Approximation is a skill
Roughly, the math game, gets its thesis sharpened. People fear math because they're told math is exact, and it is, but you don't always need to be exact. What you need in daily life is to approximate fast: to know, when the fruit vendor rings up 120,000, that your rough sum said 80-something, so something's wrong. Being approximately right in two seconds beats being exactly right in two minutes. The game trains that as a reflex, a GeoGuessr of arithmetic, but Juan and Julia land on the honest nuance: the reflex itself is exact (1+1 is 2, instantly, like a muscle), what's approximate is the judgment you build on top of it. And the same experiment feeds the next idea for the dead language app: a word-guessing game with hot/cold hints, exploring vocabulary by inference rather than drilling it.
// trace: where this idea came from
- ↳ video diary @ 29:13 (la gente le teme a las matemáticas porque son exactas, pero uno no siempre necesita ser exacto)
- ↳ video diary @ 29:36 (lo importante en la vida diaria es aproximar rápido, saber si no te están tumbando; el caso de las frutas 80 vs 120 mil)
- ↳ video diary @ 30:56 (el reflejo mismo es exacto, 1+1 es 2 al instante como un músculo; lo aproximado es el juicio encima)
- ↳ video diary @ 49:55 (el mismo concepto de Roughly alimenta la nueva idea para Severo, adivinar palabras con pistas caliente/frío)
- ↳ Entry 270-3: Memorize for speed (memorizar para responder rápido y aproximar rápido son la misma economía de tiempo)
Roughly, the mathematical-intuition game, gets its thesis sharpened here. People are afraid of math, Juan says, because they’re always told math is exact. And it is exact. But you don’t always need to be exact ▸ 29:13. What matters in daily life is being able to approximate, fast.
His example is the one everyone lives. You’re buying something, you run the rough sum in your head, we’re around 80,000, and then they charge you 120,000, and the whole value of the skill is that little alarm: how did that become 120? ▸ 29:36 It doesn’t matter that the real figure was 82,000 or 86,000, close enough. What matters is that 120 is far, and you knew it in a second at the counter, not in two minutes with a calculator. It happened to them with a fruit vendor who’d slipped an extra digit on; the rough number caught it. So the game is a GeoGuessr of arithmetic: answer in seconds, score by how close you land.
el reflejo es exacto; el juicio encima es aproximado →
Then Juan and Julia catch the honest nuance between them, and it’s the good part. The reflex you’re training is not itself fuzzy. 1+1 is 2, instantly, and there’s no being approximately right about that, it’s a reflex, a muscle, the way the Chinese kids compute on their fingers ▸ 30:56. What’s approximate is the judgment you build on top of exact reflexes: knowing 80-something is the neighborhood, so 120 is wrong. Precise instincts, roughly deployed. This is the same economy of time as memorizing for speed, the answer has to already be in your hands.
And the experiment doesn’t stay in one game. Playing with the GeoGuessr shape, Juan reaches a new concept for the shelved language app: you hear a word, you type a guess, and it tells you hot or cold, closer or further, feeding hints until it gives you the options ▸ 49:55. Vocabulary learned by inference instead of drilling, exploring the space of words and seeing how near one sits to another. The same principle, approximate your way toward the exact answer, quietly becomes a way to learn a language…
// continued in
no entry has continued this idea yet: the arc is still open