Hi, I’m Jai, a tenth-grader at Eastlake High School.

My technical journey began in fifth grade with a C++ Minecraft clone featuring custom shaders, then expanded into a 5,000-line Minecraft bot capable of complex PvP combat and strategic decision-making. By sixth grade, I was building full-stack applications on AWS using the complete serverless ecosystem: Lambda, Fargate, DynamoDB, Cognito, RDS, CDK, and associated services. I’ve developed multiple iOS and Android applications, trained custom machine learning classifiers in Python since seventh grade, and built end-to-end RAG pipelines when ChatGPT’s API first launched. I’m fluent across languages including Rust, C++, Python, Java, TypeScript, and Swift, with current focus on CUDA programming and custom PyTorch models.

My interest in Minecraft as an AI research environment emerged from a simple observation: most AI agents (i.e., MineCLIP) feel robotic because they’re optimizing for tasks humans designed, not pursuing goals they genuinely care about. Minecraft’s open-ended nature creates something unique, a complex social environment where intelligence can express itself naturally rather than being artificially constrained by researcher assumptions. After watching hundreds of humans develop authentic relationships, creative collaborations, and emergent societies in this blocky game, I became convinced it might be one of the best environments we have for studying what makes AI behavior feel genuinely autonomous versus just artificially constrained.

Mineswarm is my attempt to solve what I call “the motivation problem”: creating AI agents that develop their own objectives from internal psychology rather than executing human-designed goals. The current system uses a four-layer motivation hierarchy inspired by psychological research, where agents optimize for social reputation rather than environmental rewards. Instead of just building impressive civilizations through scale, I’m focused on creating authentic individual interactions that make you stop and think “that felt genuinely human.” The work sits at the intersection of AI development and psychology, exploring whether artificial systems can implement the functional patterns that constitute genuine motivation and emotional reasoning.