I wrote a blog post on the current trend of game decompilations using AI and where we’re going with this. It’s all very promising but oh boy are the implications scary.
I wrote a blog post on the current trend of game decompilations using AI and where we’re going with this. It’s all very promising but oh boy are the implications scary.
What’s the motivation for even doing this if not to learn? If it’s just to have a playable version of the game, emulation covers that easily. If you want to extend it, you still kind of need to know how the original is designed to understand what types of changes fit within the structure of the game.
Quenzar’s Caverns is such a simple game, re-implementing it is a great project for someone wanting to learn how a graphical application is put together end-to-end. It would even be a good beginner reverse-engineering project if you want to reproduce it precisely, given it’s a 16-bit application.
At the end of the week, you might have a reproduction of the game that you can play. But you’ll have missed out on the opportunity to learn enough about program design to understand why throwing a week’s worth of compute at this task is totally unnecessary.
The goal is to establish a workflow of working with AI, the programs, the behaviors of the models, and so forth. In any case, I always sucked at math, so learning how to program just isn’t viable for me. My strong points lie elsewhere.
My goals are not yours. I want to use AI to translate videogames and other media, reimplement games with modern QOL, and eventually, to have media of my own imagination given form.