Announcement

Collapse
No announcement yet.

Opportunity for "reinforcement learning" AI in UT?

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

    Opportunity for "reinforcement learning" AI in UT?

    Is this a topic that has been brought up before? Lets say the bots in UT use a basic self reinforcing algorithm similar to googles Alpha go and have a set of average human parameters such as reaction time and accuracy, as well as the basic programming needed to navigate the game. Whenever a player plays against a bot the information from that match could be uploaded and compiled. The program could retain information interpreted as beneficial and resulted in its success, and discard the opposite. I think UT would be a great choice because of its simplicity.

    It would be intriguing and perhaps terrifying to play against bots with a sense of human intuition, rather than inflexible programming. Could you imagine AI eventually being deceitful, perhaps triggering audio cues, or feigning weakness in attempts to ambush?

    Has there been any discussion on this?
Working...
X