The computer science discipline undoubtedly has enormous impacts on our day-to-day lives. Increasingly, key parts of our society are run by algorithms. This has created a range of notable scandals, for example, when machine learning systems fail to account for the diversity of the human population. This is why many individuals, inside and outside of computer science, argue for compulsory ethics modules in the computer science curriculum. This would be similar to compulsory ethics education that is commonly part of business and finance degrees. I thus argue this thinking is too simple, and fails to identify the actual root problems.
Arguing for or against more ethics should first consider what a usual computer science curriculum looks like. In my undergraduate degree in Aachen (Germany), this was a combination of theoretical (e.g. Alan Turing’s theories, P vs NP), applied (e.g. programming, algorithms, data structures), and technical (e.g. computer chips, networking technologies) computer science, as well as mathematics (e.g. analysis, linear algebra). The degrees at other institutions are similar, though might put a different emphasis on these four categories.
Coursework is usually delivered in the form of problem sheets and semester projects. While there exist some textbooks, most teaching is delivered through lectures and lecture notes. At the end of their degrees, many students additionally conduct a final project and a write-up.
All of these core modules have in common that they put much emphasis on technical problem solving, e.g. doing programming, conducting mathematical proofs, or building hardware prototypes. This is a natural focus because computer scientists will spend a great amount of their careers solving exactly those problems.
At the same time, there is hardly any emphasis on the engagement with academic literature or the production of coherent arguments in the form of text. Yet, being able to critically read challenging literature and, in turn, drafting a response are key facets of the usual university education in most other degrees.
Surely, one can discuss adding one or two ethics modules to the computer science curriculum, but this misses the point. Many students lack the pre-requisites to engage with such courses fully in the first place, namely a thorough education on academic reading and writing. It thus seems unlikely that computer science students will attain these pre-requisites through such few courses. How do we then expect computer scientists to become responsible leaders in technology when we don’t teach them the tools to engage at an equal level to other university graduates?
This is not to say that computer scientists don’t care about ethical considerations. The ones that I’ve met throughout my 8-year education and work at computer science departments are some of the most well informed and political individuals that I’ve met. Many spend a great amount of time discussing with peers, and often online and across the world.
Indeed, the focus on university education might be short-sighted. The problems seem to reach throughout our education systems. I ended up studying computer science at university because I knew this was relatively easy for me and I was good at it. This is something that I had learned at school, where I tended to perform better in STEM than some of my peers. Many of those who study computer science will probably have made similar experiences.
Furthermore, throughout the 12 years of my pre-university education, I kept being told by my teachers that I did not do well at writing. This naturally reinforced my view that I did not like writing that much. Yet, I also did not get much actionable feedback on how to improve. Many of those who did well in writing-based courses seemed to be able to so naturally, and probably many of the teachers teaching those subject had made similar experiences.
Back then, I hated writing a great deal — it just did not make any sense to me. This has changed since, through regularly writing for this blog and pursuing a heavily text-based PhD. Why? Contrary to what some people might argue, writing can actually be incredibly logical. You come up with an argument and translate it into text. Technical people love things that are logical and make sense, rather than just being ‘intuitive’. I am thus fully convinced that through a different approach to teaching writing in schools and universities, we could easily equip technical people with better skills by thinking more in terms of how many of their brains prefer to work.
The ability to read and write academically not only matters for making computer scientists critical and responsible future leaders in technology. They also matter for those that teach and research at universities. One of the most common complaints that I’ve heard among my friends doing computer science PhDs is that they struggle A LOT with the writing. This, in turn, can greatly impede their success and satisfaction during their PhDs, and produce fewer outstanding pieces of academic work. Communicating academic ideas remains as important as conducting the underlying research in the first place. Without a good, convincing write-up, academic work would not usually get published.
Again, that they struggle is only logical because they got hardly any education on this in their undergraduate degrees and got probably told throughout their lives that they might be good at STEM but not so much at other subjects.
Yes, we need more ethics, but not just in computer science. We also need to go a lot further. There needs to be much more thorough education on academic reading and writing, preferably as early as possible in computer science education. Reading Goethe and discussing virtue ethics should be as much part of this as learning about functional program and the Post correspondence problem. Otherwise, we’ll just produce a great number of coding servants, which might be good for business executives in Silicon Valley, but not probably for society. And, please don’t keep telling us that we don’t care about ethics, because we do a great deal.