My Kids Do Long Division by Hand

Source: Nate’s Newsletter
Author: Nate’s Newsletter
Original source: https://natesnewsletter.substack.com/p/my-kids-do-long-division-by-hand

Private backup: the full article text is archived in the private repository at archives/articles/natesnewsletter-substack-com-my-kids-do-long-division-by-hand.source.md. It is not published on the public Quartz site.

Summary

The author argues that children need both traditional cognitive foundations and deliberate AI fluency. Long division by hand, physical books, handwriting, and independent struggle are not obsolete in an AI-rich world; they build the mental models needed to direct, evaluate, and challenge AI outputs. The article frames AI as analogous to calculators but broader: the answer is neither banning AI nor handing it over uncritically, but sequencing AI use after foundational practice. It emphasizes specification quality, metacognition, productive friction, and the ability to catch AI errors as core educational skills.

Big ideas

Claims

Key evidence and examples

  • The author contrasts a child doing long division by hand with the same child learning to vibe code with Claude.
  • The calculator analogy argues that tools can be valuable after students learn underlying mechanics.
  • Examples include AI tutoring, a wrong Claude math answer requiring human sanity checking, and vague versus precise vibe-coding prompts.
  • The article’s principles include foundation before leverage, specification as literacy, director not passenger, sequence the autonomy, catch the machine, build not browse, and attempt before augmenting.

Education relevance

Very relevant for K–12 AI literacy, parenting, curriculum design, assignment sequencing, AI tutoring, and classroom norms that balance foundational skills with responsible AI use.

My notes