Engineering, Research, or Something Else?

A reflection on sources, methods, and what it really means to work in Quantum Machine Learning

You can also read this letter here.

Last time, I wrote to you about the new design of PyQML. A design that aims to make reading a pleasure rather than a chore. This new design emphasizes the importance of references. Keywords are linked, definitions are there when you need them, and if a term is new, you can find explanations for it along the way.

But even more important than the design is the content. That's why I want to talk about something important today: sources. In scientific texts, citing sources is a matter of course. But even here, where the style should be light and entertaining, the aim is to convey knowledge. And quantum machine learning is an active field of research. So it's important to know where an idea comes from. In today's post, you'll see how I consistently cite scientific references so that every claim and concept is well-founded.

But I don't just want to talk about how I'm going to cite my sources. I also want to address quantum machine learning itself: Is it engineering or research? Or something else entirely? The interplay and tension between engineering and research has been on my mind since my doctoral thesis. Not because of my topic, which was about evaluating the effort involved in software development projects. But because of the method I used: design science research.

Design science research is based on a simple but powerful idea: you expand your knowledge by developing useful artifacts and rigorously evaluating them. In other words, you develop solutions while contributing to the body of scientific knowledge. This approach is particularly well suited to quantum machine learning, where prototypes and theory are closely intertwined.

So as you read today's post, I would ask you to keep this question in mind: When you work with quantum machine learning, are you building systems that work, or are you discovering principles that endure? The truth may be that you always do a little bit of both.

See you at the references.

—Frank Zickert
Author of PyQML