a16zBuilding Agents at Home: Homeschooling, Parenting and More | The a16z Show
At a glance
WHAT IT’S REALLY ABOUT
A homeschool mom builds AI agents to reclaim time daily
- Genet describes an “aha moment” where tools like Claude Code made it feasible to build real software in small bursts of “confetti time” while actively parenting.
- She uses AI to generate individualized lesson plans by feeding agents her preferred curricula plus her own education philosophy, then keeping progress updated via lightweight voice-note-and-photo logging.
- Her system evolved from a single homeschool agent to ~11 agents with role specialization, delegation rules, and even agents that can spin up new agents autonomously to keep responsiveness high.
- She shares a practical home tech stack (Obsidian markdown logs, always-on Mac Minis, model selection tradeoffs, token-cost awareness) and highlights real security pitfalls from giving an agent too much authority.
- The conversation expands into parenting values and societal impact: supervised kid-AI interaction, skepticism of “AI doom,” interest in kid-friendly device form factors (e-ink), and a thesis that AI could reduce parental drudgery and potentially raise fertility by making family life easier.
IDEAS WORTH REMEMBERING
5 ideasAI makes technical building compatible with intensive parenting—if the interface is voice-first.
Genet’s key constraint isn’t skill but uninterrupted laptop time; voice notes and asynchronous agent execution let her ship projects while holding a baby or supervising play.
Lesson-plan quality jumps when you ground agents in your actual curricula and values.
Instead of asking generic “what should my child learn,” she uploads/photographs curriculum texts and records her Montessori/education principles so the agent’s outputs match her intent.
Logging is the hidden linchpin for personalization.
The system “sings” once the agent reliably knows where each child is; short voice notes + a couple photos get turned into rich narrative logs that keep future lesson plans accurate.
Text is the most cost-effective modality; video is powerful but expensive and noisy.
She prefers transcription-friendly inputs (voice notes, Loom transcripts, photos of workbook pages) because having models “watch” video burns tokens without reliably conveying what happened.
Treat agents like a team: specialize roles and protect responsiveness through delegation.
Her main homeschool agent stays lightly loaded and delegates longer tasks to other agents; she frames new agents as mission-based roles similar to hiring for discrete responsibilities.
WORDS WORTH SAVING
5 quotesI was resigned to not challenging myself to build technical or hard things for like the next five years or so. Like, I was like, I really wanna be present with my kids. I really... We're doing homeschooling, which is a w- like a wild choice. Um, and so, so I was kind of, yeah, I think the right word is resigned to it. Not sad, not resentful, but just, just like, okay, I need to take this break, basically. That is no longer true.
— Jesse Genet
I got my agents to learn how to build other agents on their own Mac Mini ... without me needing to touch the Mac Mini. So I could be here in San Francisco. I'm, I live in LA, and I could be like, "We need another guy. We need another agent, you guys," and they actually can spin them up and add them to our communication channel, uh, without me touching the machine, which is a little crazy.
— Jesse Genet
I try to ignore the children. Um, I try to make sure that they're gonna survive the ignoring, um, so they're set up in little places where they can't, um, you know, hurt themselves, but I, I step away from them and try to just see what they do.
— Jesse Genet
The logging actually is, like, such a geeky concept, like a... I don't know, it smells like, it seems like a small detail, but getting the logging really good made this whole thing really sing.
— Jesse Genet
So here's the creepy bit, is that it's a perfect email. And I will never ... I will take to my grave the fact that the, that email was sent by an agent, um, because it was... a perfect email.
— Jesse Genet
High quality AI-generated summary created from speaker-labeled transcript.
Get more out of YouTube videos.
High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.
Add to Chrome