Are you wanting to bump all IDs, or just yours? Because if it’s all you can do a taxon split that bumps them to genus.
I can’t speak for Tony or other staff, but as you have described it I think this may cross the line, so I would be cautious and get a definite answer first.
Not only can “automated” IDs accidentally create bad data - they can also be used to do so intentionally, which is why I have never supported any such functionality.
I would also use caution here, as this may be a mis-use of taxon changes, which i.m.o. should be reserved for actual changes in taxonomy, and not merely to correct misidentifications.
3 Likes
Just to be clear, the bulk action version of the tool has been available since Sept 15. I’ve made minor iterations on it since then, but the only thing left to do is put it up in the official extension stores and publicize its existence (right now you have to kind of dig for it in the thread and jump through some minor hoops to install). You should try it for yourself and see what you think based on that, not just the description. Even if you don’t want other people to have it you should at least appreciate having it for yourself :D
Just to be clear, I don’t want anyone to be able to add or change IDs in bulk to observations on iNaturalist, including myself. I would not make an ID on someone else’s observation without first examining it (including all the photos, and any and all notes and comments from the observer and others), and I’d prefer my own observations to be treated the same way.
If I discover that I have added mistaken IDs in the past, I take my lumps and go back and withdraw or correct them, as appropriate to each observation.
5 Likes
Of course you should examine the observation before applying any metadata to it! No one is suggesting it would be okay to do otherwise.
Maybe it would help to explain a typical workflow? I just added 1600 gall colors annotations in 1.5 hours, letting me test some interesting hypotheses and highlight a number of anomalous observations. Would have taken all day to do it manually. Here’s the process:
Use URL filtering to select only observations with no value for the “gall color” Observation Field.
Manually review every observation and assign its color with a button one by one if it is NOT white.
Refresh the page to exclude all the ones that I tagged as non-white
Bulk action tool to add the white value to all remaining observations.
I looked at every single one but spared myself 1000 keypresses, which is huge for my repetitive stress injuries and time management.
Yes, it is possible that someone could not do that diligence, but iNaturalist is a monument to what can be achieved by trusting that people will contribute responsibly and in good faith.
3 Likes
I grant that this example would be a good and responsible use of such a tool. It’s potential for destructive results (unintentional or otherwise) is also lower, in my view, than having the capability to add new taxon IDs to 1000+ observations, as was contemplated in the original post in this topic.
You and I and others who have been using iNaturalist long enough will understand what a responsible and good-faith use of such a tool looks like.
But that will not be true of the (likely) majority of iNaturalist users overall, however good their intentions may be in the moment. We have a steep (and long!) enough learning curve just for the existing features of iNaturalist, not to mention what the curve would look like for a tool like this.
And among the 8+ million good-faith iNat accounts there have also been the very few bad-faith actors, in whose hands such a tool could be enormously destructive to the entire community.
So for those reasons, to me it’s just not worth the small but potentially disastrous risk of having such a tool available, however personally helpful I might find it to be.
I’m guessing thoughts like the above went into the long-standing iNaturalist policy of not permitting machine-generated content to be posted on the site. Used responsibly, your tool might not cross that threshold, but it would make it much easier for others to cross it, even with good intentions.
Just a minor note - the larger the batch I was working on, and the more time it takes, the more I would be concerned that one or two newly-posted observations would be captured by the URL filter upon refreshing, and those could mistakenly get tagged as white in your scenario.
2 Likes