RimWorld

RimWorld

This topic has been locked
RimTalk - Warhammer 40K Lore Core
Im working on a more immersiv version of rimtalk for 40k... ChatGpt is providing the programming knowlege im providing the ideas and information chatgpt is working with... but im only one guy so maybe you guys could test the mod and use the document i provided to make this mod as immersive as possible... rimtalk is a great great mod and its an oppertunity to make our 40k colonys acting as 40k colonys
< >
Showing 1-15 of 19 comments
version 0.6.1 alpha dropped today... i still need help testing
No thanks, Rimworld doesn't need AI slop mods; we prefer mods made by actual people that can code properly.
Originally posted by VukieNyx:
No thanks, Rimworld doesn't need AI slop mods; we prefer mods made by actual people that can code properly.

No "we" don't. There's no gating keeping here, anyone can take their idea and turn it into a mod regardless of their skill and we aren't going to mock them because they can't code or draw, because most of us aren't terrible people.

This is not the modding section though, and I wouldn't recommend anyone download anything from untrusted sources unless they know what they are doing. I'm personally not interested in any sort of pawn dialog related mods, too much performance hit for something that doesn't really appeal to me.
Originally posted by Astasia:
Originally posted by VukieNyx:
No thanks, Rimworld doesn't need AI slop mods; we prefer mods made by actual people that can code properly.

No "we" don't. There's no gating keeping here, anyone can take their idea and turn it into a mod regardless of their skill and we aren't going to mock them because they can't code or draw, because most of us aren't terrible people.

This is not the modding section though, and I wouldn't recommend anyone download anything from untrusted sources unless they know what they are doing. I'm personally not interested in any sort of pawn dialog related mods, too much performance hit for something that doesn't really appeal to me.
i personally dont want mods coded by an llm
even if the idea seems cool on paper if the mods code isnt touched up before publishing then it might break something and i dont really want to risk it
plus i also have a general dislike of generative ai so that would discourage me, and others like me, from checking out the mod
now, if you like mods made with genai, then use them, i wont stop you
Originally posted by The Watcher(she/they):
i personally dont want mods coded by an llm
even if the idea seems cool on paper if the mods code isnt touched up before publishing then it might break something and i dont really want to risk it
plus i also have a general dislike of generative ai so that would discourage me, and others like me, from checking out the mod
now, if you like mods made with genai, then use them, i wont stop you

RimWorld mods rarely require significant coding and there's little room to go wrong, the mod either works or it doesn't. Most of the "AI mods" are going to work just fine or with just small little quirks that are easily overlooked. Obviously if it doesn't work well people don't want to use it, but that's hardly limited to AI mods.

That said, Harmony is maintained by AI these days, and that does have a lot of code. Just because it uses AI doesn't mean jank, the vast majority of professional programming happening these days is assisted by AI, it's fairly good at it.
Originally posted by Astasia:
Originally posted by The Watcher(she/they):
i personally dont want mods coded by an llm
even if the idea seems cool on paper if the mods code isnt touched up before publishing then it might break something and i dont really want to risk it
plus i also have a general dislike of generative ai so that would discourage me, and others like me, from checking out the mod
now, if you like mods made with genai, then use them, i wont stop you

RimWorld mods rarely require significant coding and there's little room to go wrong, the mod either works or it doesn't. Most of the "AI mods" are going to work just fine or with just small little quirks that are easily overlooked. Obviously if it doesn't work well people don't want to use it, but that's hardly limited to AI mods.

That said, Harmony is maintained by AI these days, and that does have a lot of code. Just because it uses AI doesn't mean jank, the vast majority of professional programming happening these days is assisted by AI, it's fairly good at it.

By all means, you can use your slop mods, but there is a vast, vast majority that're going to ignore them because of the fact they're slop.

Harmony is manually maintained, so you can stop lying for the sake of your pro-slop argument.
GenAI isn't coding, nor is it artisty, it's slop, and will only ever be slop.
Originally posted by VukieNyx:
Harmony is manually maintained

No it's not. Brrainz very publicly uses AI heavily in his modding and for updating and has specifically stated he uses it to update Harmony because he doesn't have time to do it the slow way. Some people were looking into their own "AI-free branch" but I doubt that is going to happen. This is the guy who released "RimGPT" which he claims was the first AI mod for RimWorld, and has similar functionality as RimTalk which this thread is about. This is also the current mod he is focused on which is all about AI integration with the game:
https://steamcommunity.com/sharedfiles/filedetails/?id=3727949765

Here's something that might blow your mind:
https://www.youtube.com/watch?v=V5yTHKVsHFg

