ChatGPT proposes and creates a game in love2d ( artificial intelligence scares me )

[glitchapp]

That may be true, but suppose that the generated code violates copyright laws. It is not nice to copy and paste code without providing attribution or credit to the original author.

I do not agree with that (with chatGPT not being able to provide sources) and I wish I could prompt it to provide sources, but I don't think stopping using it will make any difference. The generated code is also simple enough I could get to the same results (and may will) if I spend enough time learning shaders or any other field in which I used chatGPT to solve the problems. Actually you need to read code which usually comes from many sources and licenses in order to learn a new language or acquire a new skills and you do not provide the sources of everything you used to learn that new skill (you won't even probably remember).

[darkfrei]

I've asked it to find my gaps in Lua by giving me small exercises and checking if it was right solution. Is the generated exercise under license and my learned skills too?

If I create some content, all rights for it are mine. For example I've made a maze and it's unique. What if someone makes maze generator, by looking to my maze?
Is this maze generator secondary as based on this maze or it is a new product, that can makes all mazes, theoretically my maze too?

[ivan]

@glitchapp Without the appropriate sources information has little to no value.

@darkfrei Yes, you can make copyright claims for almost any creative work that you publish online.
Regarding to the maze question: I am not a lawyer but from what I understand copyright deals with creative works like music, art, books, articles and so on. You may be able to attain a copyright for your maze if it features original and creative artwork, but I do not know if you can copyright the maze layout itself.

[glitchapp]

@glitchapp Without the appropriate sources information has little to no value.

@darkfrei Yes, you can make copyright claims for almost any creative work that you publish online.
Regarding to the maze question: I am not a lawyer but from what I understand copyright deals with creative works like music, art, books, articles and so on. You may be able to attain a copyright for your maze if it features original and creative artwork, but I do not know if you can copyright the maze layout itself.

The value of what? of the things achieved? the code created and shared? the fun and acomplishement and knowledge acquired? who determines that? the market value? the job market? can we measure all of that? can we standarize those measures?

[darkfrei]

The unique maze is the same as unique book or article, I can think about it as about text. Or it can be drawn and it's actually a picture, that is also under the copyright laws.
No feature is needed, the same way, as no text and no picture need special features to be unique and to be protected with copyright.

[ivan]

By "value" I mean how useful is the information. Without any sources you do not know how truthful or how accurate is the information. If you are not 100% sure that you have the legal rights to use such information then it is not very useful or valuable from a legal standpoint.

[glitchapp]

By "value" I mean how useful is the information. Without any sources you do not know how truthful or how accurate is the information. If you are not 100% sure that you have the legal rights to use such information then it is not very useful or valuable from a legal standpoint.

I have not much idea about copyright laws but I think there are limits to what can be copyrighted and this case we are talking about a simple algorithm that renders a blue noise dithering shader, there are thousands of ways to accomplish this and I'm quite sceptical that such simple things can be copyrighted but maybe someone knows better than I about it.

[ivan]

@darkfrei I am not an attorney but from what I understand copyright applies to original, creative works. So if your maze is laid out in the shape of the Adidas or Nike logo then that would not be considered an original work.

[ivan]

I'm quite sceptical that such simple things can be copyrighted but maybe someone knows better than I about it.

Did ChatGPT collect and use the source information lawfully? Without specifying its sources we do not know and that is the core of the issue. Even if ChatGPT has transformed or altered the source data, the AI's answers are not original and NOT based on its own experience.

[glitchapp]

I'm quite sceptical that such simple things can be copyrighted but maybe someone knows better than I about it.

Did ChatGPT collect and use the source information lawfully? Without any sources we do not know and that is the core of the issue. Even if ChatGPT has transformed or altered the source data, the AI's answers are not original and NOT based on its own experience.

I don't want to turn this into a political or philosophical topic, but nothing created by humans or A.I. is new. Everything is based on acquired information and it's true that humans and machines acquires the information very differently, but at the end the results are not very different. Stable diffusion is capable of mimic art styles just as we do. As long as you do not try to impersonate the author and the composition is different, it's a new piece of art no matter if based on experience or a database.