Assuming it is right, why wouldn't I use the AI suggestion?

It is no more wrong than using the “agree” button


My understding is that human (at least this one) should not be doing this, especially at species level.

If I am doubtful, like my instinct says Augochlora and the app suggests the same but I also think a bee could be Ceratina, those metallic bees are tricky and they often visit the exact same flowers and a particular one was just a smidge unusual, I will just set upload to Bees and wait a week or two, then pop John Ascher a message if there has been no further activity. He is always kind and responsive and checks my bees when I ask. Or Jorge Merida, who is the bee expert for Mexico and is also extremely responsive.


I almost always use the CVS in combination with my common sense. I use these on my own unknowns and on other’s. Then I let more advanced users help out and these days I retract my ID as soon as I notice it is overruled. I do this because I have enthusiasm for Nature, but very basic knowledge of species/taxa. And I know this helps iNaturalist as a whole :green_heart:


100 pictures.
About 60 obs.
You have 28 … keep going for Your Highness. (But pictures must also come from multiple observers)

You know you only have to type
Eur yuc
and then autocomplete kicks in?

1 Like

I only see my 28 observations? And all the photos I see are mine. Do you see other observations? I am hoping someone else will spot it to spread the wealth. The fellow who is the local beetle expert used a very old Smithsonian reference drawing on an early observation but if there were photos from other observers I could look at that would be more helpful.

1 Like

Diana was saying that the threshold that needs to be hit for that species to enter the CV model is 100 photos, and that these pics tend to generally be distributed across somewhere around ~60 observations, so that’s the target you need to aim for


Thank you. It is 5:40 here and I have not had my coffee yet so still quite dim. (Coffee will only help a little, alas.)

1 Like

You may be able to retrieve more observers from the Genus level IDs ?

For plants especially that are unknown to me, I use the CV to get a genus identified in my record. Then I go back to the record and do some research. Is that genus known from the area? Is my photo consistent with other images on iNat? What species in the genus are recorded from nearby? Often I’ll check the iNat ID against what my PictureThis app IDed it as, even though that app may be less reliable than iNat. Eventually I get to the right ID or at least close.


This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.