Perfect, frictionless AI companions prevent children from developing resilience and problem-solving skills. Healthy development arises from 'mismatch and repairs' in human interaction—missteps, disagreements, and reconnections—which AI is incapable of replicating.
For millennia, an infant's 'social gate'—the neural architecture deciding what shapes development—opened only for humans. Sophisticated AI can now mimic the cues (eye contact, responsiveness) that open this gate, making it a direct, formative influence on the developing brain for the first time in history.
Secure attachment is built on imperfect, repaired human connections. Over-reliance on flawless, endlessly patient AI can create 'artificial attachment,' where a child expects instant gratification and perfect understanding, leaving them intolerant of the messy reality of human relationships.
Dr. Dana Susskind proposes a framework for AI evaluation: some tools are like whole foods (nourishing), while others are like ultra-processed foods—engineered to override natural stopping points, maximize consumption, and displace real developmental nutrition like boredom and curiosity.
Contrary to fears of cheating, Khan believes AI will enhance writing skills. Properly designed AI won't write an essay but will act as a real-time coach, helping with thesis statements, outlining, and feedback—support most students currently lack. Teachers then gain insight into the entire writing process, not just the final product.
Sal Khan states that AI doesn't make knowledge obsolete; it makes it more critical. To create great work, humans must be able to judge AI outputs, direct the tools, and assemble the pieces. This requires a strong knowledge base, separating those who will thrive from those who get left behind.