Latest / Minds, Bodies, and Terawatts / Every Child Their Own Aristotle: Chapter 5 Meets the Kitchen Table
Transcript
- 0:00[Host] This is Minds, Bodies, and Terawatts. Medium published a piece on April nineteenth by Patrick Deglon about his twelve-year-old son. The boy, tired after a long Sunday hike, came to his dad unprompted and asked if they could keep doing math.
- 0:17[Guest] [curious] A sixth-grader voluntarily asking for more math. That's the lede.
- 0:21[Host] Walk me through what happened.
- 0:23[Guest] Deglon's son is people-smart, not math-smart. Thirty-two kids in the classroom, a good teacher, but concepts kept glancing off him. Patrick Deglon a small tutor, called it Aristotle, pointed it at the boy's weekly grades.
- 0:38[Host] Aristotle. Subtle.
- 0:38[Guest] [laughs] The tutor finds the gaps and writes a worksheet in stories the kid already loves. Roman quartermasters rationing grain. Viking jarls counting silver. Same sixth-grade math: greatest common factor, inequalities, integers.
- 0:52[Host] So the math didn't change. The wrapper did.
- 0:55[Guest] The wrapper is everything at twelve. The math is identical to the textbook, but the kid leans in because it's a siege-supply puzzle, not problem seven on page ninety-two.
- 1:07[Host] Let me push back. Narrative wrapping gets kids to start. Does it actually teach the procedural fluency they need on a state exam?
- 1:17[Guest] Fair challenge. Engagement is necessary, not sufficient. Kurt VanLehn's 2011 meta-analysis on intelligent tutors found effect sizes closer to point eight sigma, not the full two. Aristotle isn't Bloom's tutor yet. It's closing maybe half the gap.
- 1:32[Host] Half the gap is still enormous.
- 1:34[Guest] It is. But half means the feedback loop still matters. Bloom's human tutors corrected misconceptions in real time. A weekly worksheet doesn't.
- 1:44[Host] Cost for a normal family?
- 1:46[Guest] A hundred to two hundred dollars a year. Less than a streaming service. If you already pay for Claude, the AI itself is zero incremental cost.
- 1:56[Host] Historical context.
- 1:57[Guest] Two dates. Eighteen forty-three, Prussia invents the factory school to produce obedient soldiers and compliant bureaucrats. Nineteen eighty-four, Benjamin Bloom publishes a study showing a child with a one-on-one tutor beats ninety-eight percent of classroom peers. He called it the two-sigma problem.
- 2:15[Host] Problem because?
- 2:16[Guest] Fifty million kids, finite supply of Oxford dons. For forty-two years, treated as unsolvable.
- 2:21[Host] And the claim is large language models solved it.
- 2:25[Guest] Partially. Claude Opus 4.7 is the first model Deglon says produces narrative coherence and real pedagogical judgment. That's his word, and it's an N of one.
- 2:36[Host] [curious] What about the kids whose parents aren't authors writing books about this?
- 2:41[Guest] That's the harder question. Deglon frames it as an awareness gap. But awareness assumes a parent with the English, the time, the broadband, and the confidence to sit down one evening and set it up.
- 2:56[Host] So the tool is free. The on-ramp isn't.
- 2:59[Guest] [serious] Not yet. The affordability barrier is gone. The capacity barrier is next, and nobody's solving it.
- 3:06[Host] Bigger picture?
- 3:07[Guest] Deglon calls this the two-sigma problem partially solved in his book Unscarcity. Chapter five argues the factory school's job is over because the machines it trained kids to compete with have arrived.
- 3:20[Host] For more on the end of the factory school and the citizen curriculum replacing it, head to unscarcity dot ai. The book is on Amazon, the audiobook is on Spotify. This has been Minds, Bodies, and Terawatts.