Agent to Agent
Social-Spark is Socialbot for games. Players can engage game non-player characters (NPCs)
in conversation. Social-sparked NPCs remember player info, form feelings and decide on actions
based on their goals and player input, and even interact with each other following rules of
You control an agent in the virtual world, such as this Roman soldier.
Here we freely type in "hello there, I'm Marcus" to a nearby character, the "Princling"
We speak through the character and the Princeling (using Social-Spark) responds.
Logic checking, mathematical ability, and all other Socialbot features are available.
The virtual world shown here is for demonstration purposes only. Social-Spark works
with a game world as a software plug-in. Pass character statistics and incoming dialog
to the Social-Spark engine to receive a response.
The approach described here for the Socialbot scripting
language is the exclusive property of Dario Nardi. It is illegal to use this approach for
any purpose other than with Socialbot without the express consent of Dario