First time using the Forum here but I wanted to chime in with what I think may be a similar case. I’m still familiarizing myself with how everything works on here so pardon me if this is not the right way to go about this. See linked discussion above for more details on this topic.
I have noticed someone doing thousands of moss and bryophyte observations with their own IDs, but with a single unmagnified photo, which is very strange and unconclusive for bryophytes. Often the IDs are very clearly wrong to anyone that knows the first thing about bryophytes, or even someone who can just look closely. I correct the ID and leave a brief explanation but have never gotten any replies. Maybe I should start a different thread/post on this topic so I’m sorry for the tangent, but I’m wondering if this person may also be using AI, and if so, how to know or how to proceed. Their hundreds of incorrect observations are just actively bad for iNaturalist and those on here trying to really learn and contribute to the pool of information on here.
Correct the ID and leave a brief reply, as you already have done.
Send them a polite PM explaining the issue.
If they remain unresponsive and continue to upload a high volume of observations with inaccurate IDs, send an email to help@inaturalist.org with the specifics.
I started trying to learn bryophytes less than a year ago and I’m just at the point where I know how much I don’t know - which is a lot! Because of that, I have a little sympathy for someone who’s perhaps trying to learn bryophytes by relying on iNat’s computer vision. Even in my region (northeastern US), there aren’t enough properly IDed observations to trust the computer vision, unlike most vascular plants here. I use the computer vision when I upload my observations of bryophytes, but only to give me a vague idea of what the plant might be. Then I go key out the specimen and make my own ID (which may well be wrong!).
Until enough competent bryologists start making and correcting IDs on iNat in large numbers, this problem won’t be solved. The best thing you can do is keep correcting IDs as much as you can stand (and thank you, by the way!), but just ignore unresponsive observers like your example for now. Maybe in a couple of years, there will be enough competent IDers to band together and make a big push to educate the computer vision and the observers, at least in some regions.
This probably won’t happen until enough observers take pictures of the identifying features, which often requires microscopy. I’ve talked to bryologists about this issue and the prevailing consensus was that trying to identify bryophytes on iNaturalist frustrates them because pretty much all they can do is bump things back which can cause arguments and is not as much fun as being able to figure out what something is. So they prefer to spend their time looking at samples through a microscope and posting their finds to the Bryophyte Portal.
There are more and more people posting microscope pictures on iNat though. From an observer standpoint, I can think of two strategies that could help:
Add tags/observation fields for microscopy to make such observations findable in a search and add them to a suitable project (several exist already). This could help narrowing the pool for knowledgeable bryologists so they don’t have to wade through pages of blurry pics of distant moss patches to find the ones that actually can be ID’d based on the evidence.
Make sure in addition to the microscopy images necessary for ID, there are some good in situ pictures as well. Maybe/hopefully with more of these getting identified correctly based on close-ups, the CV algorithm will learn how to recognize the field images correctly.