ASU graduate aims to demystify AI through children's book
Editor's note: This story is part of a series of profiles of notable spring 2023 graduates.
How do we understand the story of artificial intelligence?
Kacy Hatfield is a student in the Herberger Institute for Design and the Arts who aims to make the story of AI accessible to all – and is doing so by writing a children’s book. Kacy is graduating this May with a degree in digital culture, and is an undergraduate researcher for the Lincoln Center for Applied Ethics.
After graduation, she will pursue a master's degree and has been invited to participate in the Machine Intelligence Group at Draper Labs.
She shared more about her college journey below.
Question: Tell us a bit about your experience at ASU and how you came to study digital culture.
Answer: I actually came to ASU as a biochemistry major; I love chemistry and math, but the career path wasn't exactly what I wanted. I then explored career and creative job opportunities where I found digital culture and in just three days I was hooked and made the switch. And there is still so much of what I love in studying AI, and I get to integrate my love for chemistry and math into that.
I actually hadn't even heard of machine learning until spring 2021, and after my professor introduced it to us, I asked her for book recommendations. From then on, I was obsessed with AI.
Q: What inspired you to pursue undergraduate research?
A: Well, I actually did my honors thesis shortly after I learned about AI and machine learning. I decided that I wanted to pursue it, even though I really didn’t know much about the subject, and I pitched it to several professors I wanted to work with, who all were very supportive. I defended my thesis almost exactly a year after I had first learned about machine learning, and I just had such an amazing time working on my thesis that I wanted to continue doing research.
I then found the Lincoln Center for Applied Ethics, which had an undergraduate research opportunity on responsible AI. I met with Research Program Manager Erica O’Neil over Zoom, and I thought it would be the perfect continuation of my work. It’s amazing to keep doing research on this, not just to learn but to ultimately come away with more questions.
Q: You're working on a really fascinating project, in which you’re developing a children’s book on AI. Can you share more on this project?
A: The premise is an illustrated children’s book that tells the story of an algorithm named Pip — like the command in Python Programming Language — and Pip has to classify seashells on the beach. How Pip classifies them starts out in very simple terms, and as waves wash up on the shore, more advanced terminology is revealed. There’s also a character named Epoch — another term in Python — as well as a character that represents the human in the loop. All of them are placed very strategically to represent what would take place if a machine learning algorithm were to be integrated in this area.
The goal is to help people feel less scared about machine learning. I often see AI described as a black box; something that that people can't see into, and can't understand. But I think the test of a good machine learning algorithm – and a good programmer — is to translate that black box into something that is easily understood.
Part of the reason I love machine learning is because even if I dedicate my entire life to studying AI, I will never have a fully comprehensive grasp of it, because it's just always expanding and advancing so fast. I think that's key in why people feel uncertainty about machine learning, especially when the Hollywood narrative of AI is the humanoid robot that is going to take over. The thing is, these technologies are amoral, not immoral.
My goal as a researcher is to start mitigating skepticism around the subject of machine learning through this book. And this starts with younger people, but the book is also meant to be used by people of all ages.
Q: How has your time in the responsible AI research group related back to your work?
A: I love being in this research group. It's actually my second semester; last semester I did a project on the risks and mitigations of AI-powered autonomous spacecraft, which is another one of my interests. It’s so awesome to be part of a group of people that have such different backgrounds and different approaches to AI. There are so many interdisciplinary perspectives and topics brought up in discussion.
I think that in terms of responsible AI – and a lot of people may disagree with me on this – it is integral for a programmer to also be able to see the ethical implications of whatever they're employing into the world. There’s often the argument that we should wait five years before evaluating those possible impacts; when I am working on programming, I'm immediately thinking about how it may affect the real world and be used.
Machine learning is like a mirror - it's going to reflect whatever we give it, and humans are not perfect. This is why I think the conversations on ethics have to go hand in hand with the research itself, and it's really interesting to see how it comes about on all different fronts.
Q: What comes next for you in your career and future?
A: That’s the age-old question, isn’t it? I always have a list of problems that I can research! This may be a nerdy confession, but I love doing research even in my free time. I hope to direct that energy into the pursuit of a master's degree and possibly even a PhD. I have also been invited to join the Machine Intelligence Group at Draper Labs in summer 2023 as an undergraduate engineer, which is a very exciting opportunity.
The amazing thing about this field is it's always changing, and in some regard, I will always feel like a student. And because the study of AI is so new, I feel like taking the ethical and programming approach at the same time would be a lot easier to integrate than something that's already established. I hope to keep these skills as best practices in the future.
There is a lot of skepticism around AI and machine learning, and often I hear people say that it’s too complicated or complex. Everybody has the capability to understand AI, and it's not as scary as it seems. Even though it's been tremendously skewed for entertainment, which makes it easier to vilify, there are so many benefits to using machine learning, and we can employ it in the right ways to augment our human experience and not hinder it.