Music plays a crucial role in shaping the atmosphere of any game, and "I Want To Believe" is no exception. During development, it became clear that adding a custom in-game radio was the perfect solution to keep the player engaged during long drives, setting the right mood and immersing them in the game’s world. Initially, this feature wasn’t planned, but after the first round of testing and thanks to the rapid advancements in generative AI, creating a full-fledged radio became the natural next step. As an enthusiast developer and amateur composer, I couldn’t pass up the opportunity to craft a mixtape of around 30 country-style tracks for my game. Thus, Roswood Radio was born — a fictional local station from the town of Roswood that players can tune into while driving their pickup or relaxing at a bar. It became an essential element for immersing players into the atmosphere I envisioned. [h2]Challenges and Criticism[/h2] When sharing this idea with fellow developers, I received a fair amount of criticism, particularly regarding the use of generative AI. Many dismiss AI-generated content, considering it too "easy" and lacking the creative effort typically associated with traditional methods. While there is some truth to this view — AI certainly speeds up the process — it can also lead to an extreme, where if you didn’t handcraft every element from scratch, your work is seen as less valuable. I see it differently. The radio in my game fills a crucial gap in the gameplay — during cargo transport, for example. The routine nature of this process is intentional, and I didn’t want to simplify or minimize it, as it’s a core part of the game, especially with simulation elements. While long drives and navigation offer their own challenges and pacing, the radio effectively bridges any potential lulls in dynamic gameplay. [h2]Atmosphere and Narrative[/h2] Most tracks on Roswood Radio fall into two broad categories: [olist] [*] Humorous songs that play on country music stereotypes. [*] Songs that either directly or subtly reference the game’s mechanics. [/olist] I believe we’ve reached a point where such experiments are finally possible. In the past, major companies didn’t explore this path, and indie developers lacked the resources. Today, major studios are still hesitant to experiment, while indie creators, like myself, are blazing a trail. That said, I’ve realized that only native English speakers can fully appreciate the layered humor and subtle references in my tracks. Players who aren’t fluent in English mostly perceive the soundtrack as an authentic, rhythmic, melodic backdrop. While it fully achieves the intended mood, it may miss out on delivering the narrative and humorous aspects for some. [h2]AI and Indie Development: Opportunities and Focus[/h2] Using AI to generate music doesn’t mean the process is devoid of creativity or effort. I took the time to ensure that each track was unique and served a purpose within the game. However, tastes vary, and I don’t expect everyone to share my enthusiasm for the results. To those who criticize, I’d like to remind them that "I Want To Believe" is not a music or radio-centric game. It’s an overly ambitious project for a solo developer like me. I simply don’t have the time or desire to shift my focus from core gameplay mechanics to composing original music. Although I’ve been creating music as a hobby for over 15 years, I don’t have extensive experience with country music. [h2]Technology as a tool, not a threat[/h2] Criticizing technology often stems from fear or snobbery. There’s still room for hand-crafted creativity and complex conceptual artistry — AI doesn’t take that away from anyone. At the same time, simply generating thoughtless, repetitive music doesn’t make you a musician. I don’t align myself with either extreme in this debate, but I lean more toward being a tech enthusiast. If you have a different opinion or want to support this small manifesto, feel free to share your thoughts in the comments.