That's a recent video about AI development staring Tynan who is openly discussing his use of AI coding in side projects. His main point is that the hard part of development is figuring out what to make, that's the creativity aspect. Using AI to make it doesn't really take away from the process it just speeds it up and allows for faster iterations and testing.
Remember kids, it can always get worse
Originally posted by k11delta:
version 0.6.1 alpha dropped today... i still need help testing
This is straight up vile, you slopped together something and didn't even bother to test, what if it breaks the game or causes other issues?
Originally posted by Shazbot!:
Remember kids, it can always get worse
Originally posted by k11delta:
version 0.6.1 alpha dropped today... i still need help testing
This is straight up vile, you slopped together something and didn't even bother to test, what if it breaks the game or causes other issues?
They never said they didn't test it.
People have been asking for help testing things they made for a long time. There is only so much you can do on your own, and different modlists or configurations can cause problems you might not even be aware of.
Originally posted by Deadoon:
There is only so much you can do on your own, and different modlists or configurations can cause problems you might not even be aware of.
Oh yeah if only someone would know how it works so one could predict such problems, huh? Since everyone is obsessed with money or other performative nonsense: QA is a paid position, or i guess you can ask AI to test it for you.
Originally posted by Shazbot!:
Originally posted by Deadoon:
There is only so much you can do on your own, and different modlists or configurations can cause problems you might not even be aware of.
Oh yeah if only someone would know how it works so one could predict such problems, huh? Since everyone is obsessed with money or other performative nonsense: QA is a paid position, or i guess you can ask AI to test it for you.
Being a coder doesn't make you omniscient, AI isn't either. Real world testing is almost always required.
You could be a trillion dollar company with unheard of levels of QA, then release an update which passed internal testing only for it to have issues in the real world. That happened with nvidia drivers recently with a few games, including helldivers 2 for example.
Originally posted by Deadoon:
Being a coder doesn't make you omniscient,
That's why the thing you're talking about is called programming, and not "coding", ♥♥♥♥♥♥♥♥ out code isn't hard and even monkey can do it, what you want is working and ideally well thought out solution, which requires more than LLM output can provide(the thinking part).
Originally posted by Deadoon:
Real world testing is almost always required.
Exhaustive testing is impossible, which means you should predict as much as you can and cover possible exceptions.
Originally posted by Deadoon:
You could be a trillion dollar company with unheard of levels of QA, then release an update which passed internal testing only for it to have issues in the real world. That happened with nvidia drivers recently with a few games, including helldivers 2 for example.
Yeah and that's why it just reinforces my point.
Originally posted by Shazbot!:
Originally posted by Deadoon:
Being a coder doesn't make you omniscient,
That's why the thing you're talking about is called programming, and not "coding", ♥♥♥♥♥♥♥♥ out code isn't hard and even monkey can do it, what you want is working and ideally well thought out solution, which requires more than LLM output can provide(the thinking part).
Originally posted by Deadoon:
Real world testing is almost always required.
Exhaustive testing is impossible, which means you should predict as much as you can and cover possible exceptions.
Originally posted by Deadoon:
You could be a trillion dollar company with unheard of levels of QA, then release an update which passed internal testing only for it to have issues in the real world. That happened with nvidia drivers recently with a few games, including helldivers 2 for example.
Yeah and that's why it just reinforces my point.
Coding and programming are synonyms. This has been a thing since programmers filled out coding sheets.
Yeah exhaustive testing is impossible, after all you aren't omniscient which was my previous point.

Your point was that they should test things, which is what they are asking for help with. If even massive companies can't get their stuff perfect with all their money resources and testing, why do you expect a person in the internet to be able to?
Originally posted by Shazbot!:
Remember kids, it can always get worse
Originally posted by k11delta:
version 0.6.1 alpha dropped today... i still need help testing
This is straight up vile, you slopped together something and didn't even bother to test, what if it breaks the game or causes other issues?

I prefer you bring this energy to huge aAa companies who use us as free beta testers, than an individual guy who has creative ideas but can't spend massive amounts of times on coding. Especially for something non-critical as MODS. It's not a godamn critical infrastructure software such as healthcare, energy etc. It's a ♥♥♥♥♥♥♥ game lmao. You could easily make a second installation or backup your current state of the game.

Aslong as there's some effort given to understand the code. It helped me understand coding a lot better than the courses I followed before lol. I can immediately ask what function is for what, and what it's trying to-do. give special commands for documentation and commentary to written functions. It has been an amazing learning tool for me, especially because I can just itterate code fast, and break it on purpose and know why something is breaking.

Originally posted by henk:
I prefer you bring this energy to huge aAa companies who use us as free beta testers, than an individual guy who has creative ideas but can't spend massive amounts of times on coding.
Hey, check this out: https://steamdb.info/stats/releases/ (the part where we have 20k+ new releases per year with nothing of value being amongst them). Have you seen the state of steam store lately(ish) with garbage being everywhere? I personally don't want same low effort garbage mods ♥♥♥♥♥♥♥♥ up everything, that's why i'm "trying" to be prevent it, but let's be real i can't really do anything about so what remains is me shaking my fist at cloud basically.
Originally posted by henk:
Aslong as there's some effort given to understand the code. It helped me understand coding a lot better than the courses I followed before lol. I can immediately ask what function is for what, and what it's trying to-do. give special commands for documentation and commentary to written functions. It has been an amazing learning tool for me, especially because I can just itterate code fast, and break it on purpose and know why something is breaking.
Well yeah there is mild amout of guidance by "AI" and then there's people slopping whole mods without even trying to understand how it works, i'm actually not even against both, it's just slopping whole thing haven't resulted anything worthwhile so far in my experience as i gave a lot of chances to AI stuff.
< >
Showing 1-15 of 19 comments
Per page: 1530 50