Yes of course, simply check the “yes” box next to “can the community taxon be improved” and the observation will go back to needs ID.
Good thing that that’s not what you’d have to do to get the observation out of casual then.
It doesn’t
At risk of being a bit negative, what’s the point of continuing conversations like these when what these really need is input from staff. It’s often not clear if issue threads like these even get noticed / read by staff. There’s like 40+ threads related to the CVs issues alone with some going back years. This issue related to “as good as can be” also has a number of older threads, a few i was a part of.
There’s the feeling of a disconnect between the engineers and the community. The engineers of iNaturalist do not really consistently engage with the community. Though occasionally you will see them comment on bug reports.
I suppose I just don’t really know what to say anymore that’s truly valuable and I actually feel burnt out talking about the same issues over and over. Any real change or solution needs staffs involvement. Otherwise even if you brainstorm a perfect solution, if it’s not implemented. There goes your time.
To end more positively, let’s hope for the best and that iNaturalist improves.
Add a comment to the blog post for product goals July to December ?
https://www.inaturalist.org/blog/133388-our-product-goals-for-july-december-2026
We seem to have three potentially practical solutions mentioned, and there may be more:
Rescind the change, which solved the first problem but caused an arguably worse one (loss of perfectly good observations). (Don’t bother to suggest that observations sent to Casual aren’t “lost” because while that may be technically true, observations sent to Casual are lost for all practical purposes.)
Provide a pop-up when this specific situation occurs (more specific ID added to one that has been marked “No, it can’t be improved”), asking the identifier to click “Yes, it can be improved.”
Have the computer automatically remove “No, it can’t be improved” or, conversely add “Yes, it can be improved” each time a more specific ID is added to an observation marked “No, it can’t be improved.” (This wouldn’t solve every example of this problem but would take care of most of them.)
Any preferences, staff? None of the suggested solutions is perfect, but it’s time to end this accidental problem. And maybe solution #4 is even better.
We need to make a lot of noise for a prolonged period for functional problems, issues, defects and bugs to get noticed. The CV issues made it to the official identifiers poll, so it is at least acknowledged in a roundabout way.
The bungled communication of the roll-out of this one, the subsequent baffled responses by staff, and the very belated and inadequate response badly shook my confidence in whether iNat management is in touch with the needs of the community, and particularly the needs of those of us with specialties that are less easy to accommodate in a system based primarily identification of field photos.
There have been other moments of disappointment, including a recent one that I am still smarting from, but something that all of them seem to me to have in common is that an acknowledgement from staff that they recognize the problem and take it seriously would go a long way towards making us feel heard and valued. Even if the answer is also that it’s a complex problem and they don’t have a solution right now but are continuing to work on it.
Instead, the overall messaging I’ve gotten about this one is that the current system is apparently not felt to be a problem because observations would not become casual if people were using the DQA properly and sparingly.
https://forum.inaturalist.org/t/north-american-sinea-id-and-the-sorcerers-apprentice-problem/68337
https://forum.inaturalist.org/t/pasimachus-elongatus-range-does-not-include-eastern-us/67295
https://forum.inaturalist.org/t/recommendations-on-improving-the-ai-algorithm/63027
https://forum.inaturalist.org/t/use-computer-vision-on-each-photo-in-an-observation/4210/1
https://forum.inaturalist.org/t/misidentifying-orthopterans-as-earwigs/73704
https://forum.inaturalist.org/t/cv-suggestions-no-longer-accurate-after-city-nature-challenge/65001
https://forum.inaturalist.org/t/make-computer-vision-include-hybrid-taxa-on-an-opt-in-basis/60160
https://forum.inaturalist.org/t/allow-some-non-leaf-taxa-to-be-added-to-the-cv-model/63937
https://forum.inaturalist.org/t/are-genus-level-rg-observations-used-for-cv-training/63859
https://forum.inaturalist.org/t/computer-suggestion-being-too-precise/48371
https://forum.inaturalist.org/t/cv-of-ophiocordyceps-unilateralis-complex/73228
This is just what I have saved and it’s missing plenty of other CV issue raising threads.
There is literally a novel above, hours and hours of reading.
Perhaps a forum post to gather these together and help reduce the amount of duplicates would be good. Just general systemic issues that get talked about over and over.
Reminds me of something I learned in a writing class probably 25 years ago (but it’s still true). If readers keep misunderstanding something you write, it’s easy to think, “People need to be better readers!” However, you don’t get to pick your readers. You have to write in a way that communicates with the readers you have.
We who use “No, it’s as good as it can be” DQA are doing our best. Wishing for better DQA users is a nice dream. That’s all it is.
Meanwhile, I’ve concocted a paragraph to paste into ID’s when necessary to explain what a person should do if he/she adds a more precise ID to an observation where I’ve clicked, “No, it can’t be improved.”
If I’m clicking “ID cannot be improved” on an observation, it is because I do not think people should add a more precise ID.
It does not strike me as a good use of my time to provide a note (copy-paste or otherwise) every single time I use the DQA on an observation in order to proactively explain what will happen should someone decide to add a more precise ID at some time in the future.
Yes, which is why a system that makes observations casual if people aren’t paying attention is not a good system.
This could be improved, but here’s my try as a warning to people who add a species name after “No, it’s as good as it can be” has been clicked:
If you choose to put a species name on this, please watch what happens to the observation! If it becomes “Casual,” the software as written can’t cope with your improvement. Click on the observation and find the DQA (Data Quality Assessment) and click, “Yes it can be improved.” That should fix the problem (unless two or more people clicked “No, it’s as good as it can be.”)
That would also help when people choose a homonym and dump the obs in Life, due to Kingdom Disagreement. Why didn’t my ID do what I expected it to ?!