And thank goddess someone is volunteering, because I almost snorted coffee through my nose in comical understanding at Ojos co-founder Munjal Shah's explanation of why this company exists: "I am co-founding a company because I found I had 31,246 photos all named DSC0009.jpg." Ojos is using face and text recognition technology to automagically tag the vast sea of unsorted digital images in your collection. I post only a subset of the digital photos I take to Flickr — ones that I like for aesthetic reasons, or ones that capture a particular moment or event in some way. The rest of the shots… well, they lie in ignominious anonymity in the unchartered areas of my "Photos" folder. If Ojos could take them, have its way with them, and spit back a useful data set — that's going to be absolutely golden.

My worries are threefold: 1) what about the potential to "do wrong," to actually mislabel photos as things they are not? Too many errors and the cost of having to fix those will outweigh the benefit of the tagging as a whole. 2) How well will it deal with ambiguity? Will it correctly identify my obtuse abstracts? Will it recognize my friends dressed up for Halloween? 3) Seems like there's an ultimately limited potential to "do right." Many of the tags I use on Flickr aren't so much descriptive as… well, poetic. I want to see the photos I've tagged as "spring" or "love" or "home" more than I find myself needing to see photos I've tagged as "bicycle" or "building." I suppose we'll see, as soon as my knock-knock-knocking on beta's door pays off… :)

[Via Tech Crunch]