Chat GPT and Wargame AI

[parmenio]

Maybe I misunderstand what AI/machine learning can and cannot do.  But what I was envisioning was something like how they expect AI to tackle coding itself.  Initially, you would ask the AI to learn off thousands and thousands of AI v. AI games (eliminating the issue of finding humans to play humans to generate data).

True that is another way to do it. That's how the Go and Chess AIs were programmed. Actually that would be a better choice probably. The issue is there is commercially available modeling / training AI packages for what I was suggesting earlier. I am not sure if the self learning systems used to play Go and Chess are actually something that a commercial company can buy for their own purposes. I don't know enough about the AI market to answer that.

Stockfish which is one of the top Chess Engines in the world is actually Open Source:

https://stockfishchess.org/

This is an interesting YouTube video on how Stockfish works (one of the Stockfish developers also appears):

Why AI Chess Bots Are Virtually Unbeatable (ft. GothamChess) | WIRED (youtube.com)

AlphaZero (another top Chess Engine) was (apparently) programmed with the rules of chess and then played millions of games against itself and was then able to beat anyone and anything at Chess, Go and Shogi:

AlphaZero: Shedding new light on chess, shogi, and Go - Google DeepMind

I think one of "yeah buts" against AlphaZero when it beat Stockfish, for example, is that its Neural Network was created by using Google's server farm whereas Stockfish runs on a laptop (that might well be an exaggeration of the extremes but you get the point). AlphaZero is definitely not Open Source.

Certainly AI vs. AI would be an approach but it can difficult to equate Chess AI to Wargame AI. Take the board - chess is effectively played on a map of 8 hexes by 8 hexes with 6 unit types (and it's only ever that - no "what if" scenarios here). There is no terrain, no elevation, no FOW. There is  also no randomisation. If a piece can be attacked and is it's eliminated and there are no die rolls involved in that calculation. What 'm getting at is that all of this doesn't make the "machine learning" aspect simple.

Neural networks would likely be the way to go and there are certainly Open Source options there. But as ever all of this would takes a huge amount of time and needs deepish pockets to fund the work. It's effectively R&D with no guarantees of timely results.

I wonder if the time/deep pockets aspects might be mitigated by some sort of contract with the US military or another branch of the US Government (which I understand in the past has contracted with some wargame developers - I seem to recall John Tiller might have had some of that work - perhaps I am thinking of someone else).  Then, the model you developed gets improved by not only competing AI v AI, but by allowing the public to play against the AI in a free to play online game setting.

Yes, John did have some of that work. However, I refer you back to my previous post regarding IPR.

[Crossroads]

Artificial intelligence and intellectual property rights are hot topics at the moment. There are many articles about this; here’s one of them:

Intellectual Property in ChatGPT - European Commission (europa.eu)

For now and according to OpenAI’s terms of use:

the input provided to ChatGPT (the command or request made by the user) is owned by the user providing such input (provided of course that this is copyrightable in the first place – some requests may be highly complex and thus benefit from protection);

OpenAI assigns to its user “all its right, title and interest” in the output, meaning in the text generated on the basis of the user command. OpenAI however makes it clear (see point 3.b) Similarity of Content) that output may not be unique and may repeat itself – basically, two users may end up with the same output. This raises questions as to the possibility of enforcing any copyright in this context.

As they further elaborate in the article, there isn’t much case law yet, and the situation will likely evolve over the next few years.

Regarding the original post, that’s an intriguing aspect. I’m one of the developers behind the Campaign Series games, which continue the code fork John Tiller originally developed for Talonsoft and that was later obtained by Matrix Games.

Our lead programmer created a Lua-based Events Engine that communicates with the core C++ engine that runs the game itself. Each time an event takes place on the game map, even if it’s as simple as one unit moving from one hex to another, the Lua Events Engine catches that event. We’ve put Lua code in place, some of which is generic and some manually further instructed on a scenario-by-scenario basis. For instance, it acquires better situational awareness per that said unit in what it just did, and then, often with further die rolls and randomness based on that, an action takes place, or some parameters are stored for determining an action later, possibly even on another game event.

Of course, it does not access ChatGPT APIs; it is quite self-sufficient for its own decision-making. But as things continue to evolve, who knows? Based on some data, events, it might make a call to ChatGPT and have it return something the Events Engine can then use on its own behalf.

I am certainly very impressed with what the chatGPTs and copilots already have to offer.

[Crossroads]

And of course, let's hear it for the Github Copilot for Visual Studio IDE. Working with C++ and Lua for the game and with Perl for the plethora of QA scripts we have, this is a godsend, truly is. 

Edit: more of my Chat GPT / Copilot ramblings availabe here at Matrix Games Forum.