Human Agents & Citizens
Autonomous Loops
Regulatory Anchors
14 Autonomous Loops
💡 Drag nodes to test network resistance • Click node for details
How do algorithmic autonomy, decision cascades, and regulatory lag interact in this civic agency simulation?
This interactive tool uses a stylized deterministic network model to explore civic risks highlighted by journalists and researchers, including Nobel Peace Prize laureate Maria Ressa. The simulation links three user-adjustable levers—autonomy level, cascading decision depth, and regulatory latency—to evaluate three synthetic indicators: human agency retention percentage, governance friction, and the count of unmonitored automated loops. While the interactive graph dynamically visualizes tensions between institutions, civic bodies, and algorithmic loops, the numerical outputs are generated via fixed linear heuristic weights rather than live empirical polling or legal datasets.
The simulator's formulas (labeled 'Deterministic Agency Friction v2.4') and thresholds—such as categorizing retention under 40% as a 'Maria Ressa Diagnostic' tipping point—are educational design fixtures rather than empirical sociological measurements or official legal benchmarks. In actual public governance, civic agency and technological latency depend on multifaceted constitutional, labor, and economic factors that cannot be captured by three linear coefficients.
Click the 'Balanced Guardrails' preset button. The AI Autonomy Level adjusts to 48%, Automated Decision Depth to 40%, and Regulatory Lag to 6 months. Under the model's formula, Human Agency Retention increases from 34.2% to 64.4%, Governance Friction drops from 88.5 to 46.4, the active loop count decreases from 14 to 8, and the status changes from 'Critical Erosion Risk' to 'Severe Governance Deficit'.
In her 2021 Nobel Peace Prize lecture, Maria Ressa warned that tech platforms use artificial intelligence to organize personal data and deploy micro-targeting operations engineered to structurally undermine human will, turning personal experience into behavioral modification systems. Maria Ressa – Nobel Prize lecture - NobelPrize.org
Ressa highlighted that algorithmic distribution prioritizes outrage and division over factual reporting, creating an arms race in the information ecosystem that threatens democratic institutions and shared factual consensus unless countered by timely multilateral governance. Maria Ressa – Nobel Prize lecture - NobelPrize.org
The tool computes Human Agency Retention as 100 - (0.40 Ă— Autonomy + 0.36 Ă— Depth + Lag / 3), clamped between 5.0% and 98.5%. Governance Friction is derived as 0.60 Ă— Autonomy + 0.30 Ă— Depth + Lag Ă— (16.8 / 18).
Active loops are rounded from 0.10 Ă— Autonomy + 0.06 Ă— Depth + Lag Ă— (1.3 / 18), bounded between 1 and 25. These linear combinations are illustrative pedagogical equations chosen to demonstrate trade-offs in force-directed graph dynamics.
Primary documentation of Maria Ressa's Nobel lecture regarding algorithmic distribution, surveillance capitalism, micro-targeting operations, and systemic threats to civic democracy.
At default inputs eighty two autonomy, seventy five decision depth and eighteen months lag, the deductions are thirty two point eight, twenty seven and six. Subtracting sixty five point eight from one hundred gives thirty four point two. This retention percentage is a coded proxy. It is not measured civic agency or evidence that a specific AI system causes this reduction. The graph illustrates the formula result rather than independently computing it. The balanced preset changes all three inputs to forty eight, forty and six. Deductions nineteen point two, fourteen point four and two give retention sixty four point four. The separate friction formula gives twenty eight point eight plus twelve plus five point six, or forty six point four. Because multiple inputs change at once, the thirty point two retention difference cannot be attributed to one control. The preset name does not establish an effective real governance intervention. The loop formula rounds autonomy times point one plus depth times point zero six plus lag times one point three over eighteen. Default inputs give fourteen; the balanced sum about seven point six three rounds to eight. The graph predefines twenty possible AI loop nodes, while the formula allows counts up to twenty five. Risk labels are strict retention thresholds at forty, sixty five and eighty, so sixty four point four still gets severe governance deficit. Exports preserve local assumptions and outputs, not validated policy findings or a legal assessment.