VizWiz – crowdsourced assistance for the visually impaired

This is pretty amazing – a researcher at the University of Rochester  has created a working prototype of a system that allows remote sighted workers to help people with visual problems in nearly real-time. 

Users take a picture with their phone, speak a question, and then receive multiple spoken answers”
In other words, take a picture of something with your iPhone, ask a question “Is this a can of soup or beans”, and get an answer back in seconds.  It uses a crowdsourcing service from Amazon called Mechanical Turk where people get paid small amounts for carrying out small tasks. 
When i first heard about this yesterday, I was genuinely stunned at the elegance and utility of this idea, and am very much hoping it takes off.
There’s a video of it in use over a the project page here: