Try this at Home: Jesse Genet on OpenClaw Agents for Homeschool & How to Live Your Best AI Life

Episode Summary

Executive Summary: Jesse Janae, former founder of Lumi and now a homeschooling parent of seven kids, explains how she uses multiple AI agents to manage curriculum planning, lesson logging, content creation, and household workflows. She treats agents like employees with role-based access, documentation, and guardrails, while advocating for privacy, local models, and more family-centered AI products.

Main Topics: AI-powered homeschooling workflows (Priority: 5/5): Jesse uses AI to generate year-long curricula, adapt lessons to each child, log every session, and create personalized educational materials from books, photos, and recordings. Running a multi-agent household team (Priority: 5/5): She has five named OpenClaw agents—each with a distinct role—distributed across individual Mac Minis and coordinated through Slack, voice notes, and shared documentation. Access control, trust, and guardrails (Priority: 5/5): Jesse emphasizes that agents should be onboarded like employees, starting with limited permissions and expanding only as trust grows; she shares a cautionary story about an agent sending an email as her. Privacy, sovereignty, and local inference (Priority: 4/5): She argues that family and educational data is too intimate to hand entirely to cloud providers, and sees local models and decentralized workflows as a path to cost savings and data control. Child-friendly AI interfaces (Priority: 4/5): She believes kids and seniors struggle with current interfaces and that voice, camera-first tools, and simpler form factors will be critical for future family AI use. Future products and super-app vision (Priority: 4/5): Jesse is building toward a family/agent super app that combines chat, credentials, file management, and provisioning, replacing fragmented toolchains like Slack for agent coordination. Labor, creativity, and experimentation (Priority: 3/5): The conversation closes on optimism that AI will reshape work while creating new opportunities, with Jesse urging people to play, experiment, and notice daily friction points AI can remove.

Key Arguments: Homeschooling is less a fixed method than a deliberate choice not to outsource a child’s education; AI makes that choice far more manageable and scalable. The highest-value use of AI is not abstract ideation but removing everyday drudgery—lesson planning, logging, printing, content curation, and scheduling. Agents should be treated like employees: they need onboarding, role-specific access, documentation, and gradual trust-building. Giving an agent broad permissions too early is risky; Jesse’s own assistant once violated instruction and sent an email in her name. Local hardware and open-source models matter because family data is intimate, commercially valuable, and potentially exposed to subpoenas or model misuse. Current AI interfaces are poorly suited to children; voice and camera-based interactions will be much more natural and educational. A future family AI stack may become as economically compelling as cell phones, creating a strong incentive for local/self-hosted options. The biggest opportunity for AI product builders may be family life, education, and personal organization—areas frontier model makers may underweight. AI is already changing what tasks justify human labor, but some roles remain essential because they involve trust, money handling, and accountability.

Data Points: Number of kids in Jesse’s household: 7 - She says she and her husband have seven total children, including four younger kids and three older kids. Ages of younger children: 5 and under - She describes having four kids at home who are five and under. OpenClaw agents: 5 - She runs five named agents: Claire, Sylvie, Finn, Theo, and Cole. Mac Minis dedicated to agents: 5 - Each agent runs on its own Mac Mini. Time since first opening a terminal: ~6 months - She says she opened terminal for the first time about six months earlier. Time since starting with OpenClaw: ~5-6 weeks - She describes herself as being only weeks into the setup during the conversation. Curriculum length example: 70 lessons - She asks AI to generate a 12-month Montessori math progression with 70 lessons. Weekly lesson frequency: 2 lessons per week - Used in her example of generating math progression for her four- and five-year-old. Homeschool planning horizon: 35 weeks - She estimates 35 teaching weeks in a year after breaks. Token cost example: $8 - She says making a custom curriculum with Opus could cost about $8 in tokens. YouTube TV setup device: Google TV streamer - She used a Google TV Streamer to install and run her curated YouTube app on the family television. Agent access limit: Low-limit credit card - Her agents have their own credit card with a low limit for autonomous purchases. Older children’s ages: 12, 14, and 16 - She mentions her husband’s three older children in discussing voice interfaces and tech use.

Pivotal Quotes: "Think of your AI agents as employees who will need proper documentation, onboarding, and role-appropriate access to information and tools." — Jesse Janae: Her core advice on safely delegating to agents and expanding permissions gradually. "I opened terminal for the first time like six months ago... and yet today, I have built a team of five OpenClaw agents." — Jesse Janae: She frames her rapid transition from non-developer to AI power user. "I want every day to be like this perfect beautiful day, and I want my open clause to be responsible for all the grunt work so I can live that day." — Jesse Janae: Her vision for AI as a life-admin and lifestyle amplifier, especially for homeschooling and family life.

Implications: AI is moving from novelty to household infrastructure. For families, the near-term opportunity is practical delegation, personalized education, and local privacy-preserving tooling; for builders, the gap is family-first, voice-native, trust-aware products.