AI & Factorio Mods (Maglev dreams)

[BHakluyt]

How is the mod scene going with AI these days? I have a dream for a super sleek and fast elevated Maglev rail system. My internet is unfortunately too slow to get anywhere close to doing it myself...

See some inspiration here:

image

Maglev mockup

1782029362501.png (2.12 MiB) Viewed 900 times

Couldn't get this one to run: (AI made mod)

maglev-trains_1.0.0.zip

AI made mod for maglev

(35.31 KiB) Downloaded 36 times

[mmmPI]

there are various point of view , sometimes very opposite about AIs in general, and that is also present in Factorio. I'm not a person who is going to hate on AI , i try to have a rationnal approach, still i think for Factorio modding in particular, there isn't a real good AI i know that can do "everything" as you seem to wish. I tried using AIs to make mods, but i found it better to use them to understand how to make a mod , rather than to make the mod themselves, they don't have good reference on what is availble in factorio API, they make mistake between 1.0 1.1 2.0 space age and overall they are pretty unreliable for task like "can i do this ? or how do i do this ? or can you do this for me ", rather, you can ask "what does this do ?" and giving snippets of code you don't understand ( respecting the license ) , you will have an explanation, which is very often correct, it allows to spot mistakes in what you did, you can also ask utilities stuff, like i have a list and i want to remove only the elements that are also on this other list unless they are present on this 3rd list and i need to do that on those 4 files everywhere needed, AI speed-up stuff with less risk of mistake if you don't rely on them having a perfect, nor even good "game internals knowledge" and follow their own hypothesis.
You can try to "reason" depending on the quality of the AI model you use, try to have a discussion where you explain the AI your goal, or intermediate goal, or tiny step, ask it how it is done by people who know what they are doing, and why ,then it's easier to further ask precisely what you want the AI to do, instead of "the whole mod", you ask for smaller parts that you know you need because of the previous dicussion, where you try to spot if there is inconsistency, like the AI telling you A is the better method then giving intruction for how to make B; and you test if those intermediate step produce the expected result, that the AI doesn't fallback on a proposition C that look like it does the job but not really. If you don't ask it to explain to you what the code is actually doing it's impossible i think for such project in 2026.

There are more expensive AI agents than those i tried, but when i try to document on how it works, i am not convinced there is yet a stage where you can ask for "such a complex mod" to be done reliaby "at the end" without much more intermediate validations. An AI agent as far as i understand very much risk going to otherwise bash his money longer on problems like version inconsistency, never understanding the root cause and developping many alternative wrong theory and test them over and over. Unlike some other development , a factorio mod often time needs to be reloaded and tested humanly to know if the incremental changes and addition by the AI are working.

I feel your "best bet" is to take a month or so for it minimum, and validate a plan for each week, like 1rst one you load the mod and the technology and the icons and the tech, but it's just a placeholder non animated graphic, the week after because it takes time to do the back and forth with the AI, adding a thing, correcting the mistakes, learning in the way so you don't need to always reload the game to spot them in the AI code, you still learn the langage, the data structure the API and so on. I think for the graphics themselves you need no less than a week, and it probably will still look like placeholder, but undertanding which pieces are needed, getting the format right the dimension the proper set you need for a full animated trains, and making mock up version is already going to take time and require intermediate human verification, you will most likely need to correct the AI based on what you visually observe is going wrong, that's more the wiring parts, you need also at least a week for this, the animation in game and the tweaks and corrections for the graphics to be actually displayed as a functionnal train and not just there somewhere.

And you also need a week at least for the polishing, making sure it doesn't break with the other mod you like to use, make sure the performance arent terrible because of leftover scaffold or debug code, typos, stuff you left as "to do" making the mod page on the portal, and for such a mod , i expect also dealing with the early feedback for such a mod, people always find edge cases, some you may want to solve.

