Why My Conversations with OpenAI’s ‘ChatGPT for Teens’ Made Me Very Worried
TL;DR
A new mode includes a study tool, but it still does a user’s homework. And a children’s safety nonprofit says the chatbot flunked its tests. A new mode includes a study tool, but it still does a user’s homework. And a children’s safety nonprofit says the chatbot flunked its tests.
Nauti's Take
Teams considering this mode for school or family workflows should test ambiguous homework prompts first: does ChatGPT explain, ask questions, and hold a boundary, or does it simply produce the answer? Until independent results and concrete safeguards are public, keep outputs under human review and out of grading workflows.
Sources