What is this insect? Florida, Fall '09

At first sight, I'd say it's a cotton stainer, a Dysdercus.

It's a Scarlet Bodied Wasp Moth

Spot on, Ros.

I didn't know what it was, but I recognized it as a moth from the antenna and the body. So I googled "Florida moth". One of the images was for a polka dotted wasp moth which I recognized as similar in body style, so I googled for wasp moths and one of the images was a hit. Once I had the name I used it and found lots of hits.

(one of my interests from my work in internet design has always been how people search for information when they don't have the correct words to start from. Searches for unknown plants and animals are some of the most challenging)

It's one of my interests too, finding the right key words or the right combination of words..

One is impressed with the treasures one can find using very precise keywords..

Not only that, but it's interesting the way the human brain classifies information, especially visual information.

The "Body Style" of the wasp-like-moths was what I recognized, even though I have never seen them before. It would be nice if there was a way to search the internet through similarities in visual information. But I feel there's something deeper in my (and other people's) ability to recognize body plans in animals than just visual elements. I feel there is something innate about body plans which is possibly linked to the same instinctive ability to recognize faces (something which kids are finely tuned to).

Your ID and comments are fascinating, Ros!

I didn't know this was a moth at all... my Internet searches for "Florida Insect Bug Red Fall Blue Spot," etc., gave me nothing useful.

Now I know what the fellow is!

And PS: We were able to capture him and release him outdoors, so he's a Happy Moth!

How is the internetanswer to visual searching coming? Ive heard that the present software is wholly inadequate and expensive.

I don't think there really is any functional visual search mechanism out there yet, or even in the works. There have been a few attempts to get humans to tag images with keywords so that a visual database could be constructed, but that's just another "word-search" paradigm using humans to categorize images. It's not really a "visual" search the way I think of it.

At some point I would expect that "facial recognition" type algorithms will be used to try to categorize images, but a base categorization hierarchy will have to be defined and then the recognition software will have to be modified to target it. At least that's what I anticipate, but I'm not sure if it's being done yet.

But I've been out of the high tech world for several years now, so I may be missing some of the new stuff.