Like many people, director Valerie Veitch became interested when OpenAI first released its Sora text-to-video generative AI model to the public in 2024. Although she didn’t fully understand the technology, she was curious what it could do, and she noticed that other artists were creating online communities to share their new AI creations. The hope of connecting with people drew Veatch into the AI field, but once she got there, she was shocked to see how often the technology generated images laden with racism and sexism.
Veatch was even more troubled by the fact that her new AI-enthusiast peers didn’t seem to care that the machine they were working with was spewing hateful, bigoted garbage without being explicitly prompted to do so. The strange situation drove Veatch away from his early experiments with General AI. But it also inspired him to create Ghost in the MachineA new documentary about the technologies and schools of thought that laid the groundwork for General AI’s existence.
Instead of focusing on the potential (if highly improbable) benefits to society that General AI accelerationists swear are just around the corner, Ghost in the Machine It explores the history of technology to explain why it works the way it does now. When I recently spoke to Veatch about the film, she told me that she wanted to describe the origins of Generation AI to give people a clearer idea of the intense cycle of industry hype we’re currently living through. However, first, it had to overcome the purposeful obfuscation of the entire concept by AI firms.
“In order to use the phrase ‘artificial intelligence,’ we need to know what that phrase means,” Veatch told me over a video call. “The truth is that it doesn’t mean anything; it’s a marketing term and always has been. It’s a completely confusing, stupid phrase that has taken on its own cultural meaning, and I think it’s important to be really clear about the words we use and the meaning of those words.”
As Ghost in the Machine It is repeatedly stressed that “artificial intelligence” was originally coined by computer scientist John McCarthy in 1956 when he was trying to raise more funding for his projects. But the documentary presents the coinage of the term as one of several key points in a timeline that actually begins with the birth of eugenics in Victorian-era England. In addition to being Charles Darwin’s cousin, Francis Galton was the originator of eugenics – the racist and discredited notion that humanity could be improved through the systemic elimination of “inferior” (read: non-white) races.
While Galton definitely created Some? Useful Contribution to Academics In our interview, Veatch pointed out that it is important not to minimize the fact that his deeply held white supremacist beliefs informed the social sciences of the era. Galton and his fellow eugenicist/protégé Karl Pearson were not directly involved in the development of early computational machines. But Galton’s seminal work with multidimensional modeling—a technique he used to measure the attractiveness of African and European women—shaped Pearson’s thinking as he developed statistical tools like logistic regression, one of the foundational components of modern machine learning.
“Am I going to hug Sam Altman on camera? Is this a true film about this technology? This is propaganda.”
galton pearson helped generalize the idea People of different races were fundamentally different in quantitative ways. This type of racist thinking led Galton and his colleagues to believe that human intelligence could be measured, and that human brains worked exactly like machines. Veatch says that leap played a major role in selling the hypothetical idea of artificial intelligence to the public.
“What was really surprising to me during my initial exposure to all this was that when you look at the question of superintelligence as a documentarian or journalist, it doesn’t take very long before you get your head slammed on the bottom door of race science because it’s baked into this technology,” Veitch said, explaining that these concepts are “soaked” in eugenic thinking.
Rather than trying to refute the idea that general AI models generate hateful ideology because they have been trained on it (a concept commonly known as “GIGO” – garbage in, garbage out), Ghost in the Machine It uses its historical analysis to explain why the companies creating this technology seem so disinterested in addressing its current issues. This historical context helped Veatch understand some of his own troubling experiences with General AI when he was playing with an early version of Sora in the cast’s slack. Veatch remembers the group as a friendly, welcoming place, until another member – a woman of color – began expressing concerns about the way models approached her each time she created images based on their photographs.
“It kept her braids and it kept her fashion, but she was taking herself to an art gallery, which the program deemed ‘white space,'” Veitch explained. “My reaction was ‘What the heck,’ and I tried to explain to the group how this was actually a problem with the software itself.” No one else in the group linked to his post. “This was a Slack where, normally, there are always dozens of screaming koala emoji reactions to every post. But this time, there was nothing.”
Veatch took it upon himself to contact OpenAI directly and alerted the company to “how racist, sexist, and misogynistic the outputs[she]were seeing were – outputs where women would start growing extra breasts and walking around after two rounds of generating a scene.” Veatch thought that OpenAI would view this as a significant bug that must be fixed before more people could be encouraged to adopt Sora into their lives; Instead the company ignored his concerns.
Veatch recalled, “The response I got was basically, ‘It’s too hard; there’s nothing we can do to change it.’
That situation lit a fire within Veatch to learn why so many different forms of generative intelligence consistently behave in such ugly, disturbing ways. At first, she didn’t really think that Zoom calls with authors of white papers about the technology could be turned into a compelling documentary, but that changed when she started to see a clear line from Galton’s eugenic statistics work to the modern generation of AI organizations.
Voices included Ghost in the Machine – a mix of AI researchers, historians, and critical theorists – makes a compelling case that basically every aspect of the AI field has been deeply influenced by its historical relationship with fields of science created to support discriminatory world views. When I asked Veatch if he’d ever been interested in talking directly to the heads of companies Ghost in the Machine Takes it to work, she laughed. To get that kind of access, he said, he would have to go through all kinds of conceptual gymnastics and make compromises that would make his film complicit in the loss of General AI.
“The idea is, you know, these guys won’t trust anybody,” Veach said. “Yeah, no nonsense, and I certainly hope they don’t believe me. I don’t want them in the movie and they already talk enough in front of the media. Am I going to hug Sam Altman on camera? Is this a true movie about this technology? This is publicity.”
Ghost in the Machine will be available for Stream via Kinema from March 26 to March 28, before it airs on PBS sometime this fall.
Correction, March 21: An earlier version of this story used the phrase “logical regression” when “logistic regression” was correct.
