Skip to main content

Indie game storeFree gamesFun gamesHorror games
Game developmentAssetsComics
SalesBundles
Jobs
TagsGame Engines

I am making an AI, that will be released hopefully shorty, I don't know when release will be, but just hope you arelooking forward to it. (you will be looking forward to it, it will be 100% free!)

A topic by Arnego Gaming Studios created 5 days ago Views: 318 Replies: 37
Viewing posts 1 to 7

WSP? BTW, this is not a game advertisement, it is a news article telling you guys about the upcoming tool I am making. So it abides by the rules.

Moderator

Then it should probably go in Devlogs.

Well, I have no game to put the devlog in, but I will put future updates to the game in devlogs when it is released.

Moderator

…I just sent you a link to a forum category. You can also make unattached blog posts.

(+1)

Hi.

Can you talk more about it? what is it? what is does? what it is for? come on, do some marketing haha.

Well... It will be capable of alot of things, including coding, generating photos, videos, songs, etc... it can help with anything you need, or it can be used for just chatting. And, unlike other AI platforms where they say it is 100% free, and you have to pay to get the whole version anyways, this in fact will be 100% free, so I hope you are looking forward to this.

Marketing?🤣 I am just building the tension with this, getting people to lean in to hear more, I am actually great at marketing, have you checked out my sandbox game, Fabricatio? I have several Devlogs marketing that game, and they are in great detail.

Looking forward to it, does it have a name? I want to constantly google it till you release it😅

(+1)

🤣🤣🤣🤣🤣🤣 Well, it is going to be called RAI, which stands for Revolutionary Artificial Intelligence, I will keep posting about it until release, and I may even post a simple demo version that you can only chat with, no extra abilities, because those extra abilities prove a challenge, but yeah, I WILL be adding those extra feature, including generating pictures, videos, music/sound effects, coding, editing anything I just previously mentioned, planning things, searching the web, and more! And the best part, it is all for free! (As I mentioned like 100 times now) I'm glad someone is looking forward to my project, and have you tried Fabricatio? That is a sandbox game by me, and it is having a huge update on September 19! You should go check it out!

U got me interested XD. Are you planning to host this anywhere else besides itch.io, or is this an exclusive release?

Well, I am releasing it on itch.io, then I will make it accessible on microStudio, and anywhere else that I can get it up for free.

Awesome looking forward to it!

Check out Fabricatio, my sandbox game! It will be receiving a massive update on September 19!

already did

Did you like it?

its  a bit laggy but its still great!

It's not laggy on my computer... It might be a problem with your computer, but I'm glad you still liked it!

Is this a local system or an interface to other services?

Because you will not create an "ai" from scratch as a project. So are you adapting free model for local usage or is this yet another client interface. Or even one of those ai agents?

I know exactly what I am doing, there is a backend thingy majig, but all the information will be reveled after release.

(1 edit)

I am not sure if Ai topics or Ai projects are allowed here but I like to ask something! Since you might be a expert about what you are working on, would it be possible for you to create a ai that converts a image of a 3d model into a actual 3d model? I used Google Flow to create a bunch of pictures. they are in jpg format but still might be useful reference to help create that model I wanted for better accuracy if that might work with your model? I also need a Ai for creating png sprites instead of jpg that can be useful for editing on any paint program! I only use Ai for complicated 3d spriting and 3d modeling that require old outdated 3d software that our modern computers can’t run such as Silicone Graphics Workstation and the 2nd oldest version of Maya which I tried to install on my Windows 11 and doesn’t work!

The closest thing I could get is Anim8or from 1999. I did manage to get 3D Studio from 1996 to work but its confusing interface and design is not really meant for beginners at the time I guess. I like a old 3d software with simple interface and it must be a tool from that time in the mid or late 90s!

