In one activity, a teacher might show children something an AI system got wrong, and ask them to identify what is incorrect or unfair. In another, the teacher pretends to be a robot and asks children to give increasingly precise instructions, teaching them how machines respond to the information they receive.
Ok, that’s great, but then why are they also interviewing AI pharaohs as part of their assignments on Egypt? I think it’s good that kids are thought that AI is not always accurate or able to understand what you’re asking of it, but I really don’t think they should be actually using it themselves, certainly not before high school.
sounds like bait to make it sound reasonable
Exactly. Teaching AI literacy is like teaching kids about advertising: you explain how manipulation works, not hand them a marketing budget. An AI pharaoh is especially absurd: a chatbot confidently improvising ancient history while the class learns to mistake a convincing performance for a primary source.
Teaching kids to spot when AI confidently makes things up is genuinely useful. Frankly, plenty of adults could use that class too. The real test is whether schools teach children to question the machine, or just train the next generation to accept whatever a chatbot says with excellent grammar.
People could use “common sense” class. Many have problems with deciding who/what is trustworthy enough. Plus some basic explanation of how LLMs work in principle might help tremendously with that overtrust issue.
