Welcome to AI for Game Development! On this series, we’ll be using AI tools to create a totally functional farming game in only 5 days. By the top of this series, you should have learned how you may incorporate quite a lot of AI tools into your game development workflow. I’ll show you ways you need to use AI tools for:
- Art Style
- Game Design
- 3D Assets
- 2D Assets
- Story
Want the fast video version? You possibly can watch it here. Otherwise, should you want the technical details, keep reading!
Note: This tutorial is meant for readers who’re conversant in Unity development and C#. For those who’re recent to those technologies, try the Unity for Beginners series before continuing.
Day 2: Game Design
In Part 1 of this tutorial series, we used AI for Art Style. More specifically, we used Stable Diffusion to generate concept art and develop the visual kind of our game.
On this part, we’ll be using AI for Game Design. In The Short Version, I’ll speak about how I used ChatGPT as a tool to assist develop game ideas. But more importantly, what is definitely occurring here? Keep reading for background on Language Models and their broader Uses in Game Development.
The Short Version
The short version is easy: ask ChatGPT for advice, and follow its advice at your individual discretion. Within the case of the farming game, I asked ChatGPT:
You might be an expert game designer, designing an easy farming game. What features are most vital to creating the farming game fun and fascinating?
The reply given includes (summarized):
- Number of crops
- A difficult and rewarding progression system
- Dynamic and interactive environments
- Social and multiplayer features
- A robust and immersive story or theme
On condition that I only have 5 days, I made a decision to gray-box the primary two points. You possibly can play the result here, and look at the source code here.
I’m not going to enter detail on how I implemented these mechanics, because the focus of this series is tips on how to use AI tools in your individual game development process, not tips on how to implement a farming game. As a substitute, I’ll speak about what ChatGPT is (a language model), how these models actually work, and what this implies for game development.
Language Models
ChatGPT, despite being a serious breakthrough in adoption, is an iteration on tech that has existed for some time: language models.
Language models are a sort of AI which might be trained to predict the likelihood of a sequence of words. For instance, if I were to jot down “The cat chases the ____”, a language model can be trained to predict “mouse”. This training process can then be applied to a wide selection of tasks. For instance, translation: “the French word for cat is ____”. This setup, while successful at some natural language tasks, wasn’t anywhere near the extent of performance seen today. That is, until the introduction of transformers.
Transformers, introduced in 2017, are a neural network architecture that use a self-attention mechanism to predict your entire sequence . That is the tech behind modern language models like ChatGPT. Wish to learn more about how they work? Take a look at our Introduction to Transformers course, available free here on Hugging Face.
So why is ChatGPT so successful in comparison with previous language models? It’s inconceivable to reply this in its entirety, since ChatGPT shouldn’t be open source. Nonetheless, one in all the explanations is Reinforcement Learning from Human Feedback (RLHF), where human feedback is used to enhance the language model. Take a look at this blog post for more information on RLHF: how it really works, open-source tools for doing it, and its future.
This area of AI is continuously changing, and prone to see an explosion of creativity because it becomes a part of the open source community, including in uses for game development. For those who’re reading this, you are probably ahead of the curve already.
Uses in Game Development
In The Short Version, I talked about how I used ChatGPT to assist develop game ideas. There may be rather a lot more you may do with it though, like using it to code a whole game. You should use it for just about anything you may consider. Something that is perhaps a bit more helpful is to speak about what it cannot do.
Limitations
ChatGPT often sounds very convincing, while being incorrect. Here is an archive of ChatGPT failures. The explanation for these is that ChatGPT doesn’t know what it’s talking concerning the way a human does. It’s a really large Language Model that predicts likely outputs, but doesn’t really understand what it’s saying. Certainly one of my personal favorite examples of those failures (especially relevant to game development) is that this explanation of quaternions from Reddit:
This explanation, while sounding excellent, is totally incorrect. That is an important example of why ChatGPT, while very useful, should not be used as a definitive knowledge base.
Suggestions
If ChatGPT fails rather a lot, must you use it? I might argue that it’s still extremely useful as a tool, somewhat than as a substitute. In the instance of Game Design, I could have followed up on ChatGPT’s answer, and asked it to implement all of its suggestions for me. As I discussed before, others have done this, and it somewhat works. Nonetheless, I might suggest using ChatGPT more as a tool for brainstorming and acceleration, somewhat than as a whole substitute for steps in the event process.
Click here to read Part 3, where we use AI for 3D Assets.
