LLM 玩 8 位指挥官 X16 游戏，使用结构化的“智能感知”。

LLM 玩 8 位指挥官 X16 游戏，使用结构化的“智能感知”。
LLM plays an 8-bit Commander X16 game using structured "smart senses"

## PvP-AI：复兴8位游戏并挑战人工智能 PvP-AI是对1990年最初编写的游戏的重现，现在在Commander X16复古电脑上重获新生。虽然模拟器可以达到每秒8.6帧的可玩帧率，但由于VERA模块问题，硬件限制为每秒4帧。该游戏具有独特的机制——动态变化的竞技场、智能导弹和优先考虑淘汰而非纯粹得分的计分系统。受到LLM与复古游戏集成的启发，作者探索使用ChatGPT (gpt-4o)来玩PvP-AI。为了避免依赖视觉输入，一个“智能感知”系统——基于文本的数据，代表触觉和电磁场检测——被实施，以向LLM提供关键的游戏信息。为了促进通信，一个新的VIA2-socket功能被添加到x16-模拟器中，允许LLM、PHP接口和游戏之间进行双向交互。初步测试，记录在一系列连续的游戏中，展示了ChatGPT学习和发展获胜策略的能力，并在比赛之间携带持久笔记。未来的研究旨在扩展“智能感知”以包括视觉和听觉，进一步增强人工智能与这个经典8位世界的互动。

对不起。

原文

PvP-AI

Creating the Game

PvP-AI is a recreation of an 8-bit game I wrote back in 1990. The only traces left of the original are a few drawings and handwritten notes. Back then, writing for an 8-bit platform, it took every bit of memory and CPU to eke out 4 frames/s with the simplest of animations and backgrounds. Alas, by the time I finished it, the 1990 recession had begun, affecting many industries, including personal computers, so nothing came of it.

A few years ago, a YouTube channel by David Murray The8BitGuy caught my eye, in particular his Commander X16 retro-computer. Looking over the specifications, I realized it might be able to handle a newer incarnation of PvP-AI.

As it turns out, the emulator was able to handle it very nicely, running at almost 8.6 frames/s, with more detail and better AI! Here’s a video of it in action. The actual hardware, though… it turns out there’s a line drawing issue in the VERA module such that certain kinds of lines aren’t rendered correctly, meaning it has to fall back on a slower method. The end result is only 4 frames/s on hardware.

If you want to try it out, you can download the files from my Google Drive. I recommend using the x16-emulator, specifically R49, to run it. More details are in my GitHub EXPLORE repository under CX16 v2 – AI Demo (a.k.a. PvP-AI).

Of note are some peculiarities in gameplay, differentiating it from your typical “shoot-’em-up” 8-bit game:

the arena layout changes based on the highest level reached by any player
missiles have limited intelligence and can navigate around and above walls
if a player brings another player’s health to zero, both players advance in level
only at the final level (level 6) does losing all health eliminate a player from the game
scores are reduced when players have their health brought to zero
scores are increased when players bring another player’s health to zero or eliminate another player
the winner is always the last remaining player, regardless of score
second place goes to the eliminated player with the highest score

It turns out this game lends itself very well to “alternate strategies” one might encounter when integrating with AI — which leads us to…

Integrating with AI

Inspired by other attempts to have LLMs interact with 8-bit systems, most notably ChatGPT vs Atari 2600 Video Chess, I wanted to explore what it takes for an LLM to interact successfully with a simple game.

Unlike approaches that require the LLM to interpret visual or audio output directly, my method uses what I call “smart senses”. These are structured, text-based representations of the game world that abstract away heavy perception tasks. This lets the LLM spend less time deciphering raw data and more time doing what it excels at: reasoning about state and planning actions.

To that end, these were the accommodations to make the game compatible with the LLM:

turn-based play: the live game runs too fast for the LLM; converting to turn-based gives the LLM the time it needs
persistent notes: the LLM is instructed to maintain in-play and after-game notes so that it can carry strategies forward across games
smart sensory input, equivalent to what the 8-bit AI uses:

touch: awareness of contact with missiles, walls, and arena boundaries
EMF detection: awareness of other players’ positions and states

Like other researchers, I’m using the ChatGPT API (model gpt-4o) as the LLM because it offers strong reasoning, stable structured outputs, and affordable per-call pricing. I’m also most familiar with PHP, so I’m using it as the interface layer that connects the LLM to the game. The last missing piece was enabling two-way communication between PHP and the emulator:


	┌───────┐     ┌───────────┐     ┌───┐     ┌───────────┐     ┌────────────┐     ┌──────┐
	│“Cloud”│ <─> │ChatGPT API│ <─> │PHP│ ××× │    ???    │ ××× │x16-emulator│ <─> │PvP-AI│
	└───────┘     └───────────┘     └───┘     └───────────┘     └────────────┘     └──────┘

After some investigation into the existing capabilities of the emulator, I was able to piggyback a new feature, currently a pull request under review, and that completed the “chain”:


	┌───────┐     ┌───────────┐     ┌───┐     ┌───────────┐     ┌────────────┐     ┌──────┐
	│“Cloud”│ <─> │ChatGPT API│ <─> │PHP│ <─> │VIA2-socket│ <─> │x16-emulator│ <─> │PvP-AI│
	└───────┘     └───────────┘     └───┘     └───────────┘     └────────────┘     └──────┘

After enabling on-demand screen captures and removing sound and other non-essentials, I had a working platform to research with. And one final allowance for budgeting reasons: instead of making an API call every frame, I chose to make one every alternate frame.

As part of my investigations, I’ve recorded a series of three sequential games “ChatGPT vs PvP-AI” with persistent notes from game to game. They provide a very interesting arc from experimentation to a winning strategy by the LLM.

Further details about the LLM interface and technical setup are in my GitHub EXPLORE repository under CX16 v3 – LLM vs PvP-AI.

Given the encouraging results, for future research, I’m looking into even more advanced “smart senses” like vision, hearing, and balance.