A parent-built AI phone using a Raspberry Pi and local voice-to-text models allows children to ask questions while enabling real-time monitoring, topic redirection, and conversation control, ensuring age-appropriate, factually correct, and emotionally consistent responses while protecting children's data privacy.
Deep Dive
Voraussetzung
- Keine Daten verfügbar.
Nächste Schritte
- Keine Daten verfügbar.
Deep Dive
Parent-in-the-loop AI phone for kidsHinzugefügt:
I made an AI phone for my 5-year-old.
So, he can ask all the questions he wants and get age-appropriate, factually correct, and emotionally consistent answers.
>> The average child between two and five asks 107 questions. 107 questions per hour.
>> Oh, yeah.
>> It's one in every 33 seconds.
>> Listen, I love my children. I love teaching them and watching them all over new things they've learned. But, it sometimes takes a superpower level of patience to talk to them. Conversation with my 5-year-old is 20% entertaining poop jokes and 50% finding the nearest neighbor of what he's saying. And the rest is just me repeating myself in creative ways. Since everything is being outsourced to AI these days, why not outsource some of my frustrations, too?
So, I built a physical phone, but with a Raspberry Pi inside routing the conversation to an AI chatbot. So, now he can ask as [music] many questions as he wants, as often as he wants, in the language of his choice, whenever he wants, and I can be left alone in peace.
All right. All right. Just kidding. Kind of. The phone is real. I did make one.
But, of course, I'm not handing it to my 5-year-old and walking away.
I built this mostly to get experience with AIs and chatbots. [music] I wanted to create a very strictly controlled environment to both observe how my kids interact with AI, and more importantly, how the AI interacts with my children.
My hope is that by deliberately and carefully exploring the world of AI, I can give them practical guidance and build their awareness before they encounter them in the wild by themselves. To do so, I built a backdoor to the system so that I can monitor the conversations in the background in real time on of my devices.
I can inject comments, redirect topics, or stop the conversation altogether. I can replay old conversations under different versions of the system prompt or different language models in order to observe [music] and tune the chatbot's behavior.
I'm using local models to do voice-to-text [music] transcription and privacy screening before sending any requests to cloud-based APIs >> [music] >> so that no data of my children is leaving my house. And most importantly, I've certainly never allowed them to use it without my presence.
Seeing how on track social media has confused a whole generation of growing brains, I worry that if I'm not paying over my children as guinea pigs [music] to this new world of AI. I have over 15 years of experience working at the frontiers of robotics and AI. With all that knowledge, I feel like I owe it to my children to help them face the tsunami that's already on the way.
Ähnliche Videos
OpenHuman VS Hermes AI: Who Wins?
JulianGoldieSEO
285 views•2026-05-29
BREAKING: Microsoft’s New Image Generating Model Beat Out GPT 1.5 and Nano Banana 2
aimmediahouse
122 views•2026-06-03
Long-Running Agents — Build an Agent That Never Forgets with Google ADK
suryakunju
142 views•2026-05-30
I Made the Same Anime Fight Scene in Every AI Video Generator
NobleGooseAnime
295 views•2026-05-30
Nvidia Bets Big On AI PCs | New Chip To Power Windows Laptops | Technology | AI Updates | N18S
cnnnews18
3K views•2026-06-01
I Tested NEW Opus 4.8 on Four Projects (Updated LLM Leaderboard)
AICodingDaily
298 views•2026-05-29
3D Platformer Update - NO CAPES
SolarLune
294 views•2026-05-30
AI Doesn't Create Bias — It Inherits It
UXEvolved
176 views•2026-06-01











