Sandbar_Trekker

You’re cutting out the important context with that quote:

Occasionally I’ll have it write more substantial code for me, usually for bug fixes, and occasionally for small features. I find it does a pretty decent job here, but I keep it on manual mode and often ask it to write the code in a different way, or take a different approach.

That doesn’t sound like someone who’s just blindly copying and pasting whatever an LLM told them to do. I agree there’s absolutely a risk with using LLM generated code that could cause some copyright issues. But the way he’s using it sounds like it minimizes that risk.

The Licensing Problem It doesn’t sound like the way the dev is using the LLM would make this a high risk issue. He’s mostly using it for testing/brainstorming/pseudo code rather than generating code to use directly.

The title of this post is incredibly misleading and I can tell you didn’t read how it’s actually being used. The dev mostly isn’t using the LLM for code generation. There wouldn’t be any “potentially unattributed plagiarism landmines” to worry about.

Edit: The dev does use the code in some cases, see below.

I think you meant LLM tools?
If you’re meaning all AI tools then that also includes anything ML related. Accessibility tools such as Speech to Text models, Text to Speech models. Research tools such as those used to virtually unravel and read heavily damaged or incredibly fragile ancient scrolls, tools to help detect exoplanets, or tools for discovering cures to diseases. And more.