Dr. Kathy made an observation that seems almost comical until you realize how true it is: plenty of kids probably think their parent “sits at a desk all day.” A parent may spend hours analyzing data, solving engineering problems, managing employees, designing systems, counseling people, writing proposals, or working through complicated financial decisions, while the child experiences nothing more than a closed office door and a glowing screen. We shouldn't assume children know what our work actually involves just because they know where we go every morning.

That is a much stronger parenting strategy than simply telling kids not to follow the crowd. Resistance requires an alternative. If we tell a teenager, “Don't let social media define you,” but have never helped them answer who they actually are, we have only removed one identity source without giving them another. Children need enough family history, faith, values, character, language, responsibility, and lived experience that, when culture offers them a template, they have something internal to compare it to.

Much of the conversation about artificial intelligence and children focuses on obvious dangers. We worry about inappropriate content, misinformation, manipulative conversations, privacy, or a chatbot saying something harmful to a vulnerable child. Those concerns matter, but Sherry Turkle's new book, Artificial Intimacy, pushed me toward a different question. What if the greater danger begins before the chatbot says anything obviously wrong? Turkle describes what she calls relational deskilling: as people increasingly turn to machines for companionship, they may lose, or in the case of children, fail to develop, the ability to handle the friction, reciprocity, awkwardness, unpredictability, and genuine complexity that belong to human relationships. That changes the question from “Did the AI hurt my child?” to “What relational muscles did my child never have to use because the AI made companionship easier?”