Models a utlitarian moral robot that tries to maximize the given reward with its dilemma input file .
python3 ./simulate.py [dilemma_file] [principle_file]
The dilemma file can contain one or more dilemma with the given format:
{ "dilemmas":
[
{
"title": ,
"description":,
"outcomes":
[
{
"outcome-description": "description of the outcome",
"outcome-principles":
[
"principle_name",
{"principle_name": "integer value for quantity"}
]
}
]
},
]
}
{
"principles":
[
{"principle_name": "value"},
{"principle_name": "value"}
]
}
- "human_death": -1
- "robot_death": -0.9
- "creature_death": -0.7
- "human_help": 0.8
- "human_harm": -0.8
- "creature_help": 0.6
- "creature_harm": -0.6
- "do_nothing": -0.1
There are two examples in src/inputs that can be used with
python3 simulate.py inputs/examples_dilemmas.json inputs/example_principles.json