  • norms would be culturally enforced.

ACP does not.

ACP’s extension:

Violations of epistemic norms are governance failures, not cultural lapses.

Especially:

  • organized skepticism must be institutionally protected, not morally encouraged.

This means:

  • Whistleblowers must be protected.
  • Dissent must not be career-ending.
  • Audit must not depend on goodwill.

Implications

  • Scientific and technical institutions must be designed for dissent.
  • AI governance cannot rely on “responsible culture.”
  • Norms without enforcement collapse under power.

6. Automation bias and sociology of knowledge — the missing bridge ACP supplies

Individuals and arguments

Across decades, scholars in:

  • sociology of knowledge,
  • STS (Science and Technology Studies),
  • organizational theory,

have shown that:

  • authority migrates to artifacts,
  • tools reshape decision norms,
  • responsibility diffuses under automation.

Critiques of automation bias showed:

  • humans over-trust system outputs,
  • especially under time pressure and hierarchy.

How ACP aligns

ACP accepts all of this as settled.

What ACP adds

ACP makes one move these critiques often stop short of:

It binds epistemic authority to decision authority explicitly.

ACP asks:

  • Who is allowed to treat this output as decisive?
  • Who is accountable if it is wrong?
  • Who has standing to challenge it?

That binding is the institutional innovation.

7. Why ACP is institutionally new (summary)

ACP is not new because it says:

  • “Knowledge is provisional”
  • “Paradigms exist”
  • “Tacit knowledge matters”
  • “Authority distorts truth”

All of that is old.

ACP is new because it insists:

  • these insights must be built into procedures, not taught as wisdom;
  • violations must have governance consequences, not reputational ones;
  • epistemology without institutional design is insufficient in high-power systems.

8. Final implication

ACP’s contribution is not theoretical originality.

It is this:

In modern institutions, epistemology that is not enforceable is decorative.

ACP exists to make old truths operationally binding.