In an interview on May 12, developer Kim Joo-seok from Nexon revealed the story behind the development of ‘ASCII Survivors,’ a game composed only of text characters. This game, developed by one person, is an experiment that combines the vampire survivor genre with ASCII code. He used advanced engines and AI to increase efficiency, and emphasized the survival strategy of indie developers that combines technology while maintaining the creator's intention.
The reason why a young man who majored in urban planning entered the gaming industry was purely with the determination to “do what I like.” He started his career as a level designer at Red Duck, briefly worked at Nexon GT and NC, and then worked at Nexon Korea. Starting with level design, I have participated in game development from various perspectives, including system design and development strategy research.
He left the company early last year. The thirst for creating games and the AI technology that I had been continuously studying coincided with the belief that “I can apply it at this level.”
The game that came out like that is ‘ASCII Survivors.’ This game, which combines the vampire survivor genre with visuals composed only of ASCII code characters, sounds like a nonsensical plan when you just listen to the explanation. Monsters, bullets, and explosions are all text characters. It started with, as he said, a “nonsense idea” that it could be shown with just a text code.
On May 12th, Inven met developer Kim Ju-seok near Jeongja Station in Seongnam. The conversation, which lasted about 70 minutes, started with career stories and expanded to include specific ways to use AI and its limitations, the complexities of the notation debate, and indie developers' survival strategies. This is not the saga of a successful indie developer. This is a calm story about leaving a large company to create and release a game on his own, and what he learned in the process.
Has majoring in 'urban planning' helped you with 'level design'?
"The name is Kim Joo-seok, currently known as 'Indigo Vortex'. I have been working as a sole proprietor under my name for about 12 years, and my first company was Red Duck, where I worked as a level designer for a while and then worked at Nexon Korea for quite some time.
Although I started with level design, I have participated in game development from various perspectives, including system design and development strategy research, and I also briefly served as a job interview committee member at Nexon Korea.
After that, I moved to Nexon Games for a while, but I wanted to do more experimentation and research, so I decided to try it outside. After saying something, it seemed like I was just saying something that seemed reasonable (laughter), but it was like that.
Your major was urban planning, and I heard that it helped you a lot in making the game. How did it help?
"Urban planning is a study that deals with space. When creating maps as a level designer, I think the sense I learned in my major came naturally when I thought about things like ‘how people perceive a certain space’ and ‘what is the flow of movement?’ As I was making the map, there was a part where I thought it could look like this in the game. What I learned through real-life space, I learned through experience at various companies in how to deal with space in game development.
I think there are many people who don't know about the job called level designer itself. Even I don't know the exact concept. Can you explain?
"In Korea, as we are exposed to a lot of RPGs, "level design is about raising the character level. There was even a joke like, “Are you designing this?”
In a game, a level (sometimes we use various expressions such as map, mission, stage, etc.) refers to all the space available to the player in the process of completing objectives. To explain a little more, the term 'level', which expressed the height of the terrain in miniature games, was also used as a concept for the depth of a dungeon, and was expanded to abstract concepts such as difficulty level, ultimately becoming a term that refers to a certain unit of game space. It seems that in the past, the expression map design was used more than the term level design in Korea.
Ultimately, in a word, it is about designing the space itself based on what activities the player does in that space and what experiences he or she has through those activities.
Since you have limited experience in level design, you could always go to another company. Why did you choose one-person development?
"Experience as much as possible in various aspects of completing the game and work toward commercial goals. I think I enjoyed game development itself as I went through many different experiences. As a result, I couldn't help but think about deep integration into the existing game development process while doing personal research on AI, and as a result, I wanted to come out and do research and challenge myself.
'ASCII Survivors' made with 'Altitude Engine' and 'ASCII Code'
"'ASCII Survivors', which was recently launched on Steam, is a Vampire Survivor-style game in which most of the visuals and game expressions are composed of ASCII code. The entire screen is just letters flying around. I'm actually confused about what letter my character is (laughs), but it's a game where you have to survive. We included a basic mode that lasts for 10 minutes, an endless mode that progresses infinitely, and a mode that requires you to survive by setting conditions, and even the visuals of weapons and enemies were all made with ASCII code.
