With the AI revolution happening it would be cool if VASSAL could tap into that. One way to do that is enable VASSAL to call AI Agents.
e.g. Google
Thus a trigger might be moving a piece or firing a weapon or whatever and the agent the could use AI to decide what to do next, such as call other agents or make a change to the some entity in Vassal such as reveal another piece.
So that the game can play itself?
Go for it. But I should warn you that you won’t find the interface to be very convenient for interacting with anything programmatically until V4.
Hi all,
I’m implementing this kind of stuff at work (my current task): is there anyone actually working on that ? I would be glad to help, since I was thinking exactly the same thing recently….
Over at BoardGameGeek, the question of computer opponents (not just Artificial Neural Nets - or ANN - a.k.a. “AI”) often shows up - not least in the Wargame - General forum. See for example
TL,DR: Vassal 3 is rather poorly suited for computerised opponents - however implemented (expert system, ANN, and so on).
Vassal 4 promises to be better suited for this, but there’s no way of knowing because - so far, and as far as I know - Vassal 4 is vaporware. @uckelman is there any kind of preview that you could share? Besides design thoughts and the like already tagged in the forum. I think perhaps some transparency could focus the effort and not splinter off in different directions (e.g., this one).
Yours,
Christian
I’ve posted a short-term development plan here.
I saw a link to this post in my own post.
My prototype is in a too early stage to make statments about AI integration.
If the issue is about retrieving information about game state and feeding it to a tool, I don’t think that would be any problem.