Jason Perlow on the current capabilities of Google Glass and the easy to envision upgrades that will soon be possible:
Because Glass is an Android device, runs an ARM-based Linux kernel, and can run Android user space programs and custom libraries, any savvy developer can create code that modifies the default behavior in such a way that recording can occur with no display activity showing in the eye prism whatsoever.
And while the default video recording is 10 seconds, code could also be written that begins and stops recording for as long as needed with a custom gesture or head movement, or even innocuous custom voice commands like: “Boy, I’m tired” to begin, and “Boy, I need coffee” to end it.
You could write and side load an application that polls the camera and takes a still photo every 30 seconds, should you say … want to “case” and thoroughly photodocument a place of business prior to committing a crime, or even engage in corporate espionage. Or simply capture ambient audio from unsuspecting people around you.
[. . .]
Once you have root on a Glass headset, any number of custom software packages could be installed without Google being able to prevent one from doing things that would make your hair stand on end, such as on-the-fly image and audio processing.
This is the kind of stuff that until now, only major intelligence agencies could do with very expensive surveillance equipment. Just wait until Israeli and Eastern European startups, which are staffed with former intelligence personnel who have a huge wealth of knowledge in using this kind of technology, get a hold of this thing.