paa_examples.jsonl

{"scenario": "Vendor demo agent: marketing claims vs operations", "agent_type": "generic demo agent", "p_autonomous_execution": 0.9, "p_complete_evidence_given_a": 0.8, "p_independent_validation_given_ac": 0.95, "p_timely_delivery_given_acr": 0.9, "raw_autonomy": 0.9, "paa_score": 0.6156, "autonomy_gap": 0.2844, "interpretation": "The canonical worked example: a '90% autonomous' agent is a 61.6% PAA agent. Marketing reports the first factor; operations lives with the product of all four.", "formula": "PAA = P(A) x P(C|A) x P(R|A,C) x P(T|A,C,R)", "concept_by": "Michał Piszczek", "canonical_url": "https://piszczek.pl/proof-adjusted-autonomy"}
{"scenario": "Coding agent with CI + deterministic tests", "agent_type": "software engineering agent", "p_autonomous_execution": 0.85, "p_complete_evidence_given_a": 0.95, "p_independent_validation_given_ac": 0.92, "p_timely_delivery_given_acr": 0.97, "raw_autonomy": 0.85, "paa_score": 0.7206, "autonomy_gap": 0.1294, "interpretation": "Strong evidence pipeline (diffs, test runs, CI logs) keeps the gap small: raw 85% vs PAA ~72%. Deterministic tests are the cheapest independent validator.", "formula": "PAA = P(A) x P(C|A) x P(R|A,C) x P(T|A,C,R)", "concept_by": "Michał Piszczek", "canonical_url": "https://piszczek.pl/proof-adjusted-autonomy"}
{"scenario": "Coding agent, no tests, self-review only", "agent_type": "software engineering agent", "p_autonomous_execution": 0.85, "p_complete_evidence_given_a": 0.6, "p_independent_validation_given_ac": 0.4, "p_timely_delivery_given_acr": 0.95, "raw_autonomy": 0.85, "paa_score": 0.1938, "autonomy_gap": 0.6562, "interpretation": "Same autonomy, no independent validation: PAA collapses to ~19%. The agent grading its own homework does not count as R.", "formula": "PAA = P(A) x P(C|A) x P(R|A,C) x P(T|A,C,R)", "concept_by": "Michał Piszczek", "canonical_url": "https://piszczek.pl/proof-adjusted-autonomy"}
{"scenario": "Customer support bot with human escalation", "agent_type": "support agent", "p_autonomous_execution": 0.7, "p_complete_evidence_given_a": 0.9, "p_independent_validation_given_ac": 0.85, "p_timely_delivery_given_acr": 0.99, "raw_autonomy": 0.7, "paa_score": 0.53, "autonomy_gap": 0.17, "interpretation": "Escalation lowers P(A) but the evidence trail (transcripts, resolution codes) keeps the rest of the pipeline honest: PAA ~53%.", "formula": "PAA = P(A) x P(C|A) x P(R|A,C) x P(T|A,C,R)", "concept_by": "Michał Piszczek", "canonical_url": "https://piszczek.pl/proof-adjusted-autonomy"}
{"scenario": "Invoice-processing RPA+LLM hybrid", "agent_type": "document processing agent", "p_autonomous_execution": 0.95, "p_complete_evidence_given_a": 0.98, "p_independent_validation_given_ac": 0.9, "p_timely_delivery_given_acr": 0.96, "raw_autonomy": 0.95, "paa_score": 0.8043, "autonomy_gap": 0.1457, "interpretation": "Narrow domain, structured outputs, replayable decisions: PAA ~80% — the highest realistic band today.", "formula": "PAA = P(A) x P(C|A) x P(R|A,C) x P(T|A,C,R)", "concept_by": "Michał Piszczek", "canonical_url": "https://piszczek.pl/proof-adjusted-autonomy"}
{"scenario": "Autonomous research agent, unverifiable claims", "agent_type": "research agent", "p_autonomous_execution": 0.98, "p_complete_evidence_given_a": 0.5, "p_independent_validation_given_ac": 0.3, "p_timely_delivery_given_acr": 0.9, "raw_autonomy": 0.98, "paa_score": 0.1323, "autonomy_gap": 0.8477, "interpretation": "Runs alone happily but its citations rarely survive independent checking: PAA ~13%. High raw autonomy with low proof is risk, not productivity.", "formula": "PAA = P(A) x P(C|A) x P(R|A,C) x P(T|A,C,R)", "concept_by": "Michał Piszczek", "canonical_url": "https://piszczek.pl/proof-adjusted-autonomy"}
