Student Spotlight: Samuel Sampson

Image

Samuel Sampson is from Los Angeles and currently lives in Hermosa Beach. Samuel’s academic background is a bit of a winding path, but with grit and determination, he has made it to the other side of things. “Without going into too much detail, I didn’t apply myself much in high school unless a class really caught my interest. I became fascinated with a robotics program and spent most of my free time working on that instead of studying. When I didn’t get into any colleges out of high school, I didn’t really have any plans. By chance, a friend from that robotics team had their acceptance to a public university rescinded, and that friend pressured me into attending community college with him as a result. After taking a few classes without much direction, I decided that I was most interested in computer science, in part from my experiences in my high school robotics program. To transfer out to a university, I needed to take all of the engineering, math, and physics for most computer science programs. Having not completed the math classes that most students do in high school, it took four years of studying and a transfer to a second community college to transfer out. I was fortunate enough to get accepted into a few good schools, including UCSD, where I ultimately transferred to. Instead of a Computer Science degree, I received a B.S. in Cognitive Science from UCSD. I originally intended to double major along with Computer Science there, but the program was too impacted to matriculate into Computer Science in the end. I took a minor in Computer Science as a consolation prize."

I helped found a 501(c)(3) nonprofit organization (Friends of Falkon Robotics) to support my former high school robotics program. We’ve raised over seventy thousand dollars in donations and grants... I feel I owe my career in large part to this team, so it feels good to be able to contribute to its ongoing success in some small way.

"Throughout my seven-year journey to a bachelor’s degree, I worked as a software engineering intern a handful of times. Each internship opportunity was through some connection I had made along the way—one time I met software engineers that were regulars at the coffee shop I worked at, one time a professor offered me a job at a company he founded after I did particularly well in his class! Once I got close to graduating, I applied broadly, and was fortunate to get a job offer from Google as an Engineering Resident (a provisionary role with a chance to become a full-time engineer) and have been working for Google since then as a Software Engineer. I currently work on the Google Cloud Shell and Cloud Workstations products. Since then, together with some friends, I helped found a 501(c)(3) nonprofit organization (Friends of Falkon Robotics) to support my former high school robotics program. We’ve raised over seventy thousand dollars in donations and grants to help fund this team so far and hope to ensure that the team continues to have a positive impact on students far into the future. I feel I owe my career in large part to this team, so it feels good to be able to contribute to its ongoing success in some small way.”

So why OMSCS? Samuel’s primary reason for enrolling in a master’s program was to get a chance to work with machine learning technologies. Although working in industry as a web developer is challenging and interesting in its own way, Samuel always wanted to work with machine learning in the back of his mind. He focused on ML and neural networks in his undergraduate studies, and ever since graduating, he missed that kind of workflow. One motivating factor that convinced him to finally pull the trigger and begin applying to programs was when a coworker got accepted to a Ph.D. program. His colleague was accepted to UCSD (where Samuel went for undergrad) and was able to transfer to a team that worked with machine learning within Google, working while completing their Ph.D. “To say I was jealous would be an understatement, and in part just having worked closely with someone who started their graduate studies after entering the workforce showed me that it wasn’t too late for me to go back.” Another motivating factor for Samuel was learning about OMSCS. He didn’t want to stop working to be able to pursue a degree, but before hearing of OMSCS, he didn’t know that fully online, asynchronous programs with high quality courses offering master’s degrees from universities like Georgia Tech existed. Once he made the choice to apply, OMSCS was his target program, but he made sure to apply to other programs just in case he was not accepted. Although he was accepted to a few programs, he ultimately decided that OMSCS was the right fit for him!

What Samuel likes the most about OMSCS is how integrated the OMSCS community is. “It’s hard to see from the outside looking in, but between Slack channels, Discord, Reddit, Ed Discussion forums, and Teams meetings, there’s unending opportunities to connect with other students for discussion and collaboration. I’ve even had the chance to meet up with some other OMSCS students and alums at official Georgia Tech events here in Los Angeles; it’s a really incredible community to be a part of. I also appreciate the academic rigor of the program so far. Some of the courses in the machine learning specialization I’m pursuing are notoriously difficult, but I personally would prefer a program that requires effort to one that’s easy to get through.” As for research, Samuel has been working on an augmented reality project, but he is still in the early phases of development. The goal is to custom tailor the optics and other hardware for a specific use case, and to use prototypes of the device to collect more data and refine the software and hardware iteratively. For now, he has built rough prototypes with off-the-shelf components, as well as rudimentary ML pipelines for data collection, processing, labeling, and training. Aside from this consumer-focused project, he is also interested in doing research with neural networks more generally. “I’d be interested in relating artificial neural networks and their underlying assumptions with biological neural networks and how their learning differs. Additionally, I’m generally interested in the impact of neural network structure on their ability to learn effectively in different situations.”

