Pamela Tudge and Ketra Schmitt
Technological determinism (TD) is the idea that technology alone shapes society, and by extension, history. Science Technology Studies (STS) and the many disciplines that intersect with it have conclusively rejected TD. Social Construction of Technology (SC) suggests the opposite of TD—that society influences technology and that ultimately technology and its impacts are socially constructed. At the most basic level, technological ideas emerge from organisms that are embedded in and supported by society. Much of current STS scholarship involves refining and interrogating SC within specific technological constructs.
TD nevertheless persists in the minds of many. Inspired by Gregg Zachary’s piece in this issue [1] on artificial intelligence (AI) and TD, we decided to write about how AI engages capital and the role that technodiversity could play in resisting this impact. Zachary interrogates the deterministic thinking often articulated among those who explain or defend job loss. He evokes decades of STS scholarship that both debunk TD and look for ways to better communicate this to students. Zachary focuses on AI and determinism because TD is a “sticky” belief that is hard for many to reject. SC, in contrast, is a lot less sticky—despite being the prevailing theory among scholars.
Zachary draws our attention to the “PR problem” that plagues social construction. Unlike TD, SC is not a story that we often hear in public narratives about technology’s impacts. As his essay demonstrates, communicating SC is an urgent issue, and one that contributors to and readers of IEEE Technology and Society Magazine are well-positioned to address.
While wealth inequality and governance are beyond the scope of this article, we note that the same mechanisms of profit and power are at play in the development and deployment of AI tools. Perhaps, it is more precise to say that the mechanisms of profit could be at play here, because so far, AI companies have not generated profits. Rather, massive amounts of capital and environmental resources are being poured into this technology, with no realistic prospect of profit in sight. The most likely path to profitability lies in removing jobs from humans and replacing those tasks with AI tools. The results have been decidedly mixed, varying from rehiring fired workers to chaotic deployment of untested tools [2], [3]. Other firms have integrated AI into workflows to increase productivity with existing staff. Emerging findings indicate that AI-integrated workflows do not reduce work at all [2]. However, reports indicate that companies often make workforce reductions on the strength of promised efficiency, or to maximize profits [4], strengthening the evidence that AI-induced job loss has more to do with the concentration of wealth and power than the technology. Thus, the AI profit conversation is speculative at best, at least when it comes to shareholders.
What is not speculative is this: AI firms have harnessed vast amounts of capital from other technology companies, and they have used it to create a hugely disruptive technology.
What is not speculative is this: AI firms have harnessed vast amounts of capital from other technology companies, and they have used it to create a hugely disruptive technology. Whether or not profitability materializes, that disruptive potential—and particularly the ability to eliminate human workers—is a core attraction of AI technology.
In our collective imagination of the role that automation can play in human life, replacing human workers is not the only way that automation has been imagined in the workplace. Figures 1 and 2 show visions of automated tasks easing the human condition. In Figure 1, automated machinery allows an older man to enjoy news, music, and projected entertainment as well as food and drink while reclining at home. In Figure 2, a maid’s work is eased by automated tools to clean the floor. Figure 3 tells a slightly different story. Here, the architect can direct robots on a worksite. Perhaps this is easier for the architect? But we are left to wonder—what has happened to the human workers? Still others have imagined this story in reverse, with robots directing human workers. In Figure 4, we see a robot overseeing humans doing physical labor.
For the few who have concentrated power and wealth through technology, TD is a useful narrative: it casts their choices as inevitable. But for the rest of us, TD is a trap that tells us our only role is to adapt. Yet, if technology is socially constructed, then it can be reconstructed. The question is not whether we have agency, but what we build with our technology and our agency.
Figure 1. “We’ll All Be Happy Then” by Harry Grant Dart, 1911. This cartoon sketch depicts the luxury that automated technology could bring humans. (Source: Wikimedia Commons, Public Domain.)
Figure 2. “En L’An 2000” by Jean-Marc Côté, 1899. A maid operates machinery that cleans the floor. (Source: Wikimedia Commons.)
This idea is urgent because AI is not simply responding to consumer demand; it is being integrated into our lives in ways that are difficult to avoid—a “technology push” that assumes our compliance rather than our consent. And what is being pushed is not aimed at meeting our real social and environmental challenges, but at maximizing technology uptake and revenue.
