Machine Vision - a black intersectional method of inquiry

Iris Scanners and Recognition — Biometric Identification Techniques

The use of biometric information is becoming increasingly common around the world. In a political, economic and social climate where media circulates fear of terrorism and refugees, it does not take a Black Mirror episode to uncover the potential for this technology to be used and exploited in ways that are scary. It brings to mind the use of eugenics in Gattaca, and how it created a new hierarchy utilising gene mapping to privilege some and discriminate against others. From a black intersectional analysis viewpoint, it is clear that this technology will almost definitely be used to police black bodies, and the bodies of people of colour, in a more rigorous, potentially dangerous and harmful way. Consequently, there are many questions that require our attention — who is creating this technology and to what end? Are the people creating it aware of or critical about the ways in which it will be used? Does an individual have any power in not using such technologies? It is clear that this will disproportionately affect the Global South in a concerning way, especially as borders, migration and immigration are more politicised and policed than ever. It will mean it is harder for people to move around this supposedly globally connected world, it will mean that people of colour, people from lower socio-economic backgrounds, people further away from white, cis, hetero, able, wealthy and male will face, at the benign end, inconvenience, and at the most extreme, persecution, imprisonment, and potentially death.

http://www.m2sys.com/blog/guest-blog-posts/iris-scanners-recognition-biometric-identification-technique-airport-security-systems/

*

Machine Vision that analyses artworks

Analysing artwork seems a more innocuous way to use machine vision. This specific example used a training set of 80 000 artworks to build AI that could relatively accurately decide which artist a painting was created by. Using a black intersectional lens, this can be approached with many questions and critiques that could be angled at the internet and technology more widely — who makes these tools? Is a true cross-section of humanity taken into account when thinking about the history of art? Without even knowing the details about this training set, I would guess that of the 80 000 training images, very few, if any, would be from indigenous art practices from Africa, Australia, Asia, North and South America. Like the people who create the technology we use every day, it is highly likely that the people who created this AI come from an anglo-European-centric viewpoint, analysing art from a tiny slice of art history, which also happens to be a huge proportion of documented art history. Erasure of history, practices, languages, culture, people, and so much more, through colonisation, has been and will continue to be woefully common, however the internet does offer an opportunity for those who have been silenced to document their history, their experiences. The internet is at once a perpetuator of this silencing, as well as a vessel for voices.

https://aimatters.wordpress.com/2015/05/12/the-machine-vision-algorithm-beating-art-historians-at-their-own-game/

MachineVision.jpg