An electrical substation transformer, representing the energy infrastructure needed to power AI computing systems.
What is AI made of?
This activity goes beyond defining AI to explore what is AI actually made of, where the aim is for students to reflect on how the way we ask questions about technology helps us to have deeper, and more critical reflections about it. “What is it made of” provides an interesting complement to questions such as “who benefits?” and “who is harmed?”
Using the PowerPoint slides on the button below, students will investigate a set of images (e.g. copper miners, an active ethernet cable, a data center, legislation on AI microchip manufacturing, screenshots of an X post from Andrew Tate, Python code for scraping data from websites like X, a picture of data labelers in the Philippines, code for neural learning etc.). The aim is to examine the whole sociomaterial spectrum of the AI supply chain, even including an image of a shark and another of an orangutan, who are all implicated in “making AI.” There are alternatives for how to use the cards and ideas for follow up activities.
By the end of the lesson, students should be able to better visibilize the real “cost” of AI and counter the idea that AI is a robot or magical sparkle somewhere in a digital fantasy land parallel but separate to our own.
A more complete vision of what is needed to “make” AI can be found on Kate Crawford and Vladan Joler website Anatomy of an AI System.
…AI is neither artificial nor intelligent. Rather, artificial intelligence is both embodied and material, made from natural resources, fuel, human labor, infrastructures, logistics, histories, and classifications. AI systems are not autonomous, rational, or able to discern anything without extensive, computationally intensive training with large datasets or predefined rules and rewards. In fact, artificial intelligence as we know it depends entirely on a much wider set of political and social structures. And due to the capital required to build AI at scale and the ways of seeing that it optimizes AI systems are ultimately designed to serve existing dominant interests. In this sense, artificial intelligence is a registry of power.
— Kate Crawford (Academic), Atlas of AI: Power, Politics, and the Planetary Costs of Artificial Intelligence, 2021
A fluorite mineral sample, a source material used to produce chemicals needed in semiconductor manufacturing.
Close-up of a microchip with visible circuit patterns and internal structures.
A clean room where technicians work with equipment used in semiconductor manufacturing.
More about the author
Dr. Reed developed this lesson and agreed to host it on the Civics of Tech website. The activity was developed with funding from the Audencia Foundation.
Heidi Reed
This activity was developed by Heidi Reed. Heidi Reed is a socio-legal scholar exploring the business and society relationship. With a PhD in Applied Social Science from the Hong Kong Polytechnic and a JD in Law from Indiana University, her research draws on a multidisciplinary background in law, psychology, and anthropology. Her legal education encompasses American, international, and Chinese law through collaborations with The Hong Kong University. Her work focuses on the ethical dilemmas and value conflicts inherent in corporate responses to societal challenges. Prior to academia, she worked with human rights organizations across the U.S., Asia, and Africa.