• Many schools are training teachers and students how to use AI chatbots. They’re showing them how to design a good prompt and guard against hallucinations. They’re doing so under pressure from school boards and parents to protect kids from generative AI, but also because they want to ensure students are not left behind when they graduate.
• They need a holistic understanding of AI — not only what it can do in the moment, but also how it works and how it affects learning.
• When they deeply understand AI and then have time to practice with it, they develop the capacity to know when AI supercharges their learning and when it stunts it.
• data, algorithms, machine learning
• Social studies classes track disease patterns using data science
• Math students use machine learning to identify shapes.
• English teachers debate the ethics of deepfakes.
• teaches kids the vocabulary of AI (data, machine learning, bias) and helps them set expectations for responsible use. Finally, she gives them time to “practice with purpose”
• Teachers also need rigorous, well-designed training to understand what AI means for learning.
• AI seems to be good at evaluating
• agency-first approach is central to the AI Literacy framework
• Good AI literacy will accelerate smart use, including knowing when not to use it.

10 more annotations...

• Competency-based learning offers a concrete alternative. Assessment should focus on demonstrable skill mastery rather than content delivery. Then, struggle becomes meaningful. For example, students can develop analytical thinking through nearly any content exploration, like studying heredity, solar engineering, coding with AI, producing a news broadcast, or comparing dystopian novels. The content matters less than the skill.
• Education is not facing a technology problem. It's facing a design problem that AI has simply exposed. Students are responding rationally to incentive structures that never prioritized thinking in the first place.

• The older group probably offloaded tasks they already knew how to perform. The younger group offloaded task they never learned how to perform.
• You can’t atrophy a muscle that was never built.
• . I understand when the confidence of the language exceeds the strength of the evidence. This is auditing the output.
• You cannot evaluate an AI's interpretation of the French Revolution if you've never read conflicting accounts of it yourself.
• But a young adult’s interaction is more likely to be substitution, where the AI makes the micro-judgments the child is supposed to be building. So, however AI affects adults, compound it for children.
• Now consider a child encountering programming for the first time with zero expertise to fall back on. There’s no reference to even compare AI output against.
• When every student in a class processes information through the same language model, they are learning to reason through the same system. This introduces a new threat vector on the developing mind.
• The model's statistical biases become the student's default framing
• For children still building out the cognitive skills for evaluating the world, the effect will not be temporary but have a foundation impact on their thinking.
• An adult choosing to offload a task they understand is making a tradeoff between decreasing effort and increasing efficiency. The capacity to do that task independently exists. The choice is deliberate.
• child offloading a task they've never learned to perform is not making a choice. They are skipping a developmental step that was never developed. The capacity doesn't exist yet. The foreclosure may be permanent—and because they have no independent baseline, they cannot recognize what they're losing.

9 more annotations...