Fooling face recognition algorithms

Computer systems built to represent human identities have contained with them many ontological assumptions about what it is to be an individual and what personal identity is. These systems define the human as a “what” ie: that which can be measured, not as a “who” ie: our inner self.

If the state of the art in computer science can produce a unique feature that describes an individual as such, what good does that do the individual if this knowledge is only leveraged against them?

Interesting paper that attempts to trick state of the art face recognition algorithms in recognising weird looking shapes as human faces.

The sub text being that technological identity and self identity are very different things; you are just not your fingerprint, retina print or your face etc. (or even all of them combined).