Meet the Video Game Designer Who’s Leveling Up with AI

SeungGeon Kim, a technical game designer in South Korea, grew up playing video games on Nintendo DS and Wii. He used to dream about working for the gaming giant at some point in his career, but in recent years, he’s become more focused on his own ideas. “Right now, I’m developing games for myself,” he said.

When Kim learned about ML-Agent, an add-on to the game engine Unity that helps train agents through reinforcement learning, a bigger picture began to take shape. “It’s a reinforcement learning library that enables [developers] to embed smart agents right into games,” he explained. “They feel and move like real players, and I thought if I knew about data science, like machine learning and statistics, I just might be able to work with those new libraries more.” 

Building on computer science with data science

While Kim had earned his bachelor’s degree in computer science, he grew more interested in data science and the future of that field. He planned to earn his master’s degree in person, but like so many learners, he had to shift his plans once the COVID-19 pandemic halted international travel. “I was planning on going to Germany or the US to get my master’s degree, but the virus came out and the visas…you know.” 

With that possibility seemingly grounded, Kim began exploring other degree options. “I was trying to decide whether I should stay in Korea or try to find some college that offers good quality online education,” he said. Then he came across the University of Colorado Boulder’s Master of Science in Data Science. 

The curriculum caught his attention, as did the fact that he could enroll in the graduate degree program based on his performance in three pathway courses. CU Boulder’s master’s in data science program offers performance-based admission, so if students show their aptitude in one of two pathways—computer science or statistics—by passing three required courses, they’re admitted to the full degree program.

Earning a master's degree online

The online master’s degree entails taking required courses, like machine learning and statistics, before choosing from a series of electives, such as deep learning or software architecture. “Taking all those courses helps me a lot instead of taking some tutorials or short certificates,” Kim said. 

His favorite turned out to be statistics. “The statistical model courses were the best, I think,” he said. “The professor was really good at teaching subjects and giving examples and the quality of assignments was really good.” 

Learn more: Deep Learning vs. Machine Learning: Beginner's Guide

Kim had experience with online learning thanks to his bachelor’s degree. He knew that studying certain subjects virtually made more sense. “It saves time and there’s really lots of overhead when you need to go to campus,” he said. “You need to get a house when you go to campus, you have to pay for food. If you do it online, you can do it from your home. I was able to complete the degree while working full-time as a game designer.”  

He was also able to move through his coursework at a faster pace because he controlled how much he worked on at a given time. “If it wasn’t for this program, I wouldn’t have been able to get a degree this fast,” Kim said. 

Now Kim is nearing the end of his time in the program, and he’s excited to apply what he’s learned to the games he is developing. “Thanks to the degree, I was able to read all the documents, follow tutorials, and implement new kinds of AIs in my game,” he said. “That was the basic purpose, and now I’m very satisfied because I was able to fulfill my purpose.”