Technologies Do Indeed Have Politics
Three guiding questions from a paper that everyone in our tech-centric world should read
The folks at the helm of the most influential technology companies in our world today are trying to make science fiction come to life. But their laser focus on the technical details means they often lose sight of the broader context. If we want to avoid undesirable outcomes, we’ve got to pay attention to the specific systems that these technologies are oriented toward.
Technologies don’t exist in a vacuum and they are not neutral.
Unfortunately, not enough developers of technology fully appreciate this. Today I’ll share three questions that we should all be asking of the technologies around us.
The Absent-Minded Professor is a reader-supported guide to the themes that tie together technology, education, and human flourishing. If you want to support my work, the best way is to purchase a paid subscription.
As I’ve gotten increasingly interested in thinking about the philosophical and societal implications of technology, I’ve been discovering entire fields and thinkers who are helping me along the way. One of the discoveries I made early on was that there is a whole field of scholarship devoted to understanding and wrestling with these issues: Science, Technology, and Society (STS). STS emerged in the 1960s and encompasses a few different themes including the history of technology, the philosophy of technology, and the study of how technology interacts with society more broadly through public policy and government.
What I appreciate about STS is the way that it frames the conversation about technology holistically. It’s easy to go through a program in a technical or scientific field like engineering, chemistry, physics, or computer science without paying much attention to how the technical concepts you learn fit into the broader picture of society and a vision of human flourishing. This a recipe for disaster. It’s easy for even the best of intentions to lead to harmful impacts.
Another aspect that excites me about STS is that it is an inherently interdisciplinary field. It brings together engineers, natural scientists, social scientists, historians, and humanists. And this interdisciplinarity isn’t just a happy accident. The breadth of the issues requires a broad coalition. To understand the impact of technology in the context of society we need to know both the technical details and the philosophical and historical context.
I write these essays each week to try and cultivate conversation around the impacts of technology and the interaction between technology and the good life. It’s also a big part of what I’m trying to do at Harvey Mudd. I want the engineers I’m training to explore these concepts and develop the curiosity to ask critical questions about the technologies they are using and developing.
In many ways, I feel like I’m late to the game on STS. I wish I had discovered this field earlier. But I’m doing what I can now to play catch up and educate myself about STS and how it might help shape my practices as an engineer.
As I continue to ask my colleagues in STS both at Mudd and beyond, I’m building my own STS reading list and curriculum. As I read, I want to invite you to engage with these important questions too. Today is the first of what I hope will be a series of essays over the coming months to share my thoughts and reflections on some of what I’m reading. I hope you’ll join along. If there’s interest in going deeper, I’d love to continue the conversation with you in the comments or via email.
“Do Artifacts Have Politics?” by Langdon Winner
Langdon Winner’s 1980 paper “Do Artifacts Have Politics?” quickly emerged in my queries as a canonical paper in the field of STS. After reading it, I’d go as far as to say that every engineer and natural scientist needs to read this as part of their undergraduate program. It’s a concise, accessible, and challenging analysis of the way that technology interacts with the systems and structures around it. More importantly, it lays out a set of questions to help us understand how to think about the societal implications embedded in our tools and machines.
The lessons that I took away are in the form of three questions that I think we should all be asking. Here are the themes:
The Connections Between Impact and Intention: What is the intended vs. actual impact of a technology and how are they linked?
Timing and Trajectory: Where are critical decision points for shaping the trajectory of a technology?
Embedded Organizational Structures: What political and organizational structures are embedded in a technology?
1. What is the intended vs. actual impact of a technology?
Winner opens his piece with two examples of a technological artifact with a direct social impact: Robert Moses’s bridges on the parkways in Long Island, New York and the installation of pneumatic molding machines in Cyrus McCormick’s reaper manufacturing plant in Chicago in the mid-1880s. Winner uses these two examples to highlight the way that a technology can be designed to achieve a certain social impact.
Moses’s Bridges
Robert Moses was an urban planner and public official who held a variety of roles in government from the early to mid-1900s. The reason Moses is cited in Winner’s essay is for his low-clearance parkway bridges that were built on Long Island and weren’t tall enough for buses to pass beneath.