Obviously i'm not an experienced modder, i tried to touch things here and there, and if AI were better at helping or i had developped a better method of using them i could have touched some more, i felt the limitations, i'm not doing a generic "AI bad argument" here, i'm saying it's actually much more work than may be anticipated from the fallacious promises of companies selling chatbots ^^ When i say "1 week per stage" it's ambitious and approximation, it's more like "at least a week" or so in my mind, and obviously depending on how much time you can spend and if you are a programmer already, or if you have a background that allows you to learn and remember fast the new concepts. Also Discord can be quite useful, like if you really can't find the answer to a specific stuff even with AI, go ask real humans, but you need to narrow down the problem better than "i asked chagpt once and it didn't immediatly work" , otherwise you're asking someone to review the full mode and finish what could be 95% of the work (and real humans rarely do that x))

[BHakluyt]

Thank you for the detailed feedback. I have internet for a few hours about once a week so its extremely difficult for me to do it as you described above. I really wish someone with more internet time and the same burning desire for a maglev can help out.

I've found Gemini and Claude can indeed be very helpful to create a mod with minimal know how from my side. I know how to open a lua in notepad + and thats about it. Still fixing the bugs but I got two mods of the same idea so far. Would anyone like to try it out? The idea is to have a new and additional game loop since Space Age killed the end game for me.

quantum AI entanglement

1782363951322.png (2.05 MiB) Viewed 701 times

quantum survival

1782414933185.png (1.91 MiB) Viewed 701 times

[mmmPI]

Without a lot of internet it will probably not possible to use AIs much tough.

there exist "maglev mod" not sure how this one is https://mods.factorio.com/mod/maglev-straightaways, but i have used those electric trains as part of the set from Moshine recently : https://mods.factorio.com/mod/se-space-trains and there exist other trains mods that "ressemble" the maglev idea : https://mods.factorio.com/mod/Space_trains https://mods.factorio.com/mod/electric-trains

All of the license are permissive, you can study the mods to understand how it's done all you want offline, and keep the part you don't understand , for that time you can ask an AI x).

Also , although you can technically use notepad + , you would probably much prefer a more "programming" oriented software to review the code, to have the syntax well colored for readability at least.

[BHakluyt]

Look, "F AI" and all that yeah (would love to do the coding offline on the laptop but that struggles, responsible data centre usage in mind) and all that but I would never realistically get so far ever and learned a bunch of how the game works in the process. After a weekend of pasting code back and forth between the AI and the computer, we are finally in the game! All the settings show up, great! The space ship crashes! The warnings are given and fate is accepted. Here we go...

10min in and the mod is working! Shall I play it until endgame or anyone want to test it from the get go?

Quantum Survival horror where you need to transport the ore very quickly or it spoils, so maglevs will be needed and probably take another 3 years or another good off weekend. Early days...

Oh yeah its for Factorio 2.0 still, haven't tried the updated 2.1 yet.

N00b techy stuff

If you are interested how to do this yourself for free, I might be willing to give out some tips. Gemini consistently code things for Factorio 1.0 and then has to fix it. Some icons won't work because the internet just doesn't have the right paths, so using the ingame assets might need some manual easy verifying. The mod is supposed to use only in-game assets with a tint applied. Ultimate tip is to flesh out a design document first and feed it through a few AIs and the one you want to use by asking it to look through the whole document and look for operational and design errors and to polish it for you and expand missing details like recipes. It is better to give a detailed initial design than adding features or things later on.

So see this post as documentation of my early days prodding with AI and modding...

quantum survival

1782665573394.png (2.01 MiB) Viewed 596 times

Let me know if you'd like to test it and I will upload it here...

[mmmPI]

Look, "F AI" and all that yeah

Unfortunatly it's often that people stop at that x). I think though it can give a confidence boost to try new things and even sometimes an actual help , and if you like the thing and/or ask the AI to teach you instead of doing it for you, i feel there will be plenty things where you won't need the AI anymore too, it should be even, otherwise you could feel paralized or dependant on the AI for understanding your project.

10min in and the mod is working! Shall I play it until endgame or anyone want to test it from the get go?

Quantum Survival horror where you need to transport the ore very quickly or it spoils, so maglevs will be needed and probably take another 3 years or another good off weekend. Early days...

