Add an "ai-assisted" tag or something similar, especially for Resources people make

What’s wrong, and what I do not agree with, is the mentality of “Just give up, it’s not worth it” etc…

It is worth fighting for, it is worth talking about. There are so many places that straight up do not allow AI, and a lot that require a label to be put on if it was used. Why would it be hopeless to do the same on the Godot forums when so many other places already have this?

The question is if it actually works for those other places. AI code checkers have prooven to be unreliable. The Linux kernel for example is used in litterly hundredthousands of servers. If such a lable was effective why would they first try to make it mandatory to then abandoning the idea completely later ?
To me it’s a bit like my earlier employers where one bank thinks it’s a good idea to outsource to India only to withdraw from there a year later and take everything back. Then the next bank does make the same mistake.
If the lable can’t be trusted then what use does it have ? It’s like all those “bio” lables on food. You can’t thrust them either.

So what if the slop wasn’t made by AI at all ? It’s still usless.
It’s like that forrest of dependecies most applications have since quite some time.
I still love this talk by Joe Armstrong. That is so my experience in IT.

Do you think it’s possible for food labels to always be correct and accurate? Do you expect the authorities checking this to notice every incorrect or straight up dangerous food that is sold?
No, of course not. But does that mean we should go “Eh its impossible to tell if the label is correct, why bother making rules that you need to label it as such?”

Let people have the choice, please.
If some people decide they want to avoid AI assisted things, let them do so. Even if the label is not a 100% correct, and people might not use it when they should, it should still be there.

Yup, let people have a choice.

This isn’t a hard issue. We simply want to have a choice.

There is something depressing to me about seeing heavy-AI-use works being posted. For starters, we can often only guess, because the poster has tried to not make it obvious.

Why is this important? Well with AI art for example, there is a part of our brain that empathizes with the artist when looking at a piece of art. The work, the emotion, the love put into it. When there is suspicion about it being AI art (but you’re not quite sure), that part cannot activate. It’s in a troubling in-between state, which is caused by (totally valid) trust issues.

I’m no Christian, but AI art lacks the spark of humanity, most of us can agree:

And I do think that in large-scaling business production, AI coding might be required to keep up with business demands. But with your own game, your own idea, I think it’s important to keep the spark of humanity, even with your own code, because even that has some human value to it.

Whether we like it or not, games are a form of art.

So when we see someone post an impressive coding project on this forum, knowing they put hours of love and tears into it by coding everything themselves, their own systems, their self-designed architecture, is a totally different outlook.

Whether you realize it or not, art is shaped by the context in which it is viewed. That context should be clear.

Easy Agree 100%

itch.io added an genAI tag.

They also had a surge of AI slop that they deleted

and added extra guide lines for AI-assisted slop:

I think it’s time this forum should add a tag. I don’t care if it’s ai-assisted, genAI, or human-made

Is using results from google’s ai overview “ai-assisted”?

Literally? Yes.
Is it what other people mean by AI-assisted most of the time? I think otherwise.
Genuinely it makes no difference either way.
If you make something of bad quality without AI, or something of good quality with AI, it may as well be the same. If you make something good with AI, or something good without AI, it is practically the same result; the only minor difference is people see ‘AI’ and have the villager-sees-ogre deal.

Using an LLM for learning is different than ‘vibe-coding’ (that is what it is called?),
using Google’s AI overview is basically googling it nowadays. An LLM is nothing special, nothing much different than an autocomplete, if you use it, you use it, if you don’t, you don’t.
I would hope this did answer your question, and perhaps any follow-ups, etc.

I am hoping that people don’t view using results from google.com as “ai-assisted”, given how it can be perceived negatively depends on how it is being used in your project. I mean if the definition of “ai-assisted” means from using results from google to actually pay for tokens from whatever.ai site to help write your code/ make assets then it’s not fair for my project to be branded as such.

