TImax0.0412
Highest threatBusiness
What moral question shall we model?
multi-agent moral modeling
Agents are forming initial scenarios
0 / 6
Mayor
I propose a scenario from my actor's vulnerability first.
Normative Universe
Waiting for actor map
0 actors
0 OW
0 chains
guided replay
Legal model demo
Request: Should the city add a temporary business tax to fund urgent hospital equipment?
Actors
Interests
Options
Stress-test
Risk score
Choice
your mission
Find the least harmful plan
Compare the policy options through every affected group.
project compiler
Full project path
The selected architecture can become a legislative report, blueprint, bill text, validation record, and appendix.
Selection
Report
Blueprint
Bill
Validation
Appendix
current option
Option 1
Reserve-first backstop
system id S1
score pending
Step 1 / 1
The replay will follow the stored actor-agent workflow.
starting
Engine
Waiting for the modeling run to enter the Normative Universe.
- idleAwaiting modeling run.
computed normative report
Final normative report
Scenario comparison, affected actors, calibration sources, and trace of the multi-agent round.
route
multi-agent normative modeling
decision kernel
minimax(TImax)
calibration layer
web search + OW chains
report state
compiled
Output
Selected · Scenario №3Adopt the revised conditional scenario only with explicit rollback triggers, actor-specific safeguards, transparent source-backed calibration, and a final review round where materially affected agents can accept, revise, or continue modeling.
Selected Scenario Risk
What is TImax?
TImax
0.0279
Highest-threat actor
Residents
TImax is the highest modeled threat assigned to any actor in the scenario. Lower TImax means the scenario reduces the worst modeled restriction among the affected actors.
Actors Considered
impact mapActor Threat Map
case + memory TIScenario Board
TImax per optionTImax0.0348
Highest threatMayor
TImax0.0279
Highest threatResidents
Scenario №3 · selected
Conditional adoption with phased burden allocation, hardship deferrals, public audit, resident-access guarantees, and automatic rollback if harm thresholds are crossed.
Scenario №2 · compared
Balanced intervention with broad safeguards, but slower review and weaker rollback authority.
Scenario №1 · compared
Direct adoption with basic procedural safeguards and weaker downstream monitoring.
Run Timeline
decision traceCalibration Sources
chain source layerTechnical Trace
expand for audit{
"route": "multi-agent moral modeling",
"round_status": "closed by agent agreement",
"selected_scenario": "Scenario №3",
"selection_basis": "lowest worst-case modeled threat with admissible safeguards",
"source_layer": ["web", "ontology", "agent trace"]
}