The impetus for planning this was as I continued to play and analyze various Bamseo Like games on both mobile and PC, and at some point I thought, ‘Couldn’t this be done with ASCII code?’ Since I used to like roguelike games, ASCII visuals naturally came to mind, but when I looked for it, I found that there was no Bamseolike with ASCII visuals. So I thought, ‘I want to do it experimentally anyway, so let’s give it a try.’ I started with the ridiculous thought, ‘Wouldn’t it be possible to show it with just a text code?’ (laughs).
Is there any other reason why you chose the Godot engine for game development?
"I have been developing with Unreal Engine for most of my career. This time, I started developing in Unreal Engine. Since it is both a game and an experimental project, many open source people around me have been saying that “Godot will be the trend in the future,” and while I was somewhat familiar with Godot’s most recent game, I wanted to experiment with applying AI to development, so I chose Godot.
How did you like the Godot engine when you used it yourself? Compared to Unreal.
"Wouldn't it be the most sensitive thing to give a score (laughs)? Considering that I am someone who has mainly used Unreal, I think it is around 80 to 90 points out of 100. The advantages are clear. I felt it was light when I was studying, but once I started using it in earnest, I realized that development was much lighter and more enjoyable than I expected. I think it's a really good engine, especially for making 2D games. Because it is open source, access to plug-ins and sources is free, and the engine code itself is clear, making it easy to dig into.
The downside is that there is still a lack of references. There were many cases where I found the answers to be disappointing, so I often had to look at the code diligently. 3D has some shortcomings compared to Unreal, but it's not something you can't do. Overall, I think it is an engine with ample potential for future development, and if it is your first time developing, the altitude engine is also worth considering as an option.
Do you think you will continue to use Godot Engine in the future?
"For those who want to continue playing 2D games or want to make light games, Godotado I think it would be a good option, and my perception of altitude has improved a lot, and I have a lot of potential to use it in the future.
AI and game development — uses, limitations, and notation debate
"인공신경망이 떠오르면서 그 이후로 꾸준히 스터디를 해왔어요. 게임 내에서의 사용만이 아니라 게임 개발 과정에 대한 AI의 접목도 생각해 볼 수준이 되었던 것 같다고 느꼈죠.
이번에 가장 핵심적으로 도움이 된 건 개발 속도와 반복의 효율을 높이는 부분이었어요. 플레이를 디자인하는 과정에서 다양한 가설을 빠르게 실험하고 새로운 가설을 세우는 데도 굉장히 도움이 됐고요.
구체적인 예를 들면 사운드 작업이에요. 이 게임이 터미널 비주얼이다 보니 8비트 사운드가 어울릴 것 같았는데, AI를 통해서 제가 원하는 형태의 8비트 사운드를 시뮬레이션할 수 있는 툴을 직접 빌드했어요. 그 툴로 기반 사운드를 만들고, 제가 편집해서 원하는 감성으로 다듬었습니다. 원하는 방식의 툴을 직접 만들 수 있다는 점이 좋았어요.
AI를 활용하면서 부정적으로 느끼거나 주의해야 한다고 느낀 점이 있나요?
"결국 게임이라는 콘텐츠는 재미와 경험을 전달해야 하는데, 그걸 AI에게 다 맡겨버리면 창작자의 어떤 의향이나 디렉션 없이 그냥 생성된 결과물이 되는 거잖아요. 그래서 가장 중요한 건 "무엇을 AI에게 맡기지 않을지"를 먼저 설계하는 거라고 생각했어요. 나는 이 게임을 통해서 사람들에게 무엇을 보여주고 싶은가, 어떤 재미를 주고 싶은가, 그 결정까지 어떤 방식으로 다룰 것인가를 전체적으로 설계하는 건 결국 창작자의 몫이에요.
그 관점에서 게임의 규칙을 설계하고, 플레이를 최적화하고, 기술적 판단을 내리는 능력은 AI 시대에 오히려 더 중요해진 영역이라고 생각해요. AI가 아무리 좋다 해도 거기엔 경험과 기술적 판단이 밑받침이 돼야 하니까요. 또 리소스 생산 과정에서 AI 활용에 대해 굉장히 다양한 시각들이 있다 보니, 그 부분은 개발하면서 더 생각하게 되는 부분이기도 했습니다.
사운드 작업 관련해서, AI가 만든 툴로 만들었다는 게 표기 문제로도 이어지지 않았나요? 실제로 스토어 페이지에 기재하셨다고요.
