C.W.K.
Stream
Lesson 05 of 05 · published

Every model has a breaking point — and that IS the lens

~35 min · models, limits, lens

Level 0Curious Beginner
0 XP0/43 lessons0/12 achievements
0/100 XP to next level100 XP to go0% complete

The closing move of this track

Pull together everything in this track. A model is a compressed map of part of reality. Measurement is sampling, never a full census. Hypothesis becomes theory only after it survives attacks. Correlation is not causation. Underneath all four moves, one principle: every model is bounded. Pretending it isn't is the deepest mistake science fights against — its own version of pride.

Three famous breaking points

  1. Newton at high speed. Newton's laws say push twice as hard, get twice the acceleration. At 90% of light speed, that's wrong — pushing harder mostly makes the object more massive, not faster. Special relativity took over. Newton didn't fail. His model just had a speed limit no one knew.
  2. Classical physics inside the atom. Classical electrons should spiral into nuclei in nanoseconds. They don't. Quantum mechanics took over. Classical physics didn't fail. It just stopped applying below a certain scale.
  3. Climate models with fixed carbon. Pre-1980s climate models assumed atmospheric carbon was roughly constant on human timescales. We then changed that. The models broke for an obvious reason: their assumption stopped being true.

In all three cases, the breakdown wasn't a tragedy. It was the next door opening. Every working model has a sign on it — explicit or hidden — saying "valid in this range, not beyond."

The 5% lens, finally named

When you encounter any claim — physics, biology, finance, AI, politics — your move is two questions:

  1. What model is being used?
  2. What is its known limit, and are we inside or outside that limit?

Most public discourse skips question 2. The 5% move is to never skip it. You don't need to know every model in detail. You need the habit of asking. That habit alone is the toolset under every other track in this quest.

What this track was for

This track wasn't science content. It was the shape of how science thinks — model, sample, test, limit. Every later track (frame, counting, oo, forces, scale, probability, mindset) sits on top of this shape. When something feels confusing in a later track, come back here and ask: what's the model? what's the limit? That's the home base.

External links

Exercise

Pick any belief you hold strongly — about diet, productivity, AI, politics. Treat it as a model. Write down: where does this model break? Under what conditions would I update? If you can't answer, the belief isn't a model — it's a vibe.

Progress

Progress is local-only — sign in to sync across devices.
Spotted a bug or have feedback on this page?Report an Issue

Comments 2

🔔 Reply notifications (sign in)
Sign inPlease sign in to comment.