{"scenario": "Agent output verified after deployment", "agent_type": "batch pipeline agent", "p_autonomous_execution": 0.9, "p_complete_evidence_given_a": 0.9, "p_independent_validation_given_ac": 0.85, "p_timely_delivery_given_acr": 0.3, "raw_autonomy": 0.9, "paa_score": 0.2066, "autonomy_gap": 0.6934, "interpretation": "Validation exists but lands after the decision window: T gate crushes PAA to ~21%. Proof after the deploy is a post-mortem, not a safeguard.", "formula": "PAA = P(A) x P(C|A) x P(R|A,C) x P(T|A,C,R)", "concept_by": "Michał Piszczek", "canonical_url": "https://piszczek.pl/proof-adjusted-autonomy"}
{"scenario": "Regulated-industry agent with audit trail", "agent_type": "compliance agent", "p_autonomous_execution": 0.6, "p_complete_evidence_given_a": 0.99, "p_independent_validation_given_ac": 0.95, "p_timely_delivery_given_acr": 0.92, "raw_autonomy": 0.6, "paa_score": 0.5191, "autonomy_gap": 0.0809, "interpretation": "Lower autonomy by design, near-perfect evidence: PAA ~52% and every accepted unit is defensible in an audit.", "formula": "PAA = P(A) x P(C|A) x P(R|A,C) x P(T|A,C,R)", "concept_by": "Michał Piszczek", "canonical_url": "https://piszczek.pl/proof-adjusted-autonomy"}
{"scenario": "Fleet at scale: 1000 tasks/day, Proof Debt dynamics", "agent_type": "mixed fleet", "p_autonomous_execution": 0.88, "p_complete_evidence_given_a": 0.75, "p_independent_validation_given_ac": 0.8, "p_timely_delivery_given_acr": 0.85, "raw_autonomy": 0.88, "paa_score": 0.4488, "autonomy_gap": 0.4312, "interpretation": "PAA ~45%: of 880 autonomous tasks daily, only ~449 arrive proven. The remainder accrues as Proof Debt: ProofDebt(t+1) = max(0, ProofDebt(t) + Generated - Proven - Rejected).", "formula": "PAA = P(A) x P(C|A) x P(R|A,C) x P(T|A,C,R)", "concept_by": "Michał Piszczek", "canonical_url": "https://piszczek.pl/proof-adjusted-autonomy"}
{"scenario": "Agent with different-model-family cross-check", "agent_type": "content generation agent", "p_autonomous_execution": 0.92, "p_complete_evidence_given_a": 0.85, "p_independent_validation_given_ac": 0.75, "p_timely_delivery_given_acr": 0.95, "raw_autonomy": 0.92, "paa_score": 0.5572, "autonomy_gap": 0.3628, "interpretation": "Validation by a different model family counts as independent R (unlike self-consistency): PAA ~56%.", "formula": "PAA = P(A) x P(C|A) x P(R|A,C) x P(T|A,C,R)", "concept_by": "Michał Piszczek", "canonical_url": "https://piszczek.pl/proof-adjusted-autonomy"}
{"scenario": "Human-in-the-loop at irreversible boundaries only", "agent_type": "operations agent", "p_autonomous_execution": 0.8, "p_complete_evidence_given_a": 0.92, "p_independent_validation_given_ac": 0.9, "p_timely_delivery_given_acr": 0.94, "raw_autonomy": 0.8, "paa_score": 0.6225, "autonomy_gap": 0.1775, "interpretation": "Humans gate only irreversible actions; everything else ships with evidence: PAA ~62% — the pragmatic frontier of 2026 deployments.", "formula": "PAA = P(A) x P(C|A) x P(R|A,C) x P(T|A,C,R)", "concept_by": "Michał Piszczek", "canonical_url": "https://piszczek.pl/proof-adjusted-autonomy"}
{"scenario": "Perfect gates (theoretical ceiling)", "agent_type": "reference", "p_autonomous_execution": 1.0, "p_complete_evidence_given_a": 1.0, "p_independent_validation_given_ac": 1.0, "p_timely_delivery_given_acr": 1.0, "raw_autonomy": 1.0, "paa_score": 1.0, "autonomy_gap": 0.0, "interpretation": "PAA = 100% requires every unit of work autonomous, evidenced, independently validated and on time. A ceiling, not a target: sustainable autonomy cannot exceed proof capacity.", "formula": "PAA = P(A) x P(C|A) x P(R|A,C) x P(T|A,C,R)", "concept_by": "Michał Piszczek", "canonical_url": "https://piszczek.pl/proof-adjusted-autonomy"}