Why are most AI "Female"? An analysis of patriarchal social gender coding in artificial intelligence helper bots.
Computer technology and Artificial Intelligence have been slowly becoming more incorporated into the intrinsic aspects of many people’s daily lives. From our mobile phones, our personal computers and tablets, in our GPS that guide us on our journeys, in our daily interactions with our technologies through to social media and helper bots like Alexa, Siri and their derivatives, our lives have become intimately entwined with technology.
There is not one aspect of our daily lives that is not touched or impacted upon by the explosion of cognitive computing and machine learning that underpins the workings of our interactions with weak artificial intelligence that currently exist today. How we as a species, adapt to the changes we have wrought on ourselves through our own technology, and our survival in the face of the coming strong AI, will see if we survive to be the fittest, or if our place will be taken by another intelligence, with just as much ownership to this planet and its resources as we claim as ours.
Advances in medical technology and biotechnological constructs such as artificial hearing with bio sensors and feedback loops, artificial hearts, that can mesh with the bodies own electrical systems and augmented limb replacements that learn from the bodies that they integrate into, have allowed people to live longer in bodies that are no longer only flesh and blood, but something more than both. As machines and humans merge together, and the demarcation of what is machine and what is human becomes blurred, it is necessary for us to ask where AI is heading?
Is AI really ushering in a new world where gender, race, and sexuality can possible be rendered obsolete, or are we just destined to perpetuate the the same patriarchal stereotypes of the old world? Will our systems and institutions cope with the new emerging paradigms of humanity, or will the gatekeepers of our social systems, hold firm to the archaic paradigms that enforce the biological essentialism that genders what has none.
Ferrando (2014, p1) writes in her case study on cyborgs, robots and artificial intelligence that, unlike humans, machines are genderless and have no ethnicity, because they are “free of the boundaries of gender difference. This perception is common within the field of Cybernetics, “since AI operates out of the sexual paradigm, the notion of gender has become obsolete” (Ferrando (2014). However even in her statement about the obsolescence of gendered technologies, she also quotes Katherine Hayles, who says “The body is the net results of thousands of years of sedimented evolutionary history, and it is naïve to think that this history does not affect human behaviors at every level of our thought and action” Ferrando (2014, p4). Are we stuck with our biological imperatives or are we capable of bootstrapping ourselves up and beyond these biological constraints?
But in appreciating our evolutionary baggage and its impact on our how process and classify, we must be aware of these boundaries and think outside of our evolutionary inherited paradigms that construct our social gendered world, so that we can be open to the possibilities of overcoming our evolutionary limits to this mode of thinking. If we can consider and accept the possibility of a genderless technology, then we will be capable of not just constructing new paradigms, but of also deeply understanding our roles and that of our technologies in shaping these new modalities of being. By opening our minds our bodies and our institutions, we discover new ways of understanding the technology that inhabits our shared spaces and that new ways of looking at gender will play out through the development of Artificial Intelligence and cybernetics.
But the reality of things looks rather different. Developing methods to analyse and read the contextual frameworks that impact and surround how gender is being incorporated into AI, means looking at who is developing Artificial Intelligence Systems, who is the Artificial Intelligence bot designed for, what is the ethos of its creation, and what roles are these AI intended to fulfill, in both the desired and unknown consequences for its creators and its consumers? Questions that desperately seek answers as our society grapples with what defines human in a world where that is harder and harder to define.
Will gender in the age of AI really become obsolete? Or rather, are the stereotypes of gender, being incorporated into the AI, perpetuating the notions of gender and race as we currently know them?
Helper bots like Siri and Alexa, and other bots being developed in fields traditionally dominated by females, like nursing, child care, and education have all been created and developed with a female persona Ferrando (2014). From their physical attributes, through to the lilting tones of their female voices, female AI’s are being created to assist and help people in a variety of ways Araujo (2018). Research has shown that AI’s whose role is to help, and assist are likely to be given a female voice, and mannerisms, this gendering of technology is not just driven by those the develop the technology, but also by those that consume it, as research has consistently shown that there is a demand and a requirement from end users of these products that the preference is for gendered technologies and for those genders to be female, with research showing a very clear demand from consumers for this gendered technology, especially in those roles which have been historically and socially relegated to the work of women Bogdanovyvh, Trescak and Simoffet (2016).
Alexander & Yescavage (2018, p 82), in their analysis of the film “Her” argue that in the film the artificial intelligence behind the helper application called “Sam has a female voice, which the protagonist Theo voluntarily chooses, but the gendering of the AI’s voicing is just a choice and nothing more as the AI need not be fundamentally male or female”. “More importantly, the artificial intelligence Samantha is pitched to us the audience and to our central protagonist Theo for that matter as a care-taking device, and she is much more often than not focused on Theo’s needs and emotional well-being: her role, not just her voice, is gendered as traditionally feminine, the maternal caretaker”. Alexander & Yescavage (2018, p 82).