Lack of political will and inaction are one dispiriting side of the coin. Action, political will, and agency are on the hopeful side of that same coin. Individual human agency is the vehicle that harnesses our ingenuity and creativity. To us, this is a productive and useful way to conceptualize our current moment.
Human agency gives us an alternative way to consider how AI acts upon us. Emerging artistic interventions to AI systems put forth compelling arguments by highlighting the concept of technodiversity. In the article, “Machine and Ecology,” Yuk Hui explains that technodiversity is based on the principle that technological development should not be considered as a singular, homogenizing global trajectory [5]. What is key is this idea that we can deliberately cultivate multiple, coexisting digital ecosystems.
In developing this idea of diversity, Hui’s fundamental point is that the development of technology through history is situated within the place and culture of a people, and it is inherent to the human agency within a locality and cultural milieu. As he explains:
“We can say that technodiversity is fundamentally a question of locality. Locality doesn’t necessarily mean ethnocentrism, nationalism, or fascism, but rather, it is that which forces us to rethink the process of modernization and globalization and allows us to reflect on the possibility to resituate modern technologies” [5, p. 64].
This “locality” is not merely geographic; it is a locality of values. We see it operationalized today. As a 2025 Nature [6] feature made clear, the Chinese model DeepSeek “shocked the world” precisely by deviating from the Western playbook, prioritizing efficiency and access in a field dominated by scale and enclosure. This alternative approach proves that the homogenizing force of corporate AI is but one current in a sea of possibilities. Our agency lies in recognizing these divergent currents— from the artist reprogramming a robot in a workshop to the engineer building an efficient model in Beijing— and choosing to amplify them. Technodiversity is the conscious cultivation of these many currents, ensuring the future flows in more than one direction.
The 2025 study “How Artists Improvise and Provoke Robotics” [7] documents human agency in action. In Cat Royale, a robotic arm built for assembly line work is redeveloped to play with a cat. Its logic is not destroyed but diverted: toward care and a relationship no engineer anticipated. Broncomatic reimagines a mechanical bull through a new breath control interface and robotic manipulation system—a sophisticated hacking of off-the-shelf products into an interactive “thrill ride.” Different Bodies uses robotic sensors to translate nonhuman movement into sound, creating an orchestra of
Figure 3. “En L’An 2000” by Villemard, 1910. An architect oversees robots at a worksite. (Source: Wikimedia Commons.)
Figure 4. “Technocracy, a Bloodless Revolution” from The Technocrats’ Magazine, 1933. Artist unknown. (Source: Wikimedia Commons.)
Figure 5. Industrial robot inscribes the Bible onto a continuous scroll of paper. Installation by robotlab. Photograph by Mirko Tobias Schaefer, 2008, ZKM Medienmuseum, Karlsruhe. (Source: Wikimedia Commons, CC BY 2.0.)
machinic sensation. And in bios [bible] (Figure 5) by the German artist collective robotlab, an industrial robot slowly writes scripture onto scrolls of paper, its precision repurposed toward patience and ritual [8]. Each project reframes what a machine is for. Together, they demonstrate that technodiversity is not an abstract ideal but a living practice—one that asks what AI could become if it served values beyond profit.
This is human agency in its most visceral form, each project making a new relationship between autonomy, error, and collaboration. In this article, the artistic interventions show that the robot’s single purpose, as it was created, is a fiction imposed by its programmers. In interventions like these, we see the human body and imagination directly rewrite a machine’s logic. These are not niche experiments but necessary demonstrations, offering proof that our systems are the product of human choices. They chart a path forward, reminding us that the ultimate force acting upon us is not the AI models, but the humans directing this revolution—what is necessary at this pivotal moment is confronting determinism to cultivate technodiversity.
Author Information
Pamela Tudge is a research associate at the Systems Risk Laboratory, Concordia University, Montreal, QC H3G 1M8, Canada. Her work critiques communication technologies and design histories to enhance public literacy on sustainability.
Ketra Schmitt is an associate professor at the Center for Engineering and Society and an associate member at the Concordia Institute for Information Systems Engineering, Gina Cody School
of Engineering and Computer Science, Concordia University, Montreal, QC H3G 1M8, Canada. Email: ketra.schmitt@concordia.ca.
___________
To view the full version of this article, including references and acknowledgements, click HERE.
___________
JOIN SSIT