Stefan Lunis moved 11 times from ages 15 to 21, living in places as disparate as Tennessee and Jamaica. So when he relocated last summer to Cambridge, Massachusetts, to participate in the prestigious Summer Research Program at the Massachusetts Institute of Technology (MIT), it didn’t faze him.
He was there to “explore consciousness and the nature of reality,” Lunis says.
Studying interdisciplinary studies at UCF with minors in cognitive science and philosophy, Lunis’ research at MIT focused on the interplay of Bayesian inference with artificial intelligence (AI) and machine learning (ML) technologies. Bayesian inference describes how an observer updates their beliefs as new data becomes available. Lunis says he hopes to use the knowledge and insights he accumulates to improve AI and ML and to help shepherd these emerging technologies through social, economic and political frameworks that too often misuse world-changing innovations.
Your gift to the College of Undergraduate Studies Vice Provost and Dean’s Excellence Fund will help us propel Knights to success. Make a gift on April 13 for UCF Day of Giving.
Given the ambitious scope of his research, MIT invited Lunis to continue his work there through January 2023. Here, Lunis talks about how his background has influenced his educational path, why he thinks people and AI have gotten off on the wrong foot and how he turned a setback into a comeback.
Where are you from and how did you end up at UCF?
I was born in Baltimore and lived in Maryland until I was 13 years old, when my family and I moved to Mount Dora, Florida. My mom is an administrator for nursing facilities, so she comes in, fixes up the location and then pops over to another location. I moved around a lot, but Mount Dora became my base. I moved to Tennessee, then came back to Mount Dora, moved to Jamaica and came back, moved to New York and came back, and so on.
When I was 21, I finally settled in Mount Dora for good and started working and taking classes at Lake Sumter State College. From there, I took advantage of the DirectConnect to UCF program.
What led you to choose interdisciplinary studies (IDS) as your major?
My mom is in nursing and my dad’s a pediatrician, so I started out as a biomed major with the intention of following them into the medical field. But over time I realized that, while I could be a good doctor, I wasn’t passionate about it. I switched over to computer science and finally to an IDS major. It wasn’t a straight path, but the beautiful thing is that this circuitous route exposed me to the field of cognitive science as a whole. That knowledge has allowed me to approach artificial intelligence from a unique perspective, something that’s going to be very useful as my career evolves.
How would you describe the research you’ve been doing at MIT?
If I showed you a tomato, some lettuce and some bacon and then asked you what was missing, you’d say “bread.” Humans can fill in gaps like that because we can infer things from limited information. Large language models based on AI and ML are trained on billions of parameters, and we think there may be an implicit structure within these models that mimics that inference ability. Our research is trying to discern where that comes from.
Do you think artificial intelligence has an image problem?
AI exists in this wild space that’s really nebulous to most people right now, and that’s allowed misinformation to take hold and spread, depicting the technology as this big, bad threat to civilization. I hope that over the course of my career I can help blow away the fog and show people that there’s nothing to be afraid of. I don’t want to see AI end up like some clean energy technologies, slapped with haphazard regulations at the behest of people and groups who see it as a threat to their share price or election results.
I also want to help ensure that the rollout of AI doesn’t follow the same path as some other world-changing innovations. If there’s the potential for an emerging technology to be misused, the first people who suffer the consequences are minorities. Look at the development of highways. Where are you going to build it? Is it a field or was there a town there? We have to stop applying technology in a way that cements prejudices. Groups that I follow on Twitter, like Queer in AI and Black in AI, have opened my eyes to some of these issues, and I can take those voices into account and keep them in the forefront as my career in AI develops.
Where does that concern for others come from?
My parents are both immigrants — my mom is from Jamaica and dad is Haitian. Being an immigrant requires empathy and humility, and they passed that on to me. It also comes from the experiences I’ve had. I’ve lived in so many different locations, and I started working when I was 15. I was a certified nursing assistant, I did manual labor in a field, I tended bars. I’ve been exposed to so many different kinds of people and cultures, and I came to realize through all these interactions that localized intelligence is powerful and everyone carries it. People possess such deep and useful information from their lives. I think this imbues every person with an inherent value and should afford everyone some level of respect.
What challenges have you had to overcome?
Back in 2019, I was pursuing a degree in computer science and also working long hours at a bar in Mount Dora to support myself, and I just got overwhelmed. My coursework and GPA suffered, and I got academically disqualified. Then the pandemic hit, so I took a year to think things through and came back a lot better for it. I decided to quit my job at the bar. That’s when I made the switch to the IDS major.
How is UCF helping to prepare to graduate and enter the workforce?
UCF supported me and helped me get back on track. The college’s advisors, faculty, and staff gave me the tools and opportunities to explore my interests and the structure to help me focus on what I excelled at and what knowledge and skill I needed to build.
I can’t thank Dr. Hubert Seigneur, Associate in Solar & Energy Systems Integration, Florida’s Premier Energy Research Center at the University of Central Florida enough. He let me work in his lab, focus on research, and actually make money doing something relevant to the field I’m pursuing. With his help and the support and encouragement of my partner, Alyx, and other friends and colleagues, I’ve [gotten on the right track academically and have] developed into a person and student that that old me wouldn’t even recognize.
Finally, the entire UCF community has given me incredible support in pursuing my goals and challenging me to accomplish things I otherwise would have never thought possible.