Where are the Female Founders in Generative AI?
If AI is the next major wealth creation event, where are all of the women?
TLDR:
(1) Statistically, women are less likely to pursue career paths perceived riskier
(2) Generative AI is a field in rapid evolution, so the path to glory is less clear (than in other sectors of tech)
(3) Most likely, generative AI will be the next major wealth creation event
= We have a problem with representation in generative AI, which is likely to compound existing gender disparity issues
Most photos I see posted online of ‘Generative AI founder events’ have one thing in common: they’re all, or almost exclusively all, male. Perhaps this was just a function of my network or who posted, but not systemic? I wondered about the data: are generative AI startups more likely to founded by / led by men than by women? That’s an easy one: yes, of course, because men are more likely to found startups in general. As has been well-documented, women accounted for just 13.2% of all startup founders in 2023, so we can’t expect generative AI as a subset of the overall tech landscape to look any different.
But, could women be even less represented in generative AI startups than tech in general? To answer this, I dove into the data.
Looking at startups that raised venture funding since 2023, generative AI startups are 27% less likely to have a female CEO than the average funded startup, and this has increased over time.
Why might this be?
We’ve all heard of the statistic that women only apply to jobs that they’re 100% qualified for, whereas the threshold for men is 60%. This originally came from an internal report from Hewlett Packard. This dynamic has been observed in other contexts as well, such as the incidence of speaking up in meetings (less likely for women than men) and asking for a raise (less likely for women than men). What explains these behavioral differences? Initially the assessment was that women lacked confidence. With more study, the emerging consensus was that women misunderstood how to play the hiring game (i.e. what matters is advocacy and relationships, not just skills). Either way, women are less likely to put themselves out there when they’re not sure they’ll be successful.
Could the same thing be happening here?
At a recent dinner in San Francisco with super impressive builders, operators, and investors (all of whom happened to be women), this topic came up: we could collectively only think of three (3) generative AI companies founded by women. It’s not to say we’re foreign to the ecosystem – many among us worked in and around AI… we invested in it, we worked at companies building with it, or worked at companies building it – but we didn’t see women among the leadership pack.
As the stats show us, female founders are a rarified few. Entrepreneurship is at its core betting on the likelihood of a very unlikely outcome. For a whole basket of cultural, social, economic, and behavioral issues, women are less likely to make this leap than men.
Against the backdrop of entrepreneurship in general, generative AI is even less stable ground. More than perhaps any other part of tech today, generative AI is a field in motion. Only 18 months ago ChatGPT was launched and seared generative AI into the public imagination. Since then, billions have been raised and yet there is a short half-life on all of our assumptions. Which model is the best? Generalized models or domain-specific models? Open source or proprietary? Does data quantity matter, or just quality? Is defensibility possible for application-layer companies, or will GPT10 leapfrog it all? Co-pilots or agents? There is no “recipe” for success yet in generative AI.
Here's the problem: Given the forecasts on trillions of dollars of market up for grabs, the winners in generative AI are poised to win big. And, those in leadership roles in the new economy have disproportionate power to shape our future. Will women be missing out? Could the major models, infra tools, and apps companies of tomorrow lack a diverse perspective?
You can’t score if you don’t shoot
The responsibility is on all of us to find stability on shaky ground. We need to support our female founders (especially with the self-doubt that can be inevitable for any one at times), encourage those on the edge to give entrepreneurship a try, and specifically select for diversity in early founding generative AI teams. The reality is few people have the “skills” to be successful in this new AI economy – so most of us are on the same level playing field to learn it. Let’s pull each other in together.
Most alarming is that AI is built in our image, with our data, and with great ethical challenges and issues that we have no clear answer for and that demand a diversity of thought. It’s no longer the ‘right thing’ or a nice-to-have to be balanced and diverse it’s a fast track to disaster if just one group innovates in isolation.
Sarah—well put. There are a number of people (men) who would gladly support females in STEM entrepreneurial efforts (and some of us already do). You won’t hear much or see much of this because we know we are putting ourselves at great risk by our actions.
How? Well you point out that women should take and be encouraged/supported in taking on more risky behaviors just like men do. Over ten years ago in a large company meeting where leaders were discussing diversity and inclusion, the discussion turned that gender equality was not one of the issues the executives needed to address. The company had many female executives and directors and used a merit only based system for advancement. I interrupted—‘hold on! The merit system equates ‘accomplishments’ to being male-like and does not value female accomplishments equally but requires women to be more male-like to have their merits counted.’ The room exploded with a groundswell of support for this idea—and it’s my understanding that the women of this company have now been key to overhauling the system. (Bravo to them.) My arrogance in speaking up a decade ago eventually cost me my position in the company—as I suspected it would when I chose to speak up in the meeting.
Be careful here—if female founders don’t take on risk the same as males, how can the risk structure (in STEM entrepreneurial ventures) be changed so females are more attracted to them and more successful? This is the strategic change needed—we should not be forcing valuable female resources into uncomfortable situations and expect them to perform their best as leaders. It is an unfair and biased expectation.
So how should a female run AI organization be structured?