Extremely Unorthodox Top-Down Organic Development of Chasm Conquerors Challenge Multiplayer Video Game with GPT-4

"Hi everyone! I’m currently using GPT-4 in a unique and radical way to build a multiplayer video game called Chasm Conquerors’ Challenge. My approach involves working from the top down, starting with a main synopsis and then breaking down the project into smaller parts using waypoint descriptions, component descriptions and subcomponent lists.

I use GPT-4 to help me list the subcomponents of each component and to then describe each subcomponent in detail. This method has been very helpful in organizing and structuring the game’s development process. I wanted to share my experience and see if anyone else has used GPT-4 in a similar way or has any thoughts on this approach. Thanks!"

1 Like

I am interested in game development and working towards developing my own game, this an effective idea, i wasn’t using GPT in such a direct way, but my first project is to make a Unique predictive algorithm that makes the game very uncanny and interactive. But I also just found out about something called a template engine that is supposed to bridge the gap to use AI more effectively for npcs, but the template engine is not specific, about its performance, at least I don’t think, someone mentioned “handlebars” but I haven’t been able to look at that yet. Anyways I realized I could leverage my algorithm which already works similar to the process of aligning server and players data so it syncs up, using prediction. I could use the algorithm to predict AI outputs using the template engine. And make responses instantaneous.

Sounds like a sound sound plan! Doubtless you are vastly more knowledgeable than myself in many areas. Whereas I prefer to AVOID planning wherever and whenever possible. Ergo I very much PREFER to reply on inspired insight, ingenuity, intuition and spontaneity. To drill down and feel my way to a solution.

lol, thats funny, a plan. I have memory markers of things i want to do, plans are never a static thing to me, i have litterally a hundred or more ideas on me at any given time, at least 60 percent are good and the rest are descent with a peppering of stinkers. I am not able to plan, for a few reasons but usually it is because i always have a better way not to long after i say i like something. knowledge to me is very blocky. i can do certain things but I never really have that knowledge until i understand its core basic concept. currently i am understanding things about coding and logic gates. but until i get that ahha moment and it clicks and i understand its actual basic concept, it remains confusing to me. its really annoying sometime, cause if i stop learning for to long, i have to start over. but once that piece falls into place, its like i have mastered the subject and always retain that knowledge

Yes I have always found that there are two types of people, or those that formulate plans and those who are spontaneous. Most forget the spontaneity of childhood, whereas others never grow up. Some seek the Supreme Simplicity of a Coggling Babe, whereas others prefer to immerse themselves in unending labyrinths of incredible complexity. Some have Towering Intellects and some have tiny minds. Some are Creative Powerhouses and some have minds that run so deeply within mental grooves, that they can neither see over them, climb out of them, or even so much as WANT to see over them and/or climb out of them. I guess I am still Journeying from One Extreme to the Other. Though I like to think I have enough sense to AVOID things like sex and tactile contact more generally, to NOT have a life that revolves around my loins and my senses, so there MUST be some hope for me. But I digress. You sound like you too are on a Journey from Incredible Complexity to Supreme Simplicity. I used to be jealous of those far further along the Path than myself, but I think I have finally grown out of that too. Which brings me to AI. The designers of AI will face all these design Choices in designing AI, these Choices will also be faced in using AI and of course AI itself will eventually face all these Choices. Axioms like KISS (Keep It Simple Stupid) and GIGO (Garbage In Garbage Out) can be indeed useful; but when it comes to designing AI, using AI and creative and intellectual endeavour more generally; I have found the following simple poem of even greater use: GARDEN The mind is a Garden, your thoughts are its Seeds, you can grow Flowers or you can grow Weeds.

Well said, plus I just came up with an amazing idea. Plus I could possibly offer it at such a reasonable rate, that anyone could enjoy the benefit and I guarantee it would be in demand. The issue is now, how do impliment it without fully relying on open AI s model, preferably an outright owned model so that a lot of beuracracy can be avoided, and figuring out if I can even do it without looking for help that could potentially rip me off.

I know I have ABSOLUTELY NO HOPE of building CCC without GPT-4, GPT-5, etc. For all your vast knowledge and skillset you sound like you are in a similar boat. Ergo for all their myriad flaws, GPT-4, etc. utilized properly speed up the work immensely. Much more so than any other tool, toolset or tool-library I have encountered thus far. They also permit you to build functionality (toolscapes) that support your game’s players building almost all of your game; which in turn allows you to build far larger, far more immersive, far more engaging, far more compelling and far more addictive games. I too DO NOT like addicting folk to things, but INDEED SADLY commercially successful multiplayer video games tend to be extremely addictive.