Philosophy and Data Science

Studying philosophy will make you a better data scientist.

Now, I might be biased because I did study both philosophy and statistics as an undergraduate. Still, the soft skills that are vital for the modern data scientist are just what philosophy teaches. Asking the right questions, thinking critically about ethics, clear communication, and problem-solving are all core competencies of both the philosopher and the data scientist, and I’d like to talk a bit more about each of them here. My hope is that by the end of this article, you’ll be inspired to pick up a little Plato!

Philosophy strengthens and enriches your domain knowledge by teaching you to ask better questions.

Data science is a very broadly applicable skill set across many different domains, which means that throughout your career, you might play in many different peoples’ backyards. This makes data science an exciting field to work in, but it also requires you to rapidly get situated and ask the right questions about topics that might be foreign to you. Philosophy is, in large part, the study of how to ask the right questions. 

For example, I am very passionate about the environment, so in college, I took an elective course called “Environmental Ethics.” I spent a lot of time that semester reading, discussing, and rebutting or extending a diverse set of arguments (about the nature of nature, how we ought to relate to our environment, and what we should think about when discussing environmental policy). Grappling with that material helped me adapt quickly by asking better questions about my postgraduate work at the Environmental Protection Agency. Without philosophy, I would not have had as strong an understanding of the ways my work would impact the natural world. And if you need more convincing, the logic course I took freshman year also comes in handy when I’m writing SQL queries!

Philosophy teaches you ethics.

Statistics is the science of generalizations – making broad inferences across demographics and identities is a core part of the discipline. As a data scientist, you put yourself at risk of over-generalizing or discriminating beyond what is appropriate by applying statistics without asking the right questions. Philosophers have spent millennia formulating, refining, and studying those questions, and by studying ethics, you can tap into that vast reservoir of knowledge. Once you’re comfortable grappling with questions like “What does it mean to blame somebody?” or “Does it make sense to ‘deserve’ punishment?” then questions like “Should our loan eligibility model be race-blind?” become much easier to answer.

Furthermore, as a data scientist, your work will often be extensive in scope and your results or models general in nature. It’s easy enough to treat an individual kindly, but designing an algorithm that doesn’t accidentally harm anyone is much more difficult. How do you ensure that your algorithm comes up with the best response for the average user without absolutely ruining the lives of a small group of users? Ethicists have been discussing this very issue for decades, and engaging with that literature will make you a stronger, safer, and more ethical data scientist.

Philosophy reinforces valuable data science skills like communication and problem-solving.

Clear, concise, and well-contextualized writing is one of the core values of contemporary analytic philosophy. Concise writing should be one of the main goals of any good data scientist. The shorter and easier it is to read your paper, the more effectively it helps bring a reader into the context of the broader philosophical discussion. Getting lots of practice writing clear and easy-to-follow papers on topics that are often hideously thorny and receiving feedback on that writing is a great way to become a better data scientist.

Problem-solving is a vital skill for the aspiring data scientist, but it is also one of the hardest skills to master. You have to swim through an ocean of problems until you start to get a sense of what works and what doesn’t. If the problems of philosophy were easy to solve, philosophy would cease to exist. Spending a few years wrestling with problems that are so hard they’ve stuck around for centuries, if not millennia, will make you a stronger problem-solver and help build your intuition for what may or may not work. Like with math and coding, you’ll get really good at thinking about possible edge cases!

In general, while hard technical skills are certainly important for data scientists, our focus on those skills can lead us to overlook all the valuable soft skills we can gain from studying other subjects. I use problem-solving, ethics, and writing skills that I worked so hard on in undergrad every day. If philosophy isn’t your cup of tea, I urge you to think about how the other paths you may have walked before coming to data science can inform your current work. That being said, I hope I have piqued your interest and maybe even persuaded you to check out some philosophy books, videos, or syllabi. If you’d like to talk philosophy (or climate, or math, or music!), please feel free to reach out!

Columnist: Henry Williams