"맞아요. 결국 툴을 만들고 그걸로 조정한 것도 저고, 편집도 제가 했는데, AI가 만든 툴을 이용했으니 AI가 생성한 거 아니냐고 받아들이는 분들이 생길 수 있거든요. 상당히 민감한 주제라 많이 고민했지만, 결국 스토어 소개 페이지에 "사운드 일부는 AI로 만든 도구를 이용해 직접 제작했습니다"라고 명시했어요. 처음부터 굳이 속이거나 아닌 척할 필요는 없다고 생각했으니까요.
'AI로 만들었다'는 걸 표기하는 것에 대해 애매한 부분이 있다고 생각해요. AI가 출력한 결과를 그대로 쓰는 건 이견의 여지가 없습니다. 그런데 그 결과를 내가 10% 정도 수정했다면, 20% 정도 수정했다면, 더 나아가 거의 다 고칠 정도로 수정했다면, 이건 AI로 만든 걸까요? 표기 문제에 대해서 좀 더 의견을 주실 수 있을까요?
"스팀이 지금 AI 사용 여부를 표시하도록 유도하고 있는데, 현재로선 각국 정부나 플랫폼의 정책 가이드라인을 따라가고 있는 상황이에요. 이 과정에서 AI를 어떻게 보는지 실제로 경험하게 된 일이 있었어요. 레딧의 어떤 채널에 홍보를 해볼 수 있을까 싶어서 관리자에게 쪽지를 보냈어요. "나는 생성형 AI를 쓴 게 아니라, AI로 툴을 만들어서 그 툴로 사운드 에셋을 만들었는데, 이것도 AI 콘텐츠로 보느냐"고요. 그랬더니 "AI가 조금이라도 관여됐다면 저희 채널에는 맞지 않습니다"라고 칼같이 선을 긋더라고요. 생성형 AI를 직접 쓴 것도 아닌데, AI에 대해 선을 확실히 긋는구나 싶었습니다.
결국 이 문제는 소비자가 어떻게 받아들이느냐예요. 퀄리티가 좋으면 AI 활용도 괜찮다는 유저도 있고, 조금이라도 관여되면 안 된다는 유저도 있어요. 칼같이 잘라 말하기가 굉장히 어려운 부분이긴 합니다.
AI가 게임 개발자의 일자리를 위협한다는 시각에 대해서는 어떻게 생각하세요? 개발자들이 체감하는 공포감은 어느 정도일까요?
"저로서는 어떤 입장을 쉽게 대변할 수 없겠지만, 공포감이 느껴지는 건 무리가 아니에요. 결국 AI는 도구인데, 그 도구로 인해 위협을 받는 셈이니까요.
소비자의 감은 날카롭다고 봐야 해서, "이거 AI가 대충 만든 거 아니야?"라고 느끼는 경우와 "AI인데 진짜 잘 만들었네?" 심지어 "이거 AI였어?"라고 느끼는 경우가 분명히 갈리거든요. AI에 의해 일자리를 잃는 것에 대한 걱정과 함께 앞으로 뭐 먹고 살지에 대한 근본적인 고민도 함께 작용되는 것 같이 느껴집니다.
저도 AI에 도움을 받아서 자료를 수집할 때 여러가지 생각을 하게 되더라고요. 개발 쪽도 비슷한 고민이 있겠죠?
"맞아요, 결이 비슷한 것 같아요. 결국 이게 옳다 그르다 쉽게 말하기 어려운 이유가, 한편에서는 생산성을 높여주는데 이건 쓸 수 있는 거 아니냐고 하고, 다른 한편에서는 생산성이 문제가 아니라 일자리를 뺏아가는 거다라고 보는 시각이 있기 때문이에요.
It is very difficult for me to easily predict whether this is positive or negative. People who work for the company have a variety of experiences depending on the company's judgment, and in fact, the job posting includes information about AI, so I don't think it's unreasonable for game developers to accept that "AI is taking our jobs."
AAA or Indie — Industry polarization and indie survival strategy
"Because there is a clear consumer need for quality and scale, large-scale games I think the axis of development will continue. At the same time, as development accessibility improves and channels to promote games become more diverse, it is a natural result that indie and small-scale development is increasing. Rather than saying that medium-sized games have disappeared, I think that the boundary between indie and indie has become blurred.