First of all it wasn’t my choice to go and look for ai to help me and second, I did not ask it to make “my game”, for example last time I googled “can I put functions in array and how to call them randomly” ai overview just gave me an example, I then tested it to see how it works, and then apply it to my code. I feel this is no different than looking up an example from doc or asking a forum. Google ai basically just lookup its searches and list examples for me to try.

Personally I think it’s great because this is what I wish for from official doc; an explanation and example usage of every functions in there. It helps me learn quickly.

But how would people know this if I were to just list my project as “ai-assisted”?? How would they know I am actually “using it for learning codes” instead of “using to make my game”?

The AI overview is just that, AI. It’s no different from an LLM collecting information from several websites and (often incorrectly) summarising it for you. If you use Google Overviews to make your whole Resource that you then sell to other people, then that’s not at all different from just going to ChatGPT and asking it to make the project for you.

There’s nothing wrong with using a search engine to find answers to your problems, but please go to websites and look for answers instead of relying on the Ai overview. It takes away so much money from small websites because nobody visits them anymore, and Google is just scraping all the info off and presenting it with a bunch of inaccuracies.

What we would call this OpenAI marketing ?

https://openai.com/index/gpt-6-astra/?video=1222553704

Is it collaboration when user describe task ?

Support websites by visiting them for sure I would but to be honest google is a search engine and we use it to search for answer. Before all the ai stuff we got algorithms and we gone through first 10-20 links to find the right answer (remember all those tabs we opened?). Now it just gives answer, we check it, it works, and we’re done. This is like why carpet bomb the whole town when we could just laser guided missile a building. Google searching is not gonna be the same unless they start monetizing their ai overview and we go back to weblink algorithms, but do we want to go back?

I guess the line got to be drawn somewhere. It’s hard enough to believe when people say they don’t use ai for their project and it’s even harder now that everytime they google something they are not being “assisted by ai”. Also what happen if forums implement ai for their search? wouldn’t we all then be labeled “ai assisted” too?

Not arguing just something to think about when try to add that tag for resources, like define “ai-assisted” and “whole resource”. Currently based on responds I’d have to list my project as ai-assisted.:pensive_face:

I have not used google search at all in the last 7 years, there are alternatives, LLMs are not mandatory.

You should mark something as ai assisted if you used these ai overviews as your main source of information.

If you have a hard time doing anything without AI help, I recommend doing a few simple tasks without the use of an LLM, so you can learn.

You mentioned how difficult and inconvenient it is to search for an answer or information, but I’m afraid that’s how learning works. You won’t learn much at all if you have everything spoon-fed to you. Struggling and figuring things out one by one is how you learn and grow.

I plan to list websites like forums, youtube, and google..etc as my the main source so at least people can see that I don’t use any of that specific token ai sites like chatgpt or claude for my project.

Also I wouldn’t say it was more difficult and inconvenient to learn “then”, I’d say it is more convenient to learn “now” (given that you verify it too). Like how we have library and now internet if you know what I mean, we’d still need to verify what we read on internet too right?:slightly_smiling_face:

I was searching the net-worth of SK Hynix (The big ram guy) and Gemini showed me that the net-worth was $700B.

(It’s $145B something)

I think the impression that finding some really smart code that was taught by industry experts would be really ruined when you find out that it was all AI generated.

Steam has an AI tag, so perhaps digital assets should too.

There was a big noAI movement amoung artists and musicians, and now coders.

When you have coded solid for several months and discover that other people have just been running an agent - that is dissappointing.

A lot of Coders created NPC’s in game companies using methods like state machines, astar, NN’s, GA’s, Fuzzy logic, and its very ironic and seems out of character to read that programmers are all anti-ai now. This is why the topic seems like it deserves a comment.

Also people might want to see the difference between assets that are AI assisted and assets that are not - how can you tell whether it has been adopted by your peers? The answer is that if noone tags their post with AI then you know that AI has not become the cool popular trendy option and that most people still do stuff for themselves.

Yes a tag is a good idea.

The fact is, some people have a problem with it, so they should tag it everywhere.