Why AI needs the feminine

The necessity of more women in AI is about much more than workplace diversity.

Lately, I have been spending a lot of time around men. Whether I’m on Stanford’s campus discussing AI’s societal impacts or in a Zoom room learning about AI safety or hosting an AI Salon event in San Francisco, I am often mildly surprised to find myself outnumbered by the opposite sex, a sensation that feels new to me. I’d always been insulated by the tech industry’s gender discrepancies, having spent my 10+ year career working in tech marketing and communications — departments that tend to skew female, and where I never existed in a gender minority. 

But in the AI world, I am nearly always surrounded by men. I am enveloped by kind men who teach me things; who elevate my voice and encourage me to claim my role in leading AI conversations. I work with men who have founded AI startups and men who are chasing things and men who dream of changing the world. I interview thoughtful men who have concerns; I speak to overzealous men who have none at all. There are men who all rushing, all pushing; there are men who are pure reflection, who probe the evolving AI landscape with logic and reason.

The prevalence of men in AI shouldn’t surprise me, but somehow it does. A startling 78 percent of AI employees are male — meaning that 22 percent of AI employees are female. This isn’t to say that there aren’t women in the AI space — the illustrious Godmothers of AI made pivotal contributions to the progression of artificial intelligence, and, interestingly, in the fields of AI safety and Responsible AI, I have found there to often be more women than men working to passionatly reign in technology that’s gone unchecked.

While it’s promising that there are so many women interested in AI safety, only 12 percent of AI researchers are female — meaning an overwhelming majority of the individuals actually building and launching artificial intelligence systems are male.

This disparity will have far-reaching consequences beyond the issues of workplace diversity and wealth disparity, although those problems also need to be addressed. For example, the AI market is projected to reach 184 billion dollars and so by nature of the aforementioned ratios, approximately 80% of that wealth will go to men — further widening the wealth gap between men and women in what could be the biggest wealth explosion in human history. 

But there’s another reason, one that is more philosophical, more opaque. It requires us to think more deeply, to consider that the AI we are building is, in essence, a reflection of our society — of our values, of our laws, of our energies. Of our standards and of the way we approach our own morality. Of our attitudes towards ourselves, towards one another and towards the Earth we call our home; of our delicate and exact humanity.

Data, algorithms and computing power are the essential building blocks of AI. Each piece — and how they come together — requires hundreds of decisions to be made by the people building it: what training data should be used? How is the data sourced? What training methods should we employ? How will bias be detected and mitigated? Will the system comply with ethical frameworks?

Every decision is made by a human and — as we’ve seen — nearly 80 percent of those humans are male. This suggests AI will reflect the perspectives, histories, traumas, and wounds of men. It will be built to align with incentives that are important to the male population, to reflect the future that they hope to see. 

Why is this a problem? It’s not a problem that the technology is being built by men — it’s a problem because it’s being built by only men. Murat Durmus, author of Mindful AI, wrote on this subject:

“When women comprise half of the world’s population and are not part of the teams building AI, the technology lacks the depth and clarity from a comprehensive, balanced view of the world. This absence is not just a loss in numbers but a significant deficit in perspective, which leads to AI systems that are less effective, less fair, and, ultimately, less human.”

The masculine and the feminine

Without the feminine influence, we risk building AI that skews towards more “masculine” tendencies. Contemporary spiritual rhetoric preaches the notion of balancing “masculine” and “feminine” energies — the idea being that we all carry a dualism of both masculine and feminine energies, regardless of self-identified gender. The masculine and the feminine are the opposing — yet complementary — energies of Yin and Yang, the sun and moon, the left brain and the right brain.

Source: Pause Meditation

While the masculine is assertive, structured and logical, the feminine is emotive, intuitive and nurturing. Where the masculine is efficient, the feminine is compassionate. The masculine practices focus, while the feminine practices playfulness. Masculine energy is more stable and predictable, whereas feminine energy is more flowing and dynamic.

We must proactively ensure that AI is being built to strive towards equanimity between the masculine and the feminine. And while the (primarily) men who are building AI contain both masculine and feminine energies, it’s difficult to achieve equanimity within AI if men aren’t connected to their feminine natures. Our society isn’t set up to help men lean into their feminine — so I worry about the male capacity to ensure both the masculine and the feminine energies are reflected in an intelligence that may one day simulate consciousness and become sentient.

It’s likely that one day in the near future, AIs will coexist alongside us in a new kind of human-AI frontier. We will partner with AIs to go about our daily routines: we’ll work together, cook together, walk our dogs together. But AIs will also be integrated into the big things, the things that really matter. They will participate in the raising of our children, the teaching of our students, the fighting of our wars. How much power we give to AIs in these actions is up to us, but it’s very likely we’ll outsource more of our decision-making to AIs as the systems become more powerful and more integrated into our society.

This is why we need more women in AI today. We need an equal representation of women not only on ethics and safety boards, but also in the labs, in the startups. We need more women participating in the decisions that will dictate the kind of intelligence we will live alongside, who feel supported to share their voices, and who can embed their intuitive and creative natures within the technology that will play a major role in the lives of our children. We need artificial intelligence that is reflective of both the masculine and the feminine — AI that is as innately nurturing as it is assertive, as compassionate as it is efficient, as comfortable with “being” as it is with “doing.”

We must build AI that is reflective of all of our parts. The tender parts and the strong parts. The parts of us that yearn for oneness, for peacefulness, for solidarity with one another. For tranquility between and within nations. For the silencing of wars, for the saving of our children, for the building of a harmonious future. This is only accomplished through ensuring equal of representation of not just men and women — but through democratic representation of all of our unique voices.

Are you a woman working in AI? I’d love to hear about your experience. Tell us in the comments. 👇

Until next time,

Cecilia

---

RemAIning Human is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

---

Building community around an essential topic

Know someone whose interested in learning more about AI? Forward on this email or click the button below to share.

---

RemAIning Human is written and edited by Cecilia Callas.