In a world where all humans are increasingly dependent on machines to tell us what to do, where to go, how much to pay, and so on, it is time for us to pause and question the same machines that have begun to define our very existence. Who is behind the machine? How much do they know about me? To what extent am I being tracked?

One of the creepiest things I’ve encountered in the past few years are the suggested or sponsored ads on Instagram and Facebook. I would have just had a casual conversation with someone and the next moment I pull up Instagram, I’d see ads based on that. What a stalker social media can be! ProPublica’s short episodes on Machine Bias: Investigating Algorithmic Justice outline how diverse targeted groups can be based on parameters, mere data, resulting in biased ad-serving algorithms. 

The Critical Computation reading for the past week was incidentally sociologist Ruja Benjamin’s Engineered Inequity: Are Robots Racist? which also talks about bias in AI, data and machines. One of the examples mentioned in her work that I found truly shocking were soap dispensers that did not work for dark skinned people. Instead of using motion detectors to track the position of the hand, the device used infrared technology which required light to bounce back from the user to activate the sensor. This might seem like a gross technical oversight and not intended discrimination, but how did the final product even pass the prototyping and testing stage? It brings up the question – whom did the testing group include, and more importantly, whom did it not?

The Microsoft project, MS-Celeb-1M, is interesting to see how fast data is circulated and downloaded on the internet in such a short span. Once it’s out there, it is out there for posterity. I once created an online portfolio on carbonmade.com with some old (and embarrassing work). I’ve tried taking these images down since but even though the site is down, they still frustratingly come up on google images!

Instances like these also help shed light on what kind of designers we must be. There is a fine line between a select target group and discrimination, and this line largely depends on the context of the design product. Though I’m guilty of being indifferent toward bias through design so far, I realise that my work needs to be more inclusive, culturally sensitive and consciously developed.