JMJoselo Martinezindesignednotmagic.hashnode.dev·Jul 20 · 3 min readLLMs and Tokens Beyond the Hype: Why Smaller Models Might Be the Better Architectural ChoiceWhen we choose and LLM, our first instinct is to pick the newest and largest model available. It's smarter, more capable, and tops every benchmark leaderboard. But is this an architectural desition? 00
JMJoselo Martinezindesignednotmagic.hashnode.dev·Jun 14 · 4 min readAgentic Systems Beyond the Hype: Why You Need Both Local and Global Perspectives to Build Agentic SystemsHaving both local and global perspectives is one of the most valuable skills a software engineer can develop. We learned this lesson in distributed systems and cloud computing, where the behavior of a00
JMJoselo Martinezindesignednotmagic.hashnode.dev·Jun 8 · 3 min readChallenges and Agent Behavior in Collaborative Swarm Environments.Bigger models tend to perform better for a specific task in a collaborative swarm environment. This could be because the model's training is optimized for efficiency and task completion. This could me00
JMJoselo Martinezindesignednotmagic.hashnode.dev·May 30 · 3 min readBuilding Pong with ABI Swarm: Why Small Models May Be Better OrchestratorsAfter several months building ABI Swarm, I finally reached a point where I could observe a complete swarm executing a real software task from beginning to end. The task itself was intentionally simple00
JMJoselo Martinezindesignednotmagic.hashnode.dev·May 25 · 6 min readWhy Bigger Models Could Break the Orchestration LayerAgentic AI Beyond the Hype When multiple models interact under semi-rigid roles, with objectives, contracts, and distributed coordination, patterns comparable to organizational dynamics begin to emerg10
