See the comments promoting an app for identifying plants - they appear to have set up an iSpot account to drum up business!
Probably not - they’ve got an official iSpot AI icon, so I assume that it’s part of the OU’s recent introduction of AI identification engines to the site.
I had just come to the forum to report a different issue. After seeing a [email protected] offering on a Fumaria observation I went to look at their profile to see what else they had suggested earlier, but their activity tracker is empty.
(Also, the mouseover text on their icon is FASTCAT-Cloud, and links to the FASTCAT site, so perhaps whoever set up the account copied and pasted, and didn’t change everything that they should have.)
hmmmm - 155 suggested species - that would be a small book on its own and about 150 more than I could narrow it down to without being an expert at all.
Not many of these photo based ID apps get a good rep - let’s hope this isn’t the way forward for iSpot.
I’ve asked [email protected] if they reply to comments - I will be surprised if they do but you never know.
See the belated news item on the home page. I found the first occurrence and asked Admin - it is not Spam of course
150 ‘suggestions’ from an AI ‘recogniser’ shows that refinement needs working on. But its conclusion is spot on.
If you take the photo to Google images, it returns more than 500 (I did). iNat was also near perfect.
This will end up with iSpot having no purpose!
[email protected] Proifile shows no activity - maybe it’s dumb though watch out for sentience which is VERY Topical
Look at profile and you will see ‘it’ it is probably trained to not respond.
I have asked Admin to review the Profile https://www.ispotnature.org/view/user/80188/activity-tracker
so that it shows activity (it is blank) - we can then visit ALL occurrences in the normal way
You are right that these are two AI systems that are being linked to iSpot. Others from iSpot admin team are writing more about this (e.g. iSpot and AI: FASTCAT-Cloud | Article | UK and Ireland | iSpot Nature) so will keep my comment short and we will be running feedback sessions on them over the next couple of months. Basically the AI system is just making a comment showing the items it thinks it might be, it does not give an identification. It is up to the humans to take or leave any suggestions it makes and anyway the suggestion comment it makes will probably be set to come in after some time, not immediately, so humans have time to make an ID without any of its comments.
Its profile claims that it’s made 13 comments, 2 of which are alleged to be replies to my comments, but as the Activity Tracker is blank I can’t locate them. It’s comments didn’t show up in my Changes Tracker. (It looks as if its activity is being deliberately not reported.)
On further cogitation I suspect that “Replies to your comments” to on [email protected]’s profile means replies to [email protected]’s comments, rather than replies to my comments. (That would be a bug; it should read “Replies to your comments” when looking at your own profile, and “Replies to X’s comments” when looking at someone else’s profile, or something on those lines.)
I think it is number of comments in the Observations where IT has made an entry. Susequent replies (that are not always to IT) are counted as responses.
It seems there are technical issues that prevent ITS activity being shown in the Activity Tracker here https://www.ispotnature.org/view/user/80188/activity-tracker.
I think that is a great mistake because it means we do not know where, or when, IT is active, so cannot contribute to the process.
Here are two
Orchid | Observation | UK and Ireland | iSpot Nature
I’ve lost the Other already.
But we need them all collecting so we can spread the news to Other Users (@miked ?)
The three I’ve seen are Aldcameron’s spotted orchid, Olearam’s ramping fumitory, and Marylegg’s Scottish primrose.
You’re right - “Replies to your comments” isn’t appropriate if you’re looking at someone else’s profile. Easily changed.
I’ve looked at the Orchid post in question. JoP agreed, Lavateraguy couldn’t be sure… that’s the nature of botanical ID from photos or pictures in a book etc. And, as L says, some details are more important than others.
is appropriate to any ID suggestion.
Besded on a review in BSBI on AI apps I chose SEEK. For the orchid ( from the posted photo) is said
I find SEEK to be less definitive & often suggests, as here, that more info is needed.
For me, as I’ve said before, the attraction of iSpot is the opportunity for discussion & learning (i.e. labellum shape in Dactylorchis is more important than markings).
Last week in the Engadin, Switzerland, where I’m looking at alpine flora, I chanced upon a botanist photographing a hybrid Dactylorchis. He said it was his main interest these days. All (I’ve posted so far myself is a Great Crested Grebe. )
iForum LIVE! session: Cos4Cloud services - Why are we integrating FASTCAT-Cloud and the Plantnet API into iSpot?
As you know we have been trialling the integration of two Cos4Cloud image recognition technologies FASTCAT-Cloud and the [email protected] on iSpot since June 2022. Want to know more about Why we are integrating FASTCAT-Cloud and the [email protected] API into iSpot? you are invited to join a iForum LIVE! scheduled chat discussion in the iSpot Forum on Thursday August 4th at 3 – 4 pm BST / 4 – 5 pm CEST. Click on this link to participate*: iForum LIVE!: Why are we integrating FASTCAT-Cloud and the Plantnet API into iSpot?
*Please note you will be invited to join the iSpot Cos4Cloud User Group to participate. We are keen to share and gather the iSpot community’s feedback and have users involved in testing and this iFORUM LIVE! discussion is part of this. See more: About the Cos4Cloud Project: iSpot Cos4Cloud User Group
More in this iSpot front page news story: iForum LIVE! session: Cos4Cloud services - Why are we integrating FASTCAT-Cloud and the Plantnet API into iSpot? | News | UK and Ireland | iSpot Nature.
Hope you can join us!