Artificial Intelligence
Machines that choose, predict, classify, and speak—and the institutions that must decide what to do with them.
A Few Places to Begin
Representative entry points from across the site.
-
The Model Is Not the System
-
Crystal Cube: Forecasting Disruptive Events
Crystal Cube treated disruption forecasting as a practical systems problem: combine disparate signals, make the model useful, and confront the difference between a prediction and an operational decision.
-
Kami
An agent-based modeling library in modern C++, developed at the Johns Hopkins Applied Physics Laboratory. It provides the pieces for modeling individual actors and their environments when collective behavior cannot be treated as a simple aggregate.
Scholarship
-
Crystal Cube: Forecasting Disruptive Events
Crystal Cube treated disruption forecasting as a practical systems problem: combine disparate signals, make the model useful, and confront the difference between a prediction and an operational decision.
-
Agent-Based Modeling for Evaluation of a Wearable-Sensor-Based Disease Surveillance Network
This work used agent-based modeling to ask whether a disease-surveillance network could detect what it claimed to detect under realistic operating assumptions.
Software
-
Kami
An agent-based modeling library in modern C++, developed at the Johns Hopkins Applied Physics Laboratory. It provides the pieces for modeling individual actors and their environments when collective behavior cannot be treated as a simple aggregate.
-
Virtual Bumblebees
An artificial-life environment based on a deliberately different reading of Langton’s ant: place bees and dots on the board, then watch a simple rule system produce behavior that is considerably less simple.