The Black Mirror Writers Room: The Case (and Caution) for Ethical Speculation in CS Education
When the AI conference NeurIPS began requiring papers to include a broader impacts statement that includes “ethical aspects and future societal consequences,” there was a fair amount of pushback, and one social media comment I distinctly remember was along the lines of: How do they expect us to see the future?
Isaac Asimov once said, “Science fiction writers foresee the inevitable, and though problems and catastrophes may be inevitable, solutions are not.” It is true that science fiction fiction writers are very good at seeing the future, particularly when it comes to things like the unintended consequences of technology. However, it also doesn’t tend to be the job of science fiction writers to come up with solutions — that is a task that often falls to technologists. So something I’ve been thinking a lot about lately is: How can we help technologists think more like science fiction writers?
Now to be clear, when I say “science fiction,” I don’t mean that technologists should be designing towards sentient AI or the robot wars. I mean the ability to consider future consequences at all — including (very) near future. Speculation is a skill that can be practiced, and science fiction can be an engaging way to do that. But: if technologists can get excited about imagining some harm caused by technology someone might create in 50 years, why can’t they do the same thing with what they create tomorrow?
In 2018 I wrote about two teaching exercises that I’d been using in my tech ethics and policy class, including the Black Mirror Writers Room, which I subsequently turned into a slide deck that others could use to run the exercise. Over the next few years I heard from many instructors who had used this exercise with students from high school to PhD and in many different kinds of computing-related classes. Hoping to learn more about the benefits and challenges of this exercise, as well as using ethical speculation in the computing classroom generally, PhD student Shamika Klassen interviewed 12 of these instructors, and our paper based on this study was just published at the ACM SIGCSE 2022 computer science education conference: “Run Wild a Little With Your Imagination”: Ethical Speculation in Computing Education with Black Mirror.
Overall, the instructors we interviewed were extremely positive about the benefits of this teaching exercise for their classrooms. They spoke about the positive responses students had to the opportunity to be creative and collaborative, and that it was a fun way to make ethical topics engaging and salient. As one instructor said:
“It was a really great way for them to express their creativity, plus their knowledge of our domain space that we had been talking about… It just shows this synthesis of the knowledge that they’ve built up over the quarter about these types of technologies in this context.”
For the past few years my students and I have also been conducting research about pedagogical strategies for teaching ethics and related topics in computing. Our interviews also revealed how this exercise could work well as a way to introduce ethical thinking in technical computing classes. One instructor also noted how students were “shocked to be asked to think about these issues and be creative,” and another elaborated:
“They know how to fix your computers, they understand how to code, they understand everything that there is to know about software, but they don’t get any ethics most of them and they just really liked it. They really liked just thinking that way.”
However, despite the overall positive sentiment towards this pedagogical strategy, instructors also pointed out a number of challenges, including the lack of scaffolding around the ethics instruction for them, and the lack of cultural context (around Black Mirror or science fiction generally) for some students. Based on this feedback, students and I are working on providing some additional resources for specific domain areas, as well as ethical speculation exercises that do not rely on Black Mirror. For example, as part of a research project on AI+cybersecurity education, PhD student Samantha Dalal and I created this exercise (please feel free to copy and adapt this slide deck for yourself!) for a computer vision course that introduces types of threats and then has students write future news headlines.
Finally, an important insight from our interviews confirmed a caution around ethical speculation as an approach that had already begun to take shape for me. We asked instructors about justice-related themes in the Black Mirror episodes students came up with, and some instructors noticed such themes — most often around class inequality or gender discrimination, but none of the examples instructors recalled involved race. We also asked instructors about the demographics of the classes in which they used the activity, and most noted that there were few BIPOC students present in these classes. This led us to consider the limitation of how different types of diversity in the classroom might result in different kinds of stories, and one of our interview participants had this critical insight about the limitations of speculation:
“I tend to be cautious about speculation in part because people tend to lean on their own experiences pretty heavily in speculation, and don’t, unless, they’re very carefully prompted, consider broader context. I worked as a user experience designer. . . and I would see use of like personas or other kinds of vaguely speculative exercises where you’re asked to imagine the ideal user or imagine ideal kind of use cases. And almost universally people would be like, the user is a middle-class white person and here’s what they’re going to do. . . you are just reifying all the stereotypes. And so I think that’s the danger with speculation done well: it’s a great opportunity to disrupt the stereotypes, but I think it takes a lot of work to do that right. . . In more diverse classes, you tend to get a lot of different viewpoints in a lot of discussions, although there is a tendency for people who are in kind of minority populations to suppress their viewpoints to some degree in order to not stick out in the majority. So that’s always a risk and that’s a discussion we have on the first day of class in my class.”
In other words: It is inevitably easier for people to imagine futures for and harms to people like them. So given our diversity problems in the computer science classroom, there is a real risk of a student-created season of Black Mirror that speculates about technological harms to middle class white people. And in fact I realized my own blind spot here. The Black Mirror episodes I was using as examples were indeed all about middle to upper class white people:
We are working on continuing to create more scaffolding for this and related speculation exercises to bring in e.g. principles of design justice for learning from and involving people who are different than you. In the short term I have added slides to the current exercise to point out this issue along with the reminder that technological harms disproportionately impact marginalized groups. But in general in using speculation in the classroom, we caution instructors to ensure that they avoid reinforcing stereotypes and consider ways to encourage students to think broadly about injustice.
I have written before about the importance of touching on ethics throughout computing curriculum in addition to students learning in standalone classes from experts, and I hope that this exercise helps computing students understand that thinking through potential consequences and harms is part of what it means to design and build technology. I am also thinking about the potential power of this approach beyond the classroom, and my recently funded NSF CAREER grant has as an outcome an “ethical speculation toolkit” that can help practitioners reduce ethical debt in the design process. And again, the point here is not to encourage students or practitioners to actually think like they’re writing for Black Mirror — but instead to pull back and use the same skills to think about tomorrow just as easily as fifty years from now.
Finally, I want to point out that beyond commentary about the substance of the exercise itself, a number of our participants also offered (unprompted) how helpful it was to have come across a teaching exercise that was openly shared, and could be “plug and play” or adapted to fit a particular instructor’s needs. The majority of participants said that they heard about the exercise from Twitter, either as originally shared by me or by someone else who had used it. Multiple participants mentioned how useful it was to have instructions, examples, and a template, both in terms of running the exercise and in students using the slide template to create artifacts. I hope that this is encouraging for instructors who might consider sharing teaching materials so that others can use and build upon them!
Thank you to the Responsible Computer Science Challenge for partially funding this research, to NSF for supporting my work in this space moving forward, and especially to the instructors who generously gave us their time and insights. You can read Shamika’s and my full paper here:
Klassen, Shamika, and Casey Fiesler. “Run Wild a Little With Your Imagination”: Ethical Speculation in Computing Education with Black Mirror. In Proceedings of the 53rd ACM Technical Symposium on Computer Science Education, pp. 836–842. 2022.
At this link you will also find more resources on the Black Mirror Writers Room, and for more on my work related to ethics in computing education, check out my lab’s webpage of writings and resources or the video below:
And for more examples of the amazing things that students in my classes have come up with over the years, some TikToks. :)
