By choosing to look at the funding from the American Government on this field, I aim to tell a different story about AI.
A quick search for the word “librarian” on Google reveals images upon images of women holding books amongst big shelves, attending to patrons, reading stories for children, or stocking book shelves. Librarian is one of those professions that, like many others, such as nurse and secretary, have been associated with the female world.
If this text is about AI, you might be asking why I’m writing about libraries and librarians–but as scholars Safyia Noble (2018) in her Algorithms of Oppression and Monica Westin (2023) more recently have shown, what most people in Western countries usually understand as the internet, and what fuels the data collection of digital information that feeds generative artificial intelligence (AI) such as ChatGPT, was first started in the 1970s by groups of librarians in different universities across the US who designed systems for distributing knowledge through online networked computers–what back then just meant establishing a real-time connection between a mainframe computer and a terminal.
Women, who were called computers themselves, used to also work at NASA and other agencies, calculating complex numbers and routes that would eventually take humans to the moon, as shown by scholars Catherine D’Ignazio and Lauren F. Klein (2020). Sociologist Anthony Elliott (2021:49) writes that in the early 80s, US governmental agencies shifted their funding away from rockets and radar towards supercomputers, machine learning, and AI. The period of the cold war is when the US Department of Defense (DoD) first started funding research groups across the US on the then new field of AI. Back then, as it is now, AI was considered a question of national security, and as such, the American government made sure that through the US DoD, funding was getting to the best research groups at the time.
The term artificial intelligence had been created years before in 1956, when a small group of male scientists met at Dartmouth University for a summer research project on AI. Those early AI researchers came eventually to be known as the “forefathers of AI”–in the Americas, at least, since in Europe, Alan Turing had already experimented with replicating the human mind in machinic systems.
Computer scientist Melanie Mitchell (2019) has written about what has informally been known as the AI winter: when the US DoD stopped funding AI projects for short periods of time during the 70s and 80s, especially because research groups were overpromising and couldn’t deliver what they said they would. But as the researcher Elliott (2021) has also shown, these AI winters were often short-lived, especially because other AI research groups were soon able to convince the Department to fund their projects.
The privileged focus often given to the events in Dartmouth and even to Turing’s “imitation game” is one of the many possible ways of telling stories about AI. By choosing to look at the funding from the American Government on this field, I aim to tell a different story about AI. A history that, as feminist theorist Sara Ahmed (2017) writes, brings to the forefront the conditions of existence of this specific reality. The case of AI is just like the history of the internet, as Harvard’s historian of science professor Dr. Naomi Oreskes (2023) recently published, and just like that, when it comes to AI, “it was the government and not the private sector that took the initial risks. The key participants were not disruptors”, but they were “seasoned professionals working inside established institutions.”
Still nowadays, AI research and development in the US is largely subsidized by the country’s government. The US Department of Defense still directs millions to research projects, especially ones with real-world applications and projects that help develop and modernize the military, while the National Science Foundation, with some of its focus on academic research, also funds experimental projects that do not always produce technologies that would be conventionally applicable to reality. But if that’s the case, why, when we think about AI, do our thoughts go directly to Silicon Valley? Why do we think about venture capital firms? And CEOs like Elon Musk, Mark Zuckerberg, Jeff Bezos, and others?
The interest in telling stories that give priority to individual efforts reveals a narrative that prioritizes meritocracy. Discourses focusing on solitary disruptors developing new technologies aim to generalize the idea that innovation is only possible in a private context and that the government is lagging behind and can’t keep pace with modern transformations, especially in the tech sector. This discourse also aims to avoid and criticize regulation since the industry allegedly needs freedom to experiment with products and technologies that, more often than not, cause more harm than good to society. But only when it’s in the CEO’s and disrupters’ interests because when it’s not, they can all come together to request an AI moratorium where no one can develop technology in the field. Stories that focus on the success of single individuals are a way of not disclosing that innovation can develop in many contexts, as Dr. Oreskes (2023) writes.
In the same way that the history of computation has completely erased women’s contribution from its past, obscuring the funding of this field that is now associated with Silicon Valley and the private sector is a way of pretending every technological development is only the fruit of individual efforts of single individuals, mostly men, who can save the world from problems they’ve created themselves.
In a way, what we now understand as AI is the fruit of the collaboration of different persons, nationalities, and ways of thinking, being, and existing in the world. AI couldn’t exist today if it wasn’t for the essential funding from the US government through its different agencies, which have considered AI since its inception in the last century as a strategic field to the country.
Feminism urges us to bring to the forefront the conditions that make our existence possible–looking at the history of state funding of AI is a way of acknowledging that today’s AI industry would not be possible if it weren’t thanks to government funding and that this field ultimately does not exist thanks to the work of a single man or investor.
Ahmed, S. (2017). Living a feminist life. Duke University Press.
D’ignazio, C., & Klein, L. F. (2020). Data feminism. MIT press.
Elliott, A. (2021). Making sense of AI: Our algorithmic world. John Wiley & Sons.
Mitchell, M. (2019). Artificial intelligence: A guide for thinking humans. Penguin UK.
Noble, S. U. (2018). Algorithms of oppression. New York university press.
Oreskes, N. (2023, June 28). “The myth that may have doomed the Titan.” The New York Times. https://www.nytimes.com/2023/06/28/opinion/titanic-titan-oceangate-innovation.html.
Westin, M. (2023, June 5). “The 1970s librarians who revolutionized the challenge of search.” Aeon Essays. https://aeon.co/essays/the-1970s-librarians-who-revolutionised-the-challenge-of-search.