Ultimately, if it's a game with commercial purposes in mind, you can't help but worry about cost, but marketing has a structure where the more you invest, the wider the depth and breadth of the channels you can reach. No matter how well you use AI, there are limits. As a result, indies either have to sharpen and polish their game to the point where people who are obsessed with the game can form a community, or they can create a large-scale game that claims that if they come to our game, they have both this and that.
Another important thing is that forming a fandomized development community in advance seems to be one of the indies' survival strategies now. It's about creating a consensus in advance, saying, "If it's a game made by that developer, I'd probably play it too." Since opening a simple festival or indie pass is expensive, one strategy seems to be to save money while forming a fandom.
Are you saying that it ultimately depends on the consumer's choice?
"That is very likely. In the end, no matter how good a game is, it only has value if it reaches someone, and the process of making it reach someone is a cost. This is the so-called marketing area. In this area, the larger the capital invested, the wider the channel is bound to be reached. As a result, it is difficult for companies caught in the middle to recover the appropriate amount of costs, and the cost of publicizing the information is also higher than in the past. So in the end, polarization is inevitable.
However, I think it is a positive change that the number of channels through which small games can be known is gradually increasing as platforms for casual play have recently emerged. I know that there are places in Korea that are making similar attempts.
"The pure development period is about 2 months. To be exact, one It's 1.5 months, but due to various reasons, the process for releasing it on Steam is about 3 months in total.
How satisfied are you after the launch?
"To be honest, this project was closer to an experiment than sales or promotion. It was a process of demonstrating the things I thought about when I came out, such as ‘How can I apply AI to development?’ and ‘How can I increase development speed while maintaining my direction?’ From that perspective, there is quite a bit of satisfaction. Because it is an ASCII visual, it is true that the resource production speed is much faster than a typical game, and the time to worry has been reduced significantly compared to the past. In that respect, I think we have demonstrated to some extent the question, ‘How can I handle AI to increase speed while maintaining my direction?’
Looking at the reactions of Steam curators and broadcasters, many were surprised by the lack of graphics, but many commented, "You can feel the essential fun of Bamseo Like," and "It's an ASCII visual, but it has its own flair." Even if it didn't sell a lot, it was meaningful to confirm that people would view it well if it was made properly with ASCII visuals. To be honest, I don't think it would have been better if graphics had been added if the environment were to create a game like this (laughs), but being able to experiment like this on my own meant a lot to me.
What would you say to those who are watching this interview and playing the game themselves?
"If you really buy and play, I can only say thank you very much. And this While making this game, I also hoped that it would convey something to those who want to start developing as a solo artist or to those who think they can give it a try. Even though it didn't sell a lot, I still thought that it would be good to leave a feeling of 'there was a unique game' to those who like Bamseer Like.
For those who dream of becoming a solo developer, and for the future
Do you have any tips for those who dream of developing solo games?
"In many cases, you can't help but dream of hitting the jackpot while making a game, but I think you shouldn't forget 'what you want to convey' in the end. As development became more accessible, it seemed like it could be easy to miss the essence of making games. We are now in an era where believable games can be created with AI. However, if I change my thinking, I think it has become an environment and opportunity to focus on what I want to convey more deeply than in the past. If you have reached a level where you can completely create a game with ASCII visuals, it means that you can now focus more time and thought on ‘what kind of story I want to convey to people’ rather than the barriers of tools. I wish you could think of it that way. Well, it's a bit funny that we're talking about this when it's not even sold (laughs).
Do you have any plans for your next project?
"Of course, next time I think I should go for proper visuals instead of ASCII visuals. There are still a few things we are continuing to experiment with, and what I find important about this project is that the pace of experimentation has become much faster. We are discussing various things about AI with game artists, but we are making various preparations.
Any last words you would like to say?
"This project was not enough to show dramatic results, but regardless of success or failure, I was able to conduct many various experiments during the development process in a short period of time. I think it served as an opportunity to make more diverse preparations in advance when making the next game or doing something else.
I'm also personally continuing to research AI models, and I don't think AI will dramatically replace creators. Some people will say that they will use AI to express something by maximizing their creativity. Now is a period of transition itself. We live in an age where technological development is faster than ever before, and the faster the pace, the greater the confusion, but I think this situation has given me an opportunity to think more about what to look forward to in the future. I was grateful to be able to share a good story.
Comments (0)