You will be redirected back to your article in seconds

“A lot of computer scientists are just in the pixels of what they’re doing and not really connecting to communities that could be harmed by these practices,” says Shalini Kantayya, director of “Coded Bias,” a documentary that explores racial bias in facial recognition software, algorithms and artificial intelligence.

Kantayya joined editor Meredith Woerner for Variety‘s “Doc Dreams,” presented by National Geographic, to discuss the making of the groundbreaking documentary and the revelations that unraveled while making the film.

“Coded Bias,” which was shot in five countries and features over 25 interviews, starts with a discovery from Joy Buolamwini, a computer scientist and digital activist at MIT.

Through her work, Buolamwini revealed that a commonly-used, computer vision software system wouldn’t track her facial features until she put on a white mask. “That’s when I started looking into issues of bias that can creep into technology,” Buolamwini says in the film. Kantayya uses that discovery to dive into the many problematic layers coded into our worldwide “digital destiny.”

Kantayya notes that while we think of algorithms as “future-gazing” technologies, they are based on data from the past, which is often ridden with bias and systematic inequalities. When that data is used to predict behavior in the future, it can “mechanize” racism.

“I didn’t even know what an algorithm was when I began this film,” Kantayya says. “I think everything that I knew about artificial intelligence came from the imagination of Steven Spielberg or Stanley Kubrick or Ridley Scott. I’m definitely someone who’s watched ‘Blade Runner’ way too many times. I don’t think that my background as a sci-fi fanatic prepared me for the realities of what AI is, what it can and cannot do and how it’s being used as this invisible gatekeeper of opportunity in our world.”

The film shows how artificial intelligence can have detrimental effects when the systems that help determine the result of someone’s job or housing application, quality of health care and even the length of prison sentences aren’t screened for racial or gender bias.

“It’s like getting a pharmaceutical product with no counter indications, no usage on the label,” says Kantayya. “It’s like the wild wild west.”

On the positive side, the film is resonating with both viewers and those in the marketplace of big tech.  “Almost every major tech company has hosted a screening of ‘Coded Bias,’ which to me, I never anticipated ever” the director reveals. However missing from the list of companies hosting premieres is Amazon, which the doc puts under the lens for their past and present dealings.

“My mantra is we can’t leave the tech bros alone on this,” Kantayya says. “They need to hear and feel the ground swell from the public. Since the film premiered at Sundance, three of the largest tech companies in the world have changed their policies around selling facial recognition. IBM got out of the game, disrupted their entire business model, closed up shop they’re done with facial recognition. Microsoft stop selling to police and Amazon said they would take a one-year pause, of which we’re good for two more months.”

Kantayya puts that change squarely on the shoulders of the brave scientists featured in her doc, unencumbered by corporate interest.

However Kantayya doesn’t view big tech as the enemy, instead she hopes to change the conversation and help improve literacy about these exceedingly important algorithms. “I really see algorithmic justice as the unfinished work of the civil rights movement and that it will happen in quite the same way with our voices being heard. I think that’s what helped propel the movement forward and helped inspire these companies to do the right thing.”

“Coded Bias” is currently available on Netflix.