If a simulation shows a possible future based on assumptions, can we invert this computation to find the assumptions that suggest an acceptable future?
digraph { node [shape=box style=filled fillcolor=gold] rankdir=LR 1 node [fillcolor=lightgreen] 0 [label="Org Diagram" ] 1 [label="Action Adjustment" ] 2 [label="Ideal Constraint" penwidth=3] 3 [label="Action Solution" ] 4 [label="Person Explorer" ] 0->1 [label="Facilitates" labeltooltip="diagram.graph.json"] 1->2 [label="Serves" labeltooltip="diagram.graph.json"] 3->2 [label="Serves" labeltooltip="diagram.graph.json"] 4->3 [label="Performs" labeltooltip="diagram.graph.json"] 4->1 [label="Performs" labeltooltip="diagram.graph.json"] }
Buried in the library is a gradient descent solver that runs the calculations in a blink of an eye. github 