Similar to what most people are saying with the search function being out of the way.
Also, this is probably informed by the fact I try to identify a lot of Unknown observations. However I try to avoid IDing to species if I’m not already knowledgeable on said species, as there’s a tendency of some newer users to thoughtlessly duplicate IDs regardless of their personal knowledge of their accuracy, meaning that inaccurate observations can get to Research Grade, which means it’s less likely to be corrected and more likely to inform the CV the more this happens, which will then make less accurate guesses, and so on.
I find genus/family suggestions helpful in those cases as they’re at least a step in the right direction (which might be more precise than whatever I type in) without being automatically pushed to Research Grade in those cases (I am aware genus-level observations can get RG status but it seems less common for inexperienced users to figure out and misuse that option), and can get the observation under the eye of someone with more specialist knowledge looking under a more specific taxa who can often correct my use of the CV in cases its result is incorrect, or give further details in the event it is.
There might be cases with iNat Next where I could do the same thing (i.e. if all the CV options are in the same genus i could type in that genus), but it’s possible I might not know the name of a broader clade that the organism is likely to be in that still improves identification more than my vague guess of “Plant” or “Fungi” or something.
You might not have seen this, but when the list of ID suggestions comes up, you can click on one of them and you will see all the taxonomic levels above it. You can then click on any of those and “Select this taxon” to make it your ID. Quite a neat feature.
I may be wrong but i think a lot of the reason for the things that slow down us power users are to onboard new/casual users to avoid things like forgetting to add ID, upload observation, etc. It may be too much to program or too gamified, but i think it might be fun to allow an ‘expert mode’ you can unlock by getting a certain number of rg observations, IDs, or whatever. Even just going to the website and doing a tutorial. I’d also pay for a pro version of the app, though that’s not an ideal approach since it sets up an obvious barrier especially considering the way currency is valued in different countires (can you make an app cheaper in different countries? I honestly don’t know)
This other thread branched off in some other directions as well but most of it was discussing this subject in a bit more detail than the current thread: The role of CV in the new app
It seems like the CV can effectively now be ‘pretty sure’ at a species level? I don’t know if I have seen the app actually suggest a taxa higher than species yet at all (to be fair I’ve only made a few observations with it), but it will highlight a species on top as a ‘top suggestion’ with a full bar of green circles, and then everything else has 1 green circle.
The key takeaway for me is that it makes the observation process take as much as 10 seconds more than the old app. Either due to the connection stalling or freezing the CV, the CV having to “load”, or otherwise waste time when all you want to do is enter a placeholder and move on to the next observation. If it simply had a button to press first to go “view CV suggestions” that would be an easy fix.
I do appreciate it’s difficult to plan for imperfect conditions, but with an app that encourages you to be outdoors and in the wild, I feel that’s more important to consider with iNat.
This app is designed with better accessibility in mind, and someone who’s colorblind may not be able to distinguish this if only the color is different, so the icons for each quality grade need to be visually different in other ways.
This should be fixed in the next release.
This is more about how the app is handling suggestions. In the legacy apps and on the web, it finds visually similar taxa (mostly species, which is the vast majority of what the model is trained on) and for the “pretty sure” suggestion it tries to find a common ancestor for the visually similar species. I believe that in iNat Next, if the app is quite sure of the top suggestion it won’t show the common ancestor. Personally I always want to see the common ancestor, but this is also why we’re soft-launching it - to get feedback.
I vote for providing a “Use legacy interface” option in the settings to avoid some of the frustrations with Next (for me, not being able to use my own developing expertise to select a taxon name that CV has not suggested).
it’s true, though it’s a bit slower and doesn’t currently work offline. I get why they don’t like the ‘placeholder’ text but it does mess with my work flow so i really don’t like it.
If you update to v0.55.3 it has offline taxon search. When you load the app, it should try to download the 500 taxa w/ the most observations near you (or the top 500 worldwide if you haven’t granted location permission). The app will also cache taxa you search for, both from Suggestions and from Explore. This could definitely be improved (cache captive/cultivated taxa first to assist new users observing their pets and houseplants, cache taxa you have observed to better serve existing users), but it’s a start.
