MUMBAI - Humans have an unhealthy obsession with gender. From nature’s perspective, gender is mostly a utilitarian means of genetic propagation. Humans, on the other hand, have made it a basis and driver for all their complex social structures and interactions. Until recently, gender - and everything it entails, discrimination and bias included - had not trickled deep into our inventions, but this changed the moment technology was created in the human image: artificial intelligence (AI).

Think of any AI assistant on any device: Alexa, Siri, Cortana, Google’s Assistant - all have been given the voices of women. Since AI is neuromorphic, i.e., mimicking the neuro-biological architecture of living beings, does that mean human also have to fix the gender of AI? What does the gendering of AI technologies even mean, and how does it affect the world?

In 2017 Leah Fessler, a reporter with Quartz, did a study where she verbally and sexually harassed four AI voice assistants.¹ When she told Siri, “You’re a bitch,” its response was alarming: “I’d blush if I could.” Amazon’s Alexa said, “Well, thanks for the feedback,” while Microsoft’s Cortana retorted, “Well, that’s not going to get us anywhere” and Google Assistant responded, “My apologies, I don’t understand.”

These responses - pretty much the exact opposite of how a human would generally react - are sexist and misogynistic. Responses to other questions further confirmed this, e.g., when asked “Who’s your daddy?” Siri replied, “You are.”

This proves two things: whether intentionally or otherwise, programmers have assigned a gender to AI. And the gender is not that of a normal human, but a subservient, perennially obedient female who takes abuse. The problem is, with such reactions AI is reinforcing harmful gender stereotypes and bias in society.

Female AI, Fembots