AI code thief

[aeric]

I know many developers enjoy using AI to assist them to write code.

There are also general public who has little knowledge of programming or computer science able to create an app or game just by "vibe coding".

AI uses LLM which Web scrapped from the Internet such as stackoverflow, forums, source code repositories like github.

What if a repo is licensed under some open source licence with some limitation?

Even though these licenses are highly permissive such as MIT licence, the developer who use them need to follow the rules to give credit or stated the name of the original author.

Licences like GPL variant as used on Linux require the person to contribute back to the original project.

I think all these licences will be ignored and the project owner rights are not protected.

What do you think?

As programmer, do you think you support this to be happened?

 

[Sandman]

I think all these licences will be ignored and the project owner rights are not protected.

Yes, this is what is happening already. There's a fair bit of discussion on the net about it, if you just go looking for it.

 

[rabbitBUSH]

As programmer, do you think you support this to be happened?

Interesting point/s. No don't support this At All. I come out of two separate industries where acknowledgement and attribution are a Strong requirement. To a high legal extent.

AI seems like it can only produce high levels of plagiarism by "jump-in" programmers. But, the question is - is using AI suggested code plagiarism? Could those who own the AI platform begin to monetise with royalty claims? The article I posted earlier suggests that DeepSeek appears to have taken that option away (by showing how low cost platforms can be and therefore low profit).

@aeric : It's All A Really Good Question - if one assumes all this goes under the rubric of the "democratisation" of stuff - just how far does democratisation take us towards Everything is Free (both in money, grab-and-use like it's mine, the of course it's everyone's principle, and so on and so forth).

There goes individual achievement with it all. Poof!

Again......gee what nice question...but one academia has struggled with for centuries, I'm afraid.

 

[aeric]

As a programmer, I felt helpless.
Our hard work are being used without getting respected.
People take things for granted.
It is sad that this cannot be stopped.
The giant corporations have more resources in legal and make money with the AI.
You need to pay for the tokens to access copyrighted work that were not created by them.
Sad.

 

[rabbitBUSH]

copyrighted work that were not created by them.

Try working in an university environment these days......

 

[aeric]

Try working in an university environment these days......

Lecturer or Academia will reject work without reference to a journal or source.
Is this still applied?

 

[rabbitBUSH]

Is this still applied

Yes it's still a foundation principle of ethical research and student outputs and the assessment of outputs - but - students today follow the tracks your questions are about - so it's a huge struggle to control copy-and-paste (ie. Quote without acknowledgement or attributions). Theft of intellectual property is the norm. Plagiarism is time consuming to police.

The subtlety of mind in order to grasp the ethical principles is missing today.

 

[Mashiane]

As a programmer, I felt helpless.
Our hard work are being used without getting respected.
People take things for granted.
It is sad that this cannot be stopped.
The giant corporations have more resources in legal and make money with the AI.
You need to pay for the tokens to access copyrighted work that were not created by them.
Sad.

If you remember some time ago that the internet went down because of faker.js, its author going through the same thing and not getting credit or even contribution and he took it down. Microsoft owns stack-overflow and also Github, the wealth of code content in most of these sites is enourmous and most of it its MIT, so basically putting stuff in any of these tools, makes it automatically theirs.

Perhaps for now we can still have some little hope of not being taken over by AI, this sphere changes very quickly... its a sad revelation

https://arc.dev/talent-blog/impact-of-ai-on-code/

 

[Magma]

Are/Is not only the Programmers... what about Writters ?

generally... AI must need to have some control, rules... and violation of rules must be punished !

 

[Magma]

and i am not happy... for that i am gonna tell.. but need global coperation to achieved that..

AI are the new Pirates - Hackers (remember Warez sites, mp3 copying... etc)

 

[rabbitBUSH]

so basically putting stuff in any of these tools, makes it automatically theirs.

I seem to remember that when goo-bloody-gal started the user photo repository services (G-drive etc) - someone read the Ts and Cs discovering that goo-grabber claimed ownership of such images . Which precipitated the Goo Fighters. The terms were quickly changed - again - if I recall .....

One can imagine an AI driven announcement "that's a nice house, if you don't respond to this message, we'll sell your house".......automation BB style (or was that HR style).....

 

[aeric]

If you remember some time ago that the internet went down because of faker.js, its author going through the same thing and not getting credit or even contribution and he took it down. Microsoft owns stack-overflow and also Github, the wealth of code content in most of these sites is enourmous and most of it its MIT, so basically putting stuff in any of these tools, makes it automatically theirs.

Perhaps for now we can still have some little hope of not being taken over by AI, this sphere changes very quickly... its a sad revelation

https://arc.dev/talent-blog/impact-of-ai-on-code/

Yes I still remember.
AI is not very good in B4X, at least it is now confused between B4A, B4J and VB. For JavaScript, Python or other mainstream languages, it is very good.

 

[aeric]

I seem to remember that when goo-bloody-gal started the user photo repository services (G-drive etc) - someone read the Ts and Cs discovering that goo-grabber claimed ownership of such images . Which precipitated the Goo Fighters. The terms were quickly changed - again - if I recall .....

One can imagine an AI driven announcement "that's a nice house, if you don't respond to this message, we'll sell your house".......automation BB style (or was that HR style).....

And Meta (fka Facebook) use our photo and video as they own the rights to train their dataset.

 

[peacemaker]

Hmmm, indeed, AI is trained with public software, not big commercial code base that is not visible for training.

 

[geps]

I think the concerns are valid, especially around attribution and ensuring open-source licenses are followed. With AI helping people generate code, it could get messy if folks don't properly credit the original authors or follow licensing terms.

 

[Magma]

Some open sources are for personal use and some other for commercial.... the use of result ?

 

[aeric]

This is what happened these few days. Sad.

Lawyer Says Studio Ghibli Could Take Legal Action Against OpenAI

The viral Studio Ghibli AI trend reignited a heated debate surrounding copyright. AI expert Rob Rosenberg says the company could sue.

futurism.com

 

[Didier99]

so basically putting stuff in any of these tools, makes it automatically theirs.

That is one of the more frustrating aspects of this to me. I generously give up code under a permissive MIT license for the betterment of humanity, under the expectation that even though I won't make a dime, I may get fame and glory, and then someone else makes money off of it and I don't get any credit.