I’ve even had the chance to meet up with some other OMSCS students and alums at official Georgia Tech events here in Los Angeles; it’s a really incredible community to be a part of.

To Samuel, computer science is really a mathematical discipline that involves analyzing what kinds of things can be computed, developing algorithms for those computations, and analyzing properties of those algorithms. Software engineering, in contrast, is about producing software that optimizes two separate goals simultaneously: performance, and its ability to be understood. Good software needs to be performant, reliable in real-world situations, and able to be understood by humans to be maintained, analyzed, and improved. Computer science is not focused as much on the real-world constraints of producing useful software as it is on the mathematics underpinning that software. Moreover, what Samuel likes most about computers is the same thing he likes most about science: their reproducibility and deterministic nature. “The irony is that the closer you get to either the discipline of science or computing systems, the less reproducible and more stochastic they seem to behave. I suppose at a more visceral level, understanding and orchestrating some large computation always feels kind of cool. Setting up a machine learning model to train on some fancy GPU or setting up a cluster of machines for some production tasks always gives you a sense of control over the world that’s hard to achieve anywhere else. When there’s bugs in your code or tests that don’t pass, it’s a really addictive process of learning, iterating, and fixing things.”

Besides computer science, Samuel is also interested in cognitive science and philosophy. "I think having a modest understanding of both fields has helped me to make sense of other fields. I don’t know if I’d want to pursue these fields academically, but maybe an online course for fun or just reading books on these subjects would satisfy my curiosities. I’m also interested in physics; I’ve enjoyed all of the physics courses I’ve taken in undergrad and would probably be interested in taking some more courses in physics at some point.” With such incredible interests, who would Samuel collaborate with, past or present? “I think I’d enjoy collaborating with Benjamin Franklin. At one point I set out on reading all of the Harvard classics, but I got as far as one book: Benjamin Franklin’s autobiography. Reading his autobiography was truly life-changing for me. One of the concepts brought up in the book was about humans not having a systematic way of living, and how Benjamin Franklin attempted to develop one for himself. After Benjamin Franklin became financially successful, he set out on several endeavors that would benefit society, from inventions to implementing systems like libraries and fire departments in the early United States. I’m sure that he’d have a lot of interesting insights into how society is run today, and if there’s someone, I could model my life after, it would be him. Over the last several years, I think my answer to this question would have varied dramatically. I’ve lost confidence in some people I would have previously considered thought leaders and realized that some people making the most impact around me are critically overlooked. I don’t think that most of the ‘celebrity entrepreneurs’ in the tech world today are people I’d enjoy working with, to be quite honest. Some of them have had significant impact and have brought technology to the world that’s helped humanity quite a bit, but others seem to thrive on hype alone.”

The irony is that the closer you get to either the discipline of science or computing systems, the less reproducible and more stochastic they seem to behave.

What has been Samuel’s favorite class thus far? “I’m pretty early on in this program, so I have a small sample size to be honest. However, I’ve been really impressed with Dr. Charles Isbell. Dr. Isbell is the professor of my current class, CS 7641: Machine Learning, as well as the dean of the College of Computing at Georgia Tech. Dr. Isbell has been interacting with students directly throughout the course (in addition to the well-structured lectures) and even contributed to the primary research for some of the foundational algorithms (MIMIC) studied in the course. I’m also continuously impressed with Dr. David Joyner, who certainly deserves a shoutout for all the work he does to keep the OMSCS program running and for how active he is on all the various social media that kind of underpin the OMSCS community. People will ask questions about ongoing developments with OMSCS on Reddit, and Dr. Joyner will often reply with timely, relevant information.”

So, how does Samuel relax after a long day of debugging? He jokes that, “My job as a programmer involves work sufficiently different from my academic studies that after a long day of debugging there, I start an afternoon of debugging for my classes for a change of pace.” Additionally, he enjoys video games and playing music; in fact, he has a home studio where he writes and records music. Samuel even released an EP in 2021 on all major streaming platforms titled Bioluminescent. “It was a really interesting process to put out a more serious release, and learning about the various requirements of different music streaming platforms. I often find that I enjoy tweaking and tinkering with video games and setting up gaming environments more than I do actually playing games themselves. There’s been a proliferation of cheap Linux-based handheld gaming systems over the last few years focused on emulating older video games that I’ve spent a considerable amount of time setting up.”