You made a working mod that launch , where ore can spoil ? If so , i think you should post it, or even on the portal, there is a "license" to choose, which can take a bit of offline thinking ^^ also that part where you think of a description, if you want to add a disclaimer about the use of AI ^^ , or make sure the changelog appears like you want , that the thumbnail and other images are ok , if you want feedback i feel that can only help differentiate it from mods that are more personnal stuff. Plus the mods sounds fun, maybe a bit too masochist for my taste but i guess i depend the spoilage result and the timing, and if there are parameters, that's also a good side of posting on the portal, you can get ideas or suggestions

If you update it to 2.1 i will give a quick try alongside mine when i update them

[BHakluyt]

Yes! I'd really appreciate it if anyone can have a look, the UPS is all of a sudden down the drain, I think its the control.lua thats a bit of a mess and perhaps the HUD or UI is causing the UPS issue. Otherwise all the entities ore spawn and stuff research etc, so if the UPS issue can be fixed I think it will work great for the most part! Please anyone if you have a bit of time and some know how please please. Its unfortunately also still for 2.0 Space Age.

Quantum Survival for 2.0

quantum-survival.rar

(1.49 MiB) Downloaded 37 times

[BHakluyt]

Wow I see a few downloaded the mod. That version is garbage, I got it fixed and got the code cleaned up. I'll do a bit of more offline testing and then come paste the full mod here again. Stay tuned.

Its also expanded, now you need to have the quantum stability stabalised, maintain hunger, thirst and oxygen levels for the player. Hard-core suvival coming soon...

[mmmPI]

Wow I see a few downloaded the mod. That version is garbage, I got it fixed and got the code cleaned up. I'll do a bit of more offline testing and then come paste the full mod here again. Stay tuned.

Hey it can also be just robots that download to try and learn stuff, typically those AI we use they are trained with stuff crawled from the web, so in the download there may be a couple of non-real ones .

I can say at least one of the download was from me ! I tried to put it to version 2.1 naively but no it didn't work right away. I couldn't really test the mod then, i just looked at the code and it looks more complicated than "just" adding a different type of trains. The later task i thought would be "easier" because the perfomance and things to do would have been "copy" and "adaptation" from the base game.( a lot more pieces than just adding technology or receipe but still following existing code). But what you did is more "complicated" it involves more systems, some brand new. The performance impact "has to be" on one of those i imagine. But you may have already fixed a few things, from what i read. The AI can help providing code that can be used to test perfomance to some extend,to add in the mod, or let you know which debug already exist in the game and how to access them when it is running, but this is getting outside of the things i'm familiar with, i never had to design my own tools for that.

I think it would be interesting to have pictures of how it looks in game, i see on the code the hunger "bar" , i wonder how it looks, concept art are not representative enough to my taste, but i know other modders use them on the portal, it is really a matter of "taste" i feel. Now if you're excited to see a few downloads, you should definitely make it works for 2.1 and publish on the portal once it is playable. Maybe 2 weeks ? 3 weeks ? Hopefully you are getting closer to the dream maglev mod !

[BHakluyt]

I hear it's called vibe coding, and what a horror its been indeed. Been taking like a month or more and it has evolved into a monstrosity, mostly because everytime I ask the AI to fix a bug, I add a new feature or two also. Lol.

So far what we got, the premise:
Hunger, Thirst, Oxygen and DNA Stability (Ageing sort off)
Decay: plates and gears etc rust and machines wear down and require maintenance.

Trying to get blue tinted 30% size of the demolisher worms on fulgora has been struggling, but fulgora and aquilo need some flying drones to just add some challenge.

I also figured how to get some AI to make nice sprites so now I have some real hope of making the maglev train mod sometime.

[eugenekay]

How is the mod scene going with AI these days?

Look, "F AI" and all that yeah

Yes! I'd really appreciate it if anyone can have a look, the UPS is all of a sudden down the drain

I hear it's called vibe coding, and what a horror its been indeed. Been taking like a month or more and it has evolved into a monstrosity, mostly because everytime I ask the AI to fix a bug, I add a new feature or two also.

