http://www.bbc.com/news/technology-30100184
Scientists at Google have created artificial intelligence software that can describe the contents of photographs far more accurately than ever before.
I can't find it now, but a few months ago I posted that G Earth is about to increase the 'standard' resolution to a meter or two. That's the standard, so you have to wonder what they can do if they get really serious. Looks like SkyGnet is getting fitted for new glasses and a brain to go with it.
QuoteSoftware description: A group of young people playing a game of frisbee
I wonder what it would say if it were a group of middle-aged people playing. How does it know they're young? Facial structure? Body position? On-page context? Random guess?
If a picture of me playing would say "A middle-aged guy making failed attempt at forehand frisbee throw" then I'm going to freak out.
Or howbout "RC didn't like frisbee and The Creator Larry had deemed that The Chosen must like frisbee so We (the royal 'We') zapped him."
http://cs.stanford.edu/people/karpathy/deepimagesent/
"Code coming soon" -?
It'd be great to try out
An alt attribute need never be wrote again...