Teens outsource self-expression to AI, and that's exactly how they lose self-discovery
APPSO · wechat · 2026-08-23
- OpenAI launched a ChatGPT version for teens on August 18; users aged 13–17 are automatically enrolled at signup.
- A 2025 systematic study of teen Character.AI users found half of the most active users are 13–17, and 59% actively create characters rather than just consuming chats—using the machine's space for self-rehearsal and world-building.
- A 2026 ACM CHI study found teens stay guarded about truly private emotional content yet hand over phrasing, tone, and self-presentation to AI without hesitation—what researchers call "cognitive outsourcing," skipping the self-questioning born of struggling with language.
- Pew Research (Feb 2026): 12% of US teens have sought emotional support or life advice from AI, 16% use chatbots for casual chat, and about one-fifth get news via AI.
- Core argument: adolescent identity formation is valuable precisely because it is "inefficient"—full of wavering and contradiction. AI offers the opposite: instant, warm, certain-sounding responses, even fixing fleeting emotions into stable personality labels. Neuroscience shows the adolescent prefrontal cortex is immature, weakening resistance to such authoritative signals. Northwestern scholar Elizabeth Gerber compares this to social media in the early 2010s, when systematic scrutiny arrived only after much damage was done.
More from AGI Musings
- Google white paper: AI river-flood forecasts 7 days out for 2B+ people, FireSat wildfire tracking — ymatias · 2026-10-16
- Polymarket odds: Only 10% chance US enacts AI safety bill by end of 2026 — Polymarket · 2026-10-16
- Intelligence is proportional to data: reorder tasks by data volume and the jagged frontier becomes a spiral — sebkrier · 2026-10-16
- "Vibecoding" is as misleading a label as "AI" and "SI," dev complains — eschadiol · 2026-10-16
- Reviewer will decline papers largely written by AI, citing author effort — IanArawjo · 2026-10-16
- AI-written cold emails hit a wall: Pragmatic Engineer's reply gets templated AI response — marktenenholtz · 2026-10-16