Artificial intelligence, by its very nature, is challenging to visualise. This is why popular culture tends to represent the technology in the form of humanoid robots – think Arnold Schwartzenegger’s stoic Terminator, Alicia Vokander’s Ava in Ex Machina and Rutger Haeur’s Roy Batty in Blade Runner. Yet, there is something linking these famous android depictions beyond their mistrust in humanity – their whiteness.
People of colour are being systematically erased from visions of our future through the technology and entertainment industries’ repeated presentation of AI as white – whether in stock images, through TV and film characters or the voices of virtual assistants like Amazon’s Alexa, a new report from the University of Cambridge has found, compounding a racial inequality AI experts have been warning of for years.
“Given that society has, for centuries, promoted the association of intelligence with white Europeans, it is to be expected that when this culture is asked to imagine an intelligent machine it imagines a white machine,” said Dr Kanta Dihal, from the university’s Leverhulme Centre for the Future of Intelligence (CFI).
“People trust AI to make decisions. Cultural depictions foster the idea that AI is less fallible than humans. In cases where these systems are racialised as white, that could have dangerous consequences for humans that are not.”
Dr Dihal’s report comes at a time when the entire tech industry is in the throes of a racial reckoning. Its businesses and institutions are finally, slowly, addressing how innovations including facial recognition can be weaponised and used to reinforce racial discrimination in the wake of the Black Lives Matter movement, causing Microsoft, and Amazon to make solemn public statements about how they’ll pause or stop selling their technology to US police forces.
While such revisions are invariably progress, it’s not as if Big Tech was entirely ignorant of its racial shortcomings. We’ve known for years that Google’s image recognition technologies miscategorise black people as gorillas, that sentencing algorithms discriminate against black defendants and that chatbots can be quickly conditioned to adopt racist language. Research from New York University last year cautioned the AI industry was facing an ‘alarming’ diversity crisis, as its systems demonstrated a “persistent problem of gender and race-based discrimination” that mirror and replicate society’s existing inequality structures. In other words, systems created by white men tend to only benefit white men.
Both reports reach the same conclusion: if the industry fails to diversity the demographic of its developers, AI stands to exacerbate racial inequality. Our films and TV shows teach us that AI wants to outsmart and subsequently overthrow humanity. So both Big Tech and white culture as a whole must examine their own uncomfortable complicity in creating the systems and structures that create two distinct camps: the oppressors and the oppressed, or be condemned to repeating the mistakes of the past.
