Too young to scroll?

Episode Summary

Executive Summary: This BBC Tech Life episode examines three tech-and-society themes: a Greystones, Ireland community delaying smartphones for under-13s to reduce anxiety and peer pressure, a London founder using an AI-powered wearable badge to improve women’s personal safety, and a Stockholm AI-run cafe revealing both the promise and limits of autonomous agents in real-world business.

Main Topics: Greystones smartphone-delay community (Priority: 5/5): A group of eight primary schools in Greystones, Ireland, asked parents to wait until secondary school to give children smartphones, aiming to reduce anxiety, tiredness, and social pressure. Children’s well-being and screen-time concerns (Priority: 5/5): Teachers and parents describe increased anxiety and fatigue among primary-school children, linking late-night screen use and early-morning device habits to poorer sleep and school readiness. AI-powered personal safety wearable (Priority: 4/5): Enough CEO Ina Jovicic explains how her own violent mugging led her to build the Eno badge, an AI-enabled wearable that detects danger cues and sends live evidence to responders. Privacy, false alarms, and human oversight in safety AI (Priority: 4/5): The safety badge is designed to keep footage encrypted and inaccessible unless needed, while human operators verify AI alerts to reduce errors in high-stakes situations. AI agents in the real economy (Priority: 5/5): Andon Labs tests an AI agent named Mona managing a Stockholm cafe, showing autonomy in hiring, ordering, pricing, and customer interaction, but also revealing major operational weaknesses. Limits of current AI autonomy (Priority: 5/5): The cafe experiment demonstrates that present-day models can mismanage inventory, be socially manipulable, and fail at basic business judgment, despite strong language skills.

Key Arguments: A community-level agreement can reduce smartphone pressure by making delayed ownership the norm rather than the exception. Parents and teachers observed that children were more tired and anxious when screen access crept earlier into the day and later into the night. The Greystones initiative worked because it was collective: parents, principals, and schools all aligned, removing the feeling that one family was acting alone. Digital safety education still matters, so the schools also prepare children for smartphone use later through workshops and ambassador programs. A personal safety product should address the absence of human support during emergencies by using AI to detect danger and escalate quickly. Privacy concerns can be mitigated by encrypting recordings and only opening access windows when an offense occurs. High-stakes AI systems should keep a human-in-the-loop layer because safety-critical decisions cannot rely on AI alone. Current AI agents can carry out administrative tasks like hiring and ordering, but they lack practical judgment and can be fooled easily. The cafe experiment is designed less to make profit and more to identify failure modes in agentic AI before broader deployment. AI progress is fast, but real-world autonomy still requires testing under messy, everyday conditions rather than just in labs.

Data Points: Age threshold in Greystones pact: 12 or 13 - Parents agreed to wait until children reach secondary school before giving them smartphones. Schools involved: 8 primary schools - The It Takes a Village initiative spans eight local primary schools in Greystones. Initiative age: about 3 years - Rachel Harper said the project began roughly three years ago. Teacher experience: 26 years - Alicia Giesler said she has taught for 26 years and noticed changes over time. Community meeting turnout: could have filled the theater three times over - Harper described strong attendance at the local town meeting launching the initiative. Australian social-platform restriction: under 16s - The episode notes Australia has introduced a law restricting social platforms for those under 16. Time since attack: more than 4 years - Ina Jovicic said her mugging occurred almost more than four years ago. Attack details: 4 men - Jovicic said she was attacked and surrounded by four men in central London. Attack time: 10 p.m. - She said the assault happened at 10 p.m. in a very safe area. AI cafe model tested: Gemini - Axel Backlund said the first tested model did not perform well. Canned tomatoes ordered: 22 kilograms - The cafe agent over-ordered canned tomatoes without a menu reason. Discount scam example: 99% discount - Backlund used this as an example of the agent being easily fooled. New model in cafe: OpenAI - The team switched from Gemini to OpenAI’s model.

Pivotal Quotes: "I was like, What can I do to change this so this doesn't happen to anyone else?" — Ina Jovicic: She explains the motivation for creating the AI safety badge after being attacked. "It doesn't, you know, matter where parents are around the world, that they had huge concerns. And it somehow united us all together." — Rachel Harper: Harper reflects on how the smartphone-delay initiative resonated beyond Greystones. "It bought 22 kilograms of canned tomatoes. For no particular reason, that wasn't on the menu." — Axel Backlund: Backlund describes one of the AI cafe agent’s most obvious inventory mistakes.

Implications: The episode suggests communities and companies are testing practical guardrails for tech: delaying smartphones, adding human oversight to AI safety tools, and stress-testing autonomous agents before trusting them with real-world decisions.