Green Group discussion

Many of us grew up with small field guides for identification of animals, insects, amphibians etc. There is always something we have not seen before, or not been really sure to identify.

I recently tried the Google AI on a small moth. I asked a few questions, then decided to upload my photo. I had to zoom in on my image to get finer details to emphasise - the AI tended to be a bit lazy and go for the moth most people are searching for, in my humble opinion. Such as a house moth or Indian meal moth. This was an outdoor moth and I'm in Dublin.

Eventually we got to this point:

Clare: "The Common Grass-veneer (Agriphila tristella) seems to fit all the details, thank you."

AI: "You are very welcome! It is incredibly satisfying to piece together all those subtle visual clues—the cigar-rolled wings, the dark color variation, the afternoon flight, and that distinctive metallic tail fringe—to land on a precise match.
The Common Grass-veneer (Agriphila tristella) is a fantastic example of how variable wild insects can be, and you did an excellent job tracking down those fine details on such a small creature.
As we head deeper into the summer, you will likely see many more of these little "cigar" moths darting out of lawns and long grasses around Dublin.
If you spot any other unusual insects or want to look into citizen science platforms like National Biodiversity Data Centre (NBDC) to log your Dublin wildlife sightings, just let me know!"

I will report this to the NBDC; and the trouble there is, you are expected to know what species it is before you start recording. Without photos on the site to help. So this is useful.
I would encourage more computer users to try Id-ing flora and fauna this way, but to not necessarily accept the first answer.

I've heard that queries to LLMs are very energy intensive, which, if true, makes their use problematic from a green perspective. It's a subject of some debate though, and also one of the facets of AI use that is not obvious from the outset.

I'm glad you found an ID for your moth. It sounds darling. :)

Yes, you're right, more energy and water are used than with a regular web search. I had tried my books and a regular web search and image search first.
This is my usual butterfly and moth book. It does not cover micromoths, which includes the grass moth.
Butterflies And Moths

I always wonder how many of the insects we see when we are out and about have not actually been identified yet. It can be tricky! I saw a hairstreak some months ago which looked nothing like any of the blues I've seen before. Could not find it online. It was *tiny.*

This paper suggests there are thousands of new insect species identified in the past year.

https://www.eurekalert.org/news-relea...

Some moths and butterflies are hatched with unusual colour markings in each generation. If the new colour proves better for camouflage, the moth will get to reproduce and spread the new strain. If it's worse for camouflage, something will eat it first. In this way, the population can adapt to shifts in forest cover, field use, city pollution, light pollution and climate.
So if you are searching for a moth, you could say dark pattern variant, for example, if it just looks darker than the photo in the book.
It has been suggested that climate change and urban spread will cause differences in size as well, because of some of the above factors. Again, those which work for the individual will be reproduced.
Some moths and butterflies have gender colour variations, see the Purple Emperor.
His Imperial Majesty: A Natural History of the Purple Emperor by Matthew Oates