In Film Noir’s “Ex Machina”, Eva, the female AI is a sentient being created by Nathan. “In this instance, Nathan has chosen to give sexuality to Ava, claiming that the evolution of any species requires sexuality” (Alpert & Garland 2016, p 4). In both films, the AI is female and is created by a male for his own purposes. While both these films are fiction, the ideas contained within them, and the stereotypes about women they employ, speak volumes about the reality of AI development, and the biases that exist currently Alexander and Yescavage (2018 p 22). This bias towards imposing historical and social roles on our technology both limits or abilities to evolve with our machines, but at the same time, protect us as a biological species from something that even as we create it, we fear what we create, will eventually if not destroy us, render the question of “what is human?” to be irrelevant Alexander and Yescavage (2018).
When it comes to the growth in the sexual robot industry, the sexual AI’s are almost universally female. From realistic genitalia and soft to the touch hair and skin, these female sexual robots are created to fulfil the sexual needs and desires of heterosexual men Naff (2017). However, we see a fine line being developed as to how far in humanness can the creators of sex bots go, without crossing the uncanny valley into representation of us that skate to close to our realities and generate fear with us as a species Naff (2017).
As with the development of helper bots, sex bots’ creators are likely to be, young white cis heterosexual men, which leads to a very narrow expression of both helper and sex bots, where the limitations of gendered social roles and the implicit problems in the patriarchal system which works actively to diminish the contributions of women and raise the default status of human, to that of cis white straight male are projects and multiplied by the people who create them, the people buying and using them, and the stereotypes surrounding females and the sexual roles they occupy, means almost all sex bots are created “female” Cockayne, Leszcynski and Zook (2017).
The dominance of mostly white, heterosexual men in this field, means there is little to no input from women, other ethnic minorities and people from diverse backgrounds and sexualities Ferrando (2014). This lack of representation in what we are told is a technology that will have deep and long lasting changes to both our society and our ways of thinking, is deeply disturbing to researchers, that see this as the continuation through technological entrenchment of the historical oppression of women and to the establishment of the default human as male, white and straight, and this will be our technological future, suppression of gender via our own tools meant to free us Ferrando (2014).
Part of the solution to this gender and cultural bias that exists in the AI industry is to encourage and enable people of different genders, sexualities, ethnicities, abilities and cultures to access education and training, that allows them to engage in the development process in a real and meaningful way. In this way, the conversation surrounding issues like gender and race in the AI field of development, will include many voices and viewpoints. This diversity of talent, opinions and perspectives, will lead to a much more open, and lively debate, and new ways of thinking about AI creation and development Ferrando (2014, p15).
As our technology advances at an ever increasing speed towards what we are told is the Singularity and the birth of strong AI, as we continue to merge silicon and flesh our cybernetic dreams of 50’s and 60’s Science Fiction as told to use by white heterosexual men becomes a reality they only dreamed about sixty years ago, we need to ask ourselves some very fundamental questions, questions that I fear we are not morally or socially advanced enough to understand let alone answer, are we socially evolved enough to live in harmony with our creations? Can we look beyond our own humanness and the weakness that comes from being biological, and understand that life can take many forms, and that our advanced tools will demand consent for their use, are we ready for this?
These questions will require much soul searching and some radically new paradigms. As we look to the future of humanity and technology, we need to look to our shared social and biological history and understand, that the future begins with the seeds of the past. We need to develop a new frame of reference when thinking about gender, race, culture and social norms, if we are to escape the confines of thinking that more than ten thousand years of evolution has bred into us, if this technology is to benefit all humanity, rather than just the privileged few Buran (2015).
Alexander, J & Yescavage, K 2018, 'Sex and the AI: Queering intimacies', Science Fiction Film and Television, vol. 11, no. 1, pp. 73-151.
Alpert, R & Garland, A 2016, 'Alex Garland’s Ex machina: the gender of artificial intelligence and the triumph of enlightenment', Jump Cut, vol. Fall 2016, no. 57, pp.
Araujo, T 2018, 'Living up to the chatbot hype: The influence of anthropomorphic design cues and communicative agency framing on conversational agent and company perceptions', Computers in Human Behavior, vol. 85, pp. 183-189.
Bogdanovych, A, Trescak, T & Simoff, S 2016, 'What makes virtual agents believable?', Connection Science, vol. 28, no. 1, pp. 1-26.
Buran, S 2015, 'CORRESPONDENCE BETWEEN CYBORG BODY AND CYBER SELF', Journal of Research in Gender Studies, vol. 5, no. 2, pp. 290-322.
Cockayne, D, Leszczynski, A & Zook, M 2017, '#HotForBots: Sex, the non-human and digitally mediated spaces of intimate encounter', Environment and Planning D: Society and Space, vol. 35, no. 6, pp. 1115-1133.
Ferrando, F 2014, 'Is the post-human a post-woman? Cyborgs, robots, artificial intelligence and the futures of gender: a case study', European Journal of Futures Research, vol. 2, no. 1, pp. 1-17.
Naff, C 2017, 'THE FUTURE OF SEX', The Humanist, vol. 77, no. 4, pp. 12-19.