I grow up with Donkey Kong Country and played Rare’s other 3d rendered games such as Killer Instinct, Conker’s Pocket Tales, and the recently leaked Dream from couple days ago that went viral all over social media and youtube! I only wish we have 3d software just like Rare used with their games that make their 3d modeling artwork look so iconic like Donkey Kong Country, Banjo Kazooie, Conker, Killer Instinct, Diddy Kong Racing, Blast Corps, and Jet Force Gemini to name a few! They also used the same modeling tool for Dream as well!

Another great reason for using Ai is to restore and recreate assets taken from screenshots and video of prototypes never been released yet especially Earthbound 64 (Mother 3 on the Nintendo 64), Conker Twelve Tales, and Kirby Air Ride 64!

Hi. I know you didn't as me but I know some tools that may help. But perhaps i should warn that those tools are for local use.

But for doing a img2mesh there are huyuan3d based worlflows, not really game-ready in my opinion but maybe can help. Or 3D gen studio, but i never used it, so i can't talk about it. 

For creating art in PNG format I'd recommend Krita with the plugin Krita  AI Diffusion since you can edit the image while you made them and also can made the background transparent by just selecting and erasing it.

Another tool i recommend to background removal is RMGB-2-Studio, since is just drag and drop and also allows to process a bunch of files at once.

Hope may help. :-)

Not really much Im afraid. You see Im looking for a ai that could img2img by converting a jpg blurry image into a clear edit friendly and reuseable png format that we can easily custom with on ms painr, graphics gale, and such!

As for 3d models what you suggest I might look it up but then again Im afraid all I got is Blender and Anim8or.

(1 edit)

Not sure if i am understanding. 

Perhaps you are referring to upscaling? that can be done in Krita AI Diffusion as well. Or an AI that does everything? So far i didn't saw such thing, but what you want can be done by using different tools in a workflow. 

Or perhaps you are referring to use something like Dall-E3 or Nano Banana to turn a draft into a image or spritesheet and then using a background removal?  Blender and Anim8or are good. In fact, you may need them for retopo and animation, and some cases even texturing since many img2mesh models don't texture or made a hellish UV map.

Taking about Rare's style I think common tools used back them were softImage and RenderWare, but i can't say for sure since back then was pretty common for studios to create their own engines and tools from scratch. Today the most close thing i can think that could help to create similar aesthetic are 001-Game Creator, AGE for 2D and the CopperCube for 3D, wicht  is powered by a really old version of Irrlich, a competitor of RenderWare. Maybe combining those with a 'old school' style lora may help. 

Yes! That is actually something I forgot to mention, cause me and my siblings love making 3D pictures out of stuff we drew, but the AI used needs to be paid for. Oh wait, you ment actually turing it into a 3D model, instead of using something like tinkercad to do all the 3D modeling manually, didn't you? Even so, of course! I will happily add that to the program if it makes the users happy.

are u making the ai by scratch or using an api key?

Neither, I am using Qwen, so I don't need to pay, or even set up anything besides the code, so I am not doing either. But the whole thing is designed by me.

(+1)

I get that setting up an open-source model locally takes development work. But in tech terms, "making an AI" implies you actually built or trained the underlying model itself. Using a pre-trained model like Qwen means you're integrating an existing tool, not creating the AI from scratch.

Ok, but I'm using a backend version of Qwen, one that is used for making AI, just, go check out Fabricatio, it updated today.

Checking out Fabricatio doesn't change the definitions. Qwen is a pre-trained LLM developed by Alibaba. Whether you run it via an API or host a "backend version" locally to generate your code, you didn't create the AI. You're implementing a pre-existing tool.

Maybe I didn't create the backend, but I am giving this AI it's personality and abilities, and useable for free.  And that's all people care about. 

(+1)

Well u didnt say that u said and i quote "I am making an ai" Why didnt u say that before i asked this question💀 Exactly, customizing prompts and giving an existing model a 'personality' is totally different from building or training an AI yourself. You're just hosting a free wrapper for Qwen, not creating a brand new AI.

Ok, I didn't know that I just left something out, I'm sorry, that one is actually 100% on me.

Will you go check out the new update on Fabricatio?

no i cant the game lags 

oh.