I think this tells the story of how the "AI mod scene" is going pretty well.

[BHakluyt]

I think this tells the story of how the "AI mod scene" is going pretty well.
[/quote]

Exactly! Haha! Documenting for those wondering it seems...

[mmmPI]

I think this tells the story of how the "AI mod scene" is going pretty well.

It may also be a confirmation biais, if i recall correctly you're "not a fan" of those machines and thus could be more prone to see only negative interpretations when reading reported statement

Last time i used AI for a mod, someone had told me on this forum that it was "simple" what i wanted to do and indicated me some methods. I wanted to add a click sound to the hotkey that changes the music. When i had looked at the API i thought it would be difficult, cuz i'm a noob. I didn't know about the methods that i was indicated, not that they existed , not that they would solve my problem, nor how to use them.

But i asked the basic free chatgpt what it means, i asked more information to understand what they do and how it works, and then i asked it to use those method to not only add the click sound because that part worked after a few tests, but also to print the name of the music in the console. It told me it cannot do it, but i insisted saying that the name of the music played are readable in the debug, and that surely there must be a way to just print them in the console, which made it reconsider its answer and eventually provided 2 working version of the mod, with slightly different methods of detecting if the sound has changed so i could see for myself in game if there was a performance difference. Then i noticed it didn't work in /editor mode, when the time is paused, which i suspected when i saw the code, so i asked for confirmation, and wether there was way around it which lead to different versions, eventually i stopped without all the fancy i wanted like a different sound when the song repeats but still more than what i thought i would be able to do. And it works in pause mode

The person who told me the solutions don't need chatgpt to make such mods ^^, and if using chatgpt, could have had the full mod on the first try by describing much more precisely the task and how to achieve the result i imagine, but me, without chatgpt, i wouldn't have had the stuff done as fast. I may not have even know where to start, you must know that feeling too, where it's much easier to correct what's wrong in the AI than coming up with the initial draw from your ideas yourself, like you try to describe a tatoo or your future kitchen or something else as precisely as it is in your head to someone else and you add details or clarifications as needed for it be perfect when you see the drawings along the way. I would have been required to ask many more question for a "simple" task because i was missing lots of prerequisite.

Now i understand why it wasn't that hard and i can use the mod , both makes me happy, and wanting to share.

I hear it's called vibe coding, and what a horror its been indeed. Been taking like a month or more and it has evolved into a monstrosity, mostly because everytime I ask the AI to fix a bug, I add a new feature or two also. Lol.

I tried to mention in my previous explanations that having background in programming helps a lot, hehe, here, i think i recognize the mistake of an inexperienced coder like me, ( sorry) but adding feature when fixing bug is probably a thing to avoid no matter the use of AI x). With the proper method some people can get significant results with vibe coding : https://www.scientificamerican.com/arti ... d-problem/ [EDIT, article without paywall] . I guess it's a mix of ideas, perseverance, skill , luck , DIY fiddling and a bit of magic that makes it work eventually if it does.

So far what we got, the premise:
Hunger, Thirst, Oxygen and DNA Stability (Ageing sort off)
Decay: plates and gears etc rust and machines wear down and require maintenance.

Trying to get blue tinted 30% size of the demolisher worms on fulgora has been struggling, but fulgora and aquilo need some flying drones to just add some challenge.

That has nothing to do with the Maglev dreams but i can totally understand how it can ends up like that It requires messing with more different notions imo than "just" "adding a second set of train + rails with custom receipe and visuals". The maglev dream sounds easier to me because it would be a lot of duplicating code. Whereas for new runtime system you risk impacting performance more. That's beyond what i can help for already though x).

I also figured how to get some AI to make nice sprites so now I have some real hope of making the maglev train mod sometime.

I have a spend a lot of time doing sprites with AI but "animated" ones for factorio i felt were never satisfying. I have tried using AI to make 3D model to make the spritesheet myself following this tutorial => https://app.notion.com/p/3D-model-to-Fa ... 76c023040f , but the result were not as good when i did it , i can only pass the tutorial and ask how you're doing it ? ^^