Making An Actually Fun Game (NO Coding experience)
A creator demonstrates building a fully playable video game called "Library Survivors" in a single day using AI tools like ChatGPT, Claude Code, Leonardo, and Suno AI, without prior coding experience. The process involves conceptualizing the game, using AI to flesh out mechanics, generating code, iteratively debugging and refining gameplay, creating graphics, and adding sound effects.
Summary
The video documents a complete game development workflow from conception to completion. The creator and producer Dave brainstorm a game concept called "Library Survivors," a reverse Vampire Survivors-style game where players control a librarian picking up books scattered by misbehaving kids while managing a chaos meter. They use ChatGPT-O3 Pro to flesh out the game design document with detailed mechanics, power-ups, difficulty scaling, and win/loss conditions, producing a comprehensive markdown brief. The creator then uses Claude Code (installed in Windsurf IDE) to begin coding, starting with a detailed development plan. After setting up Node.js and the development environment, Claude Code generates the initial game foundation including game loop, state management, input systems, and rendering. However, initial playtesting reveals numerous bugs: non-functional controls, too-fast chaos meter, missing NPCs, collision issues, and stamina mechanics that don't work. Through iterative troubleshooting and detailed feedback to Claude Code, the creator refines the gameplay over multiple rounds, eventually achieving a working, fun game with proper NPC behavior, book mechanics, and progression systems. The game is then backed up to GitHub as a save state before proceeding to graphical enhancements. The creator uses Leonardo AI to generate tiled floor backgrounds and characters in a retro 16-bit Stardew Valley style, producing multiple character variations (the librarian and three different kids) with walking animations. After color-correction in Photoshop to ensure consistency, the creator generates background music using Suno.ai (two different lofi tracks for menu and gameplay) and sound effects using 11 Labs for menu interactions, book placement, and stamina depletion. The final result is a fully playable, visually cohesive, and audio-enhanced game that is described as genuinely fun and addictive. The creator plans a part two to port the game to Unity or Godot for further refinement.
Key Insights
- The creator uses Claude Code's plan mode to map out development roadmaps before implementing code, allowing for course correction before actual coding begins, which prevents wasted development time on flawed approaches.
- Through approximately 10+ iterative rounds of feedback and refinement to Claude Code, the creator transforms a non-functional game with broken controls and game logic into a playable, fun game—demonstrating that AI code generation requires significant human iteration rather than one-shot accuracy.
- The creator backs up the working game version to GitHub before adding graphics and sound, treating it as a 'save state' that can be restored if new features break the game—a critical practice for iterative development.
- The creator observes that using Leonardo's tile option when generating backgrounds prevents visible seams during tiling, directly solving a problem that occurred in an earlier attempt without this feature.
- The creator successfully completes a full game development cycle—from initial concept through coding, graphics generation, color correction, music composition, and sound design—all in a single day using AI tools, with the resulting game being genuinely enjoyable and addictive.
Topics
Transcript
[0:00] All right, so let's build a game. For this specific game, we're not going to let AI just make up the idea for the game. My producer Dave and I have been actually brainstorming an idea for a game that we thought would be fun to make. The game is tentatively called Library Survivors, very similar to the Vampire Survivors game, but you're playing as a librarian. You need to put away books while rowdy kids take more books off the shelves and then leave them scattered all over the map. As the librarian, you go around and you pick up the books and you put them back on the bookshelf. As the game goes on, more and [0:32]…
Full transcript available for MurmurCast members
Sign Up to AccessMore from Matt Wolfe
AI News: Dots, GPT-6.1 Sol, Sonnet 5.5, Gemini 4, and everything you need to know
A comprehensive review of major AI announcements from the week, including OpenAI's new Dots agent, GPT-6.1 Soul model, Anthropic's Sonnet 5.5, and Google's Gemini 4 Argon, with analysis of pricing, capabilities, and competitive positioning across different models.
The Hands Down Best Coding Model Right Now
The speaker reviews Opus 5.5, claiming it's currently the best state-of-the-art coding model at a reasonable price point. They demonstrate its capabilities by running a 20-hour test where the AI built a nearly complete recreation of the Mega Bonk game, including accurate character tiers, sound design, and UI elements.
This AI Model Does’t Use Words?!
Typesafe AI released Jev, a new AI model that outputs structured decisions (choices, scores, or booleans) instead of generated text like traditional LLMs. This approach makes Jev dramatically cheaper and faster, with input costs at 4 cents per million tokens and free output tokens.
AI News: Opus 5.5, GPT-6 Sol, Jev, Muse and More!
This week in AI saw major announcements from Meta Connect (Muse agent, new AR/VR glasses), three new language models (Claude Opus 5.5, GPT-6 Soul, Grok 4.7), significant buzz around Typesafe AI's Jev decision model, and new features from YouTube, Microsoft, Google, and Spotify. Claude Opus 5.5 emerged as the new state-of-the-art model while Jev introduced a fundamentally different approach to AI outputs focused on decision-making rather than text generation.
5 Ways To Use ChatGPT's New Features
The transcript demonstrates five practical applications of ChatGPT's new features, including computer use for software automation, advanced image editing, cloud browser integration for account access, voice-powered live agents, and persistent session agents for long-term user interaction tracking.