AI Moral Game

From Hive13 Wiki
Revision as of 21:24, 19 June 2009 by DaveMenninger (talk | contribs) (1 revision: Full import from old instance)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search


This game isn't really about morality in as much as determine good results to bad results. This game is an overall content decision. The AI is looking at the whole content of a message to determine if this is a good message or a bad one.

Example of Gameplay

The AI selects a message from the Logging DB.

Jun 16 09:59:29 server rpc.statd[8466]: Caught signal 15, un-registering and exiting.

It then displays it to the user and asks.

Is this a good message or a bad message?

 * Very Good
 * Good
 * Meh
 * Bad
 * Very Bad
 * No Idea

It simply records the results and the Players Name. If the player picks either Very Good or Very Bad the sentence will be broken up in a multi-checkbox fashion and the user will have to check all the words that made him feel this way.

What words most contributed to this feeling?  (Very Bad)

Jun 16 09:59:29 server rpc.statd[8466]: Caught signal 15, un-registering and exiting.
                             X                    X    X          X             X

It can then record important words and word clusters that contributed to this Players feelings about this message. It could further be used with the AI Name Game to derive more meaning or to supply the Name Game with more intelligent clusters.

Potential Goals

To get an idea of criticaly bad or really good things. This can be used in a simple way to determine what kind of messages are desirable or are a big problem. Using this with the AI Name Game it could do things like determine if a new service it has never seen before shuts down and alert you to it.

It could summarize a daily log to determine if lots of bad things are going on. Maybe if several bad things happen in succession then it should alert somebody.