# Corrective Intelligence **Domain:** Justice / Machine Intelligence / Governance **Doc Type:** Canonical Synthesis Node **Maturity:** Developed ## Definition **Corrective intelligence** is the capacity to perceive structural imbalance, reconstruct how it was produced, identify interventions that reduce it and route resources toward repair while preserving the agency and rights of those affected. Prediction alone identifies exposure. Corrective intelligence couples perception to obligation. A primitive risk system sees a floodplain and withdraws capital; corrective intelligence asks which drainage, housing, wetland, insurance or relocation intervention reduces danger without converting residents into disposable risk. ## Reparative Architecture [[articles/Climate Justice and Global Reparative Systems|Climate Justice and Global Reparative Systems]] describes corrective intelligence through measurement, provenance, identity, allocation, auditability, appealability and equilibrium. [[articles/Climate Justice and Reparative Equity|Climate Justice and Reparative Equity]] reframes climate data as planetary self-knowledge capable of steering repair. Corrective intelligence therefore requires: - causal rather than merely correlational modeling; - [[wiki/Programmable Reciprocity|Programmable Reciprocity]] between extracted value and returned benefit; - [[wiki/Score Separability|Score Separability]] across life domains; - [[wiki/Model Contestability|Model Contestability]] and meaningful appeal; - local and Indigenous participation in defining harm and success; and - rights constraints that remain prior to optimization. ## Distinction from Optimization An optimizer can minimize a loss function while externalizing unmeasured suffering. Corrective intelligence treats omitted harm, inherited load and institutional responsibility as part of the represented system. It is not synonymous with benevolent automation; it is a demanding constitutional form of [[wiki/Algorithmic Governance|Algorithmic Governance]]. ## Key Insight **Measurement without reciprocity becomes extraction; measurement coupled to enforceable repair can become intelligence.** ## See Also [[wiki/Reparative Justice|Reparative Justice]], [[wiki/Human Resilience|Human Resilience]], [[wiki/Objective Function|Objective Function]], [[wiki/Planetary Administrative Intelligence|Planetary Administrative Intelligence]], [[wiki/Fiduciary Governance|Fiduciary Governance]] ## Simple Reminders, Quotations, and Thoughts > "AI that we will confront is not going to be a mind in an individual machine." > **— Tim O'Reilly**, *2015, Edge annual question “What Do You Think About Machines That Think?”* [[reminders/Machine Succession/Artificial Intelligence Will Not Live in One Machine by Tim O'Reilly|Artificial Intelligence Will Not Live in One Machine by Tim O'Reilly]] > "Machines can beat humans at chess—but they have yet to design the type of mind game that will intrigue humans for centuries." > **— Irene Pepperberg**, *2015, Edge annual question “What Do You Think About Machines That Think?”* [[reminders/Machine Succession/Machines Can Win Without Inventing the Game by Irene Pepperberg|Machines Can Win Without Inventing the Game by Irene Pepperberg]] > "The big question back then was how much the performance of neural networks could improve with the size and depth of the network." > **— Terrence J. Sejnowski**, *2015, Edge annual question “What Do You Think About Machines That Think?”* [[reminders/Machine Succession/Neural Networks Scale With Size and Depth by Terrence J. Sejnowski|Neural Networks Scale With Size and Depth by Terrence J. Sejnowski]] > "It is time for our thinking machines to grow out of an adolescence that has lasted now for sixty years." > **— Thomas A. Bass**, *2015, Edge annual question “What Do You Think About Machines That Think?”* [[reminders/Machine Succession/Thinking Machines Must Grow Beyond Adolescence by Thomas A. Bass|Thinking Machines Must Grow Beyond Adolescence by Thomas A. Bass]]