According to evidence provided by Moses’ biographer, Robert A. Caro, the reasons reflect Moses’ social class bias and racial prejudice. Automobile-owning whites of ‘‘upper’’ and ‘‘comfortable middle’’ classes, as he called them, would be free to use the parkways for recreation and commuting. Poor people and blacks, who normally used public transit, were kept off the roads because the twelve-foot-tall buses could not handle the overpasses.
In the time since Caro’s biography of Moses was published, Moses’ supposedly racist motivation for the height of the overpasses has come into question and largely been debunked. However, even if the height of the overpasses was not directly motivated by racial prejudice, the point remains that the design of the overpasses did have the effect of preventing tall vehicles such as buses from traveling on them. This shows how technologies can have direct social impacts.
McCormick’s Union-busting Machines
The second example that Winner cites is Cyrus McCormick’s pneumatic molding machines. In the mid-1880s McCormick added these new machines to his manufacturing plant. They were extremely expensive at an estimated cost of ~$500k. Rationally, at such a cost you would assume that the technology must have offered a significant increase in efficiency or productivity for the manufacturing process. However, Winner cites the work of Robert Ozanne who suggests that McCormick’s motivation for purchasing the machines lay elsewhere.
At the time, Cyrus McCormick II was engaged in a battle with the National Union of Iron Molders. He saw the addition of the new machines as a way to ‘‘weed out the bad element among the men,’’ namely, the skilled workers who had organized the union local in Chicago. 4 The new machines, manned by unskilled laborers, actually produced inferior castings at a higher cost than the earlier process. After three years of use the machines were, in fact, abandoned, but by that time they had served their purpose—the destruction of the union.
There is a connection between intention and impact but it’s complicated
These two stories help to highlight the way that technology can be designed and intentionally used for non-technical reasons. While in these cases Winner makes a direct link between Moses and McCormick’s intentions, the link need not always be this direct.
As another example that highlights this point, Winner points to the work of coalitions of handicapped people in the US in the 1970s who showed the ways that many of the machines and structures that we use every day were not designed with handicapped people in mind. This effectively excluded them from public life. The Americans with Disabilities Act of 1990 highlights the many ways that these unintentional design decisions impacted people with disabilities by enacting policies to help resolve them.
A secondary lesson to draw from the Moses example is that connecting intention and impact is challenging. Just because something looks intentionally insidious, doesn’t always mean it is.
As we think about the technologies we develop and use, we need to consider not only their intended uses but more importantly, the impact of the technology in practice. Even technologies created with the best of intentions can lead to harmful outcomes. And while intentions are meaningful, they are only part of the equation. While the intentions of the creators of technologies, positive or negative, will surely shape our judgment of them, their real-world impacts must be assessed independently of the motive behind their creation.
2. Where are critical decision points for shaping a technology’s trajectory?
A second point that really resonated with me from Winner’s article is about the timeline of technology development and how the amount of control one can exercise over the impact of a technology diminishes as time progresses. Winner points out that for some technologies there really is a point of no return—once the cat’s out of the bag, there’s no going back. With this in mind, we need to value the early stages of a technology’s development and leverage the opportunity to shape its overall impact. Winner writes:
By far the greatest latitude of choice exists the very first time a particular instrument, system, or technique is introduced. Because choices tend to become strongly fixed in material equipment, economic investment, and social habit, the original flexibility vanishes for all practical purposes once the initial commitments are made. In that sense technological innovations are similar to legislative acts or political foundings that establish a framework for public order that will endure over many generations. For that reason the same careful attention one would give to the rules, roles, and relationships of politics must also be given to such things as the building of highways, the creation of television networks, and the tailoring of seemingly insignificant features on new machines.
I can’t help but think about this passage in connection to the recent conversation about two particular areas of technology: AI and autonomous vehicles. In both of these fields, there is a significant push to get over the initial hurdles which have the potential to limit the trajectory that the technology may take over the coming years. Of course, the folks trying to push these technologies have at least an implicit sense of the point Winner is making here. They know that the faster they can move past the tipping point of the relevant “material equipment, economic investment, and social habit,” the better for their ultimate economic outlook. Whether their motives are to profit or something more altruistic, they want to quickly push the technology past the point of no return. It’s embodied in the classic Silicon Valley mantra: move fast and break things. Of course, the assumption is that anything broken along the way isn’t that important. If only.
As we think about technologies we need to consider carefully how the choices we have continue to diminish as the technology develops. If the restriction of choice happens slowly and incrementally then maybe we are ok to take a more permissive stance. However, if we are in a situation where the “original flexibility vanishes for all practical purposes,” this ought to give us pause.
3. What political and organizational structures are embedded within a technology?
The last question that I’m asking after reading Winner’s essay is helping me critically think through how technologies may be linked to specific ways of managing them. Winner highlights a few examples that are helpful in articulating the way that technologies contain inherent biases toward particular styles and systems of management. Two notable examples that Winner discusses to make this point are that of a ship and nuclear vs. solar energy.
Each of these examples illustrates the way that a specific technology influences the structure in which it exists. For example, the design of a ship implicitly suggests the need for a hierarchical model of control with a single person, the captain, who has the final, ultimate say. For it to operate properly, the various systems of the ship (e.g., propulsion, navigation, communication, etc.) must be tightly coordinated or else it is unable to function at all. Although the specific technical details say nothing explicit about how the ship should be governed, in practice there are very few options. A ship does not lend itself to a democratic structure and all but requires a captain.
Other examples demonstrate alternate political implications. Consider nuclear vs. solar power. Nuclear power plants imply a centralized authority to control and carefully manage their operation. On the other hand, the decentralized nature of solar power enables much more democratic control of energy. With solar panels and a battery system, a homeowner can decouple their energy needs from the centralized control of the grid.
You probably already feel the implications of this tradeoff in some way. Every year I get several notices from my electric company encouraging me to install a remote killswitch on my thermostat that would enable the power company to turn off my A/C during periods of peak demand in exchange for a discount on my energy bill. This type of reasoning shows how the technological structure of the grid is oriented toward a specific way of managing it. The electric grid implies collaboration and sharing of resources.
Of course, in this example, I am not compelled to accept my electric company’s offer (at least yet). I have the choice to opt into the program. But such a program only makes sense in the broader context of an electric grid.
Artifacts and politics today
The questions and principles that Winner lays out in his essay from 1980 are still relevant over forty years later. Here are a few examples.
Autonomous vehicles
Autonomous navigation is perhaps one of the most significant areas of technological investment over the past decade. There are a number of companies in the space including Tesla, Waymo, Cruise, Zoox, and Uber. What can we learn by asking the questions prompted by Winner’s article?
Intended vs. Actual Impact
The goal of autonomous vehicles at a high level is to make transportation safer and more convenient. With a self-driving car equipped with digital eyes and ears, we can trust that we’ll be safely shuttled from point A to B in an efficient and safe manner, without needing to intervene or pay attention. Given the number of people texting and driving these days, the bar to beat keeps getting lower and lower.
But there are questions. Will self-driving cars reach the technical sophistication needed to make them safe and reliable? What about the ethical issues to navigate—in the world of autonomous vehicles the trolley problem is no longer a hypothetical question. It’s an explicitly programmed behavior. And what are we to use our newfound commute time doing? I’m sure that there is a lot we can do to redeem our commutes, but in practice will we spend our time well or fritter it away on our phones while the computer drives?
The lens of intention and impact encourages us to ask about the motivation of the companies developing these technologies. What’s the end game? What might the unintended consequences be?
Timing and Decision points
Depending on how you feel about self-driving cars, it’s important to remember that the creators of these vehicles have a particular agenda. That’s not to say that they are lying or being manipulative, but the reality of marketing is that you’re always being sold something. You’d best know what.
There are several relevant dimensions here: the first is the importance of the support of the government to give these companies space to operate. It is critical for these companies to convince the government to give them permission. So they’ve targeted the San Francisco Bay Area, Austin, and Phoenix, all tech-friendly areas, as their first testing ground.
These companies clearly understand Winner’s point about the original flexibility vanishing once the initial commitments of material equipment, economic investment, and social habit are made. This is an explicit aim of scaling quickly.
This is also why the public perception of these companies is so important. Just last week news broke that General Motors subsidiary Cruise was accused of misrepresenting itself to regulators over a recent incident where one of its autonomous vehicles ran over a woman and dragged her over 20 feet before coming to a stop. The days following an incident like this are tender moments for a company like Cruise and the entire autonomous driving world. It’s critical that they maintain the trust of both the government (who gives them permission to operate) and the broader public who ultimately will be needed for these vehicles to reach mainstream adoption.
Embedded political and organizational structures
Finally, autonomous vehicles also carry within them an implied political structure although you may not fully realize it until you buy a car without a steering wheel. Someone will be steering the car, it just won’t be you.
Ultimately the design of an autonomous vehicle will point toward different infrastructure requirements as well. As autonomous vehicles become increasingly mainstream, it’s likely that the design of roads will change to better support them, perhaps at the cost of usability for human drivers.
The artificially intelligent elephant in the room
Just yesterday OpenAI had its first OpenAI DevDay. You can watch the keynote here. They announced new custom GPTs, a new version of GPT-4 with a 128k context window, new modalities accessible via their API, and more.
It’s tough to keep pace with the new developments in the AI space and seemingly harder every day. Trying to figure out answers based on the current capabilities of these tools is a Sisyphean task: you’ll get something figured out and then a new model or tool will be released and completely change the landscape.
It’s in times of rapid technological change that fields of inquiry like Science, Technology, and Society and thinkers like Winner can offer us stable footing. While answers are often fragile, questions are robust. As we approach the technologies of autonomous vehicles, AI, or anything else the best thing you can do is find a good set of questions to guide you. The three from Langdon Winner’s essay are a good place to start.
The Book Nook
This week I finished Build the Life You Want by Arthur C. Brooks. Oprah is also listed as a co-author, but her contributions are mostly in the form of a few short notes interspersed throughout the book. The book by and large is written by Brooks.
I’d listened to several podcast conversations featuring Brooks before I picked this up so I already had a sense of the general flow of what he was going to say. Suffice it to say that it was about as self-helpy as you might expect for its title, but not in a bad way.
A few general takeaways that I appreciated:
The breakdown of the “macronutrients of happiness”: enjoyment, satisfaction and purpose
Eliminating the happiness/unhappiness dichotomy. It’s a continuous scale and we shouldn’t strive for a state of happiness (which doesn’t exist) but rather try to build practices that help us to be happier.
Listen, there’s no silver bullet here, but you weren’t expecting one, right? What Brooks brings to the table is thoughtful reminders of what you probably already know: a meaningful life is built on the foundation of family, friendship, work, and faith.
The Professor Is In
In E155 we’re out of the lab phase and moving into final projects. The students submitted their proposals last week and will present them in design reviews this week. Here’s the list of projects that they’re planning to build:
Spinning volumetric persistence-of-vision display
Split-flap clock with LED display
Snake game with VGA monitor output
Digital keyboard synthesizer
Custom phototransistor imaging array with image compression co-processor
Motion-sensing Ironman helmet
Digital maze game
Rhythm game
I’m excited to see what they are able to do and look forward to sharing more updates on the finished products next month!
Leisure Line
If you haven’t yet seen Nate Bargatze’s appearance on SNL, you’re in for a treat. Bargatze is just straight-up funny and safe for work too!
Still Life
It’s funny how drastically your childhood experiences can shape your perception of the world. Growing up in Connecticut, I remember thinking that lizards and praying mantises were rare creatures. In Connecticut, they were.
Southern California is a different story. We see what childhood me thought of as “exotic animals” on the regular. Here’s a photo of a praying mantis that kiddo #1 spotted in the driveway as we were out cleaning this last week.