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

[ivan]

No! Most of my experiences and sensory information are entirely new and not based on previously acquired data. It is possible that some people's ideas are entirely original too and that is why we have copyright.

[pgimeno]

At least where I live, obtaining copyrighted works without paying for them is not illegal; what is illegal is distributing them. If that's the same in the countries that ChatGPT gathered information from during its training, then I don't think that building the database was illegal. But it's quite possible that it emits copyrighted work, and that's where it would be violating copyright laws.

[ivan]

Good interjection by Pgimeno.
I would also add that we are all open source proponents around here (Love2d is an open source project) whereas ChatGPT is closed source and sponsored by a big corporation.
TLDR: do not trust ChatGPT or the media hype around it

[glitchapp]

Good interjection by Pgimeno.
I would also add that we are all open source proponents around here (Love2d is an open source project) whereas ChatGPT is closed source and sponsored by a big corporation.
TLDR: do not trust ChatGPT or the media hype around it

If a software makes a shader or solve whatever problems and the solution works, you can trust it to solve specific problems. That does not mean that you trust the creators or the company backing it at all. I believe people get too emotional when talking about this topic and may tend to make false associations.

Regarding the open source proponents many people contribute to open source and also create closed source programs to afford a living or because that is their profession, one thing does not make the other incompatible.

[ivan]

Yes, we all support the advancement of AI software, but it needs to be done in accordance with the law. If there are any legal doubts with regards to answers generated by ChatGPT then do not trust them. Your argument "if ... the solution works, you can trust it" will not be acceptable in any courtroom.

[glitchapp]

Yes, we all support the advancement of AI software, but it needs to be done in accordance with the law. If there are any legal doubts with regards to answers generated by ChatGPT then do not trust them. Your argument "if ... the solution works, you can trust it" will not be compensable in any courtroom.

Laws have been created to protect the interest of the ruling class and they are no measure of what is right or wrong. Many atrocities are made within the law. Copyright laws were created to protect closed software, and then open source license were created to use those laws to the opposite goal.

What is wrong is using A.I. for the wrong goals and I believe much of those wrong uses of A.I. will be made within the law.

[ivan]

Many laws are set in place to protect you and me from the large and greedy corporations (take a look at GDPR for example). Copyright laws are intended to protect authors, artist and content creators.

Also, please note that a program may be "open source" where its source code is publicly available however unless a specific license is included along with the code then you have to assume that the code is copyrighted.

[ivan]

With regards to online privacy, the most outrageous violations (or "threats" as you call them) are associated with the large multinational corporations like Google and Facebook. GDPR is a European regulation intended to reduce the power of such corporations.
As you keep chatting with ChatGPT, keep in mind the following: the AI can save everything you type and do whatever it wants with that data. There is little you can about it, especially if you are located outside of the EU.

[pgimeno]

If a software makes a shader or solve whatever problems and the solution works, you can trust it to solve specific problems.

Yes but you can't trust it to be copyright-free code.

[glitchapp]

If a software makes a shader or solve whatever problems and the solution works, you can trust it to solve specific problems.

Yes but you can't trust it to be copyright-free code.

can a simple shader that renders a blue noise dithering effect be copyrighted?
I have no idea if that's possible and where is the limit of what it can be copyrighted but it seems like a too simple task to be copyrighted, correct me if I'm wrong.