The discussion showcases a paradigm shift from viewing AI as a simple tool to conceptualizing it as a team of specialized employees. Jene has assigned distinct roles to her agents—chief of staff, developer, curriculum planner—and manages them as a collaborative unit to run her household and educational initiatives.
Despite having no prior coding experience, Jene has successfully deployed a sophisticated multi-agent system. She leverages her product management skills to define goals, document processes, and creatively solve problems, demonstrating that deep technical expertise is no longer a prerequisite for advanced AI implementation.
Jene uses her AI agents to create hyper-personalized educational experiences, from generating a full Montessori math curriculum to analyzing video lessons to identify a child's specific learning gaps. The agents also inventory physical toys to integrate them into lesson plans and power a custom, parent-approved YouTube app.
Jene's work reveals the inadequacy of existing communication platforms like Slack for managing teams of AI agents, citing issues with identity confusion and context management. Her solution is to build a new "super app" that natively integrates chat with credential, file, and API key management for seamless agent provisioning.
A strong emphasis is placed on the importance of running AI models locally to ensure privacy, reduce costs, and avoid dependence on a few large technology companies. Jene predicts that local models will achieve performance parity with today's top-tier cloud models within 18 months.
Keep pulling the thread on Jesse Genet.