I'm with you on honesty, and I've certainly seen people tacitly trying to pass off AI outputs as human written. But I think we've reached a point where, in lots of contexts, we can't reasonably assume human authorship by default any more. (We can reasonably want it and push for it! I just mean we can't literally expect it.) So even when we would prefer openness, I think 'lying by omission' is too harsh a characterisation for people who choose not to declare AI authorship but don't actively try to cover it up.
The replies to my comment kind of say everything that needs to be said about what I actually said; people can't even put aside their overt hostility to all things AI long enough to acknowledge that it makes no [good] sense to draw specific attention to LLM usage.
Ironically, it's precisely because of those folks that nobody with half a brain would acknowledge that they use LLMs to bring some or all of their ideas to life.
You can't demand that all people should disclose their bellybutton rings AND openly salivate for cancelling people with bellybutton rings.
Honesty is the whole problem with ideas like this. If you're the kind of deluded idiot that considers LLM-generated crap "your code", stating exactly how little you had to do with it is not in your advantage. Far easier to maintain the lie.
If you review the code generated by the LLM in response to your specification and you own the responsibility for what that code will do in the real world... congratulations, it is now Your Code in every way that actually matters.
In a few years, your position will be revealed as getting angry for someone that called you with your number stored on their phone - instead of memorizing it and dialing the numbers manually - and didn't start the conversation by admitting that they haven't memorized your number.
They sure don't, but often insight into/alignment with the story and development process makes all the difference for which projects people choose to contribute to.