How I built LennyRPG

2026-03-17 · Source: Lenny's Newsletter · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering, Gaming & Interactive Media · Depth: Intermediate, long

Summary

Lenny's newsletter has released its entire archive and podcast transcripts as AI-friendly Markdown files at LennysData.com, alongside an MCP server and GitHub repo, challenging readers to build innovative AI projects. This initiative was spurred by the success of community-created tools from previously shared podcast transcripts, with a standout example being "LennyRPG" by Ben Shih, a non-technical product designer. Shih developed "LennyRPG", a "Pokémon-style RPG" where players answer product questions from podcast guests, utilizing a suite of AI tools including ChatGPT, Claude Code, Cursor, GPT-4o, and the Phaser 3 game engine. His development process involved AI-assisted ideation, PRD generation, systematic processing of 300+ transcripts for quiz questions, automated generation of over 250 pixel-art avatars, and integration of game mechanics and a Supabase-powered leaderboard. This project demonstrates the significant potential of AI tools to empower non-technical users in creating complex, engaging applications, underscoring the value of well-defined project requirements for effective AI collaboration.

Key takeaway

A non-technical product designer rapidly developed LennyRPG, a full-featured AI-powered game, by systematically leveraging LLMs (Claude Code, ChatGPT, GPT-4o) across the entire product lifecycle. This involved AI-generated PRDs, automated processing of 300+ transcripts for quiz questions, and consistent generation of 250+ pixel avatars, significantly accelerating development. This workflow offers AI/ML professionals a blueprint for rapid prototyping, scalable content generation, and empowering non-technical users to build complex applications from structured data.

Topics

• AI-Assisted Development
• Game Development
• Large Language Models
• AI Data Preparation
• MLOps

Code references

• RSamaium/RPG-JS
• ComposioHQ/awesome-claude-skills

Best for: Machine Learning Engineer, Software Engineer, AI Product Manager

Related on AIssential

• I Asked ChatGPT, Claude and DeepSeek to Build Tetris — Model performance for coding tasks varies significantly across LLMs, impacting development cost and user experience.
• HybridRAG: A Practical LLM-based ChatBot Framework based on Pre-Generated Q&A over Raw Unstructured Documents — HybridRAG pre-generates QA pairs from unstructured documents to enhance RAG chatbot speed and accuracy.
• #343 Vibe Coding and the Rise of the Non-Developer Builder with Matt Palmer, Developer Relations at Replit — AI-driven "vibe coding" empowers non-developers to rapidly build and deploy custom applications, accelerating product delivery.
• Nemobot Games: Crafting Strategic AI Gaming Agents for Interactive Learning with Large Language Models — Nemobot extends Shannon's game taxonomy using LLMs to create interactive, self-programming AI game agents.
• Building Vestimate (Part 2):The Synthesis Engine and Dual-Stage RAG — Vestimate integrates vector search and constrained LLM generation within a Dual-Stage RAG architecture to deliver context-aware, hallucination-free clothing recommendations.
• I had ChatGPT build me a free PDF editor because I didn't trust it to change my files - it worked! — Non-deterministic AI can effectively generate deterministic programs for precise, repeatable tasks.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Lenny's Newsletter.