Policy on use of LLM tools and bug fixes

[Richard Kenner]

> Any sequence of bytes 

Any sequence?

So an auto-complete done by an LLM counts, but one generated by
a less-sophisticated plugin doesn't?  What's the difference?

> computed by a local or remote large language model

I notice you didn't say "generated".  But what does "computed" mean?
Do you only include literal code or do you include algorithms?  So if
I asked an LLM to generate pseudo-code to create some specific GCC
tree and then manually translated that pseudo-code into C, that doesn't
count, but if I asked the LLM to generate the C directly, it does?
What's the difference?