The school kid projects gone wrong are, a not nice, part of iNat. I flagged up a joke ID too. Now we can have the ID hidden by a curator - which is a small improvement.
Oh, I know about them.
This time, it seems that the gall and leaf miner community was very busy feeding the AI, meaning that many host plants were identified as their parasite (without the presence of it). And probably some mistranslations of bones (“beans”) and grass into completely different species, my guess in a non-english language. And it seems India’s AI sees snakes everywhere 
But then you find little gems like that owl or insect hiding somewhere you never saw it even when closely looking.
2 Likes
that is from Afrikaans. In Cape Town I set my common names to English and Afrikaans which helps me pick up where iNat trips over multilingual.
But CV tends to offer random animal, or virus / rust first - when the subject is the leafy plant. Okay when you know what you are observing, but not a good ‘working as intended’ for very confused newbies.
Now down to 19,250 unreviewed New Mexico CNC weekend observations 
5 Likes
I’ve switched from confirming correct CV plant IDs to disagreeing with incorrect CV fungi IDs that people blindly agreed with (not even close) or agreed because it does look like it but they don’t realize there are dozens more that look the same.
My CNC 2024 ID count so far
Baton Rouge, LA: 795
Houston/Galveston, TX: 89
San Antonio, TX: 41
Dallas/Fort Worth, TX: 18
5 Likes
ugh. there are so many bad identifications and unassessed captives now. i guess we have a whole year to work through them until the next big wave.
6 Likes
You have … thru 5 May … to ID all the obs …
Im poss ible t ask !!
2 Likes
May 5th of what year in what century?? ;-)
8 Likes
Went through all the non-prickly-pear cactus (around 500 observations) and several more pages of random and common garden plants last night and this morning and 18,600 unreviewed remain. Up to 16% captive in the ABQ CNC project.
6 Likes
Well, that may be when the CNC folks tally their “official” statistics, but the unidentified–and badly identified–observations will remain to taunt us forever.
6 Likes
We keep plodding along, hoping for more identifiers.
It has been satisfying this time to use the new DQA and sweep the multi-sp out!
6 Likes
After working haphazardly through the Americas projects, I’m trying to narrow down to just my local area and see how much I can chip away. I can’t look at the total numbers or I’d despair XD
4 Likes
Re unidentified, to put it in context last’s year’s April Unknowns, mostly from CNC23, are down to around 20K or so by now, and it’s only been a year!
My personal haphazardness is to work in random mode. So instead of seeing the big pile barely whittling down, I think of each display page like a bingo card. Will I see my expected Bouganvillea or Madagascar Periwinkles or Tridax procumbens showing up that I can click on right away? Usually!
What’s on everybody else’s bingo card for this challenge?
8 Likes
You take is actually quite fun… thanks 
! I did a few and my Bingo is for sure everytime any animal shows … I did even find a mammal that was easy to put to species on the second card.
Superbingo is for sure if there is even a spider. Found some of those es well… always around 5-ish animals on there as well
2 Likes
2 Likes
this IS madness, but I ran through a page like that just for the novelty. I haven’t considered a personal bingo card, but I can almost always find a Prunus serotina on my local pages
1 Like
I have “finished” the “Plantae to Dicot” bin in Europe on the CNC Global Project ! There is plenty left to do but I like small victories :)
9 Likes
I’ve been whittling down on erroneous species from Houston and San Antonio CNCs. People just clicking whatever the CV says. Think I’ve pretty much got fungi covered so I thought I’d look at fish. I see there is one IDed as Fugu so I assume it is some other pufferfish species. I open the photo and it is a tangled pile of fishing line vaguely in the shape of a pufferfish. They didn’t even get that ID from the CV. That “CV suggestion used” symbol wasn’t there and when I looked at the CV results it thought it was coyote fur, dog vomit slime mold, spittlebug spit, a fire ant hill, spider webs, or white flowers. What the actual F?
6 Likes
during the CNC, i expect that there will be a few observers who ID based on whatever the CV says, but this year, i’ve seen more than one identifier in my area seemingly making some IDing based on whatever the CV says, which baffles me. i get there are (bad) incentives for people to bump up their own species counts, but why would an identifier just pick whatever the CV says for other people’s observations?
Sometimes people trust the algorithm more than themselves. I saw one earlier that was IDed as goldfish (which they were) then withdrawn and then they accepted the CV of Garabaldi (an orange colored Pacific coast saltwater fish). Maybe those people you mention think they are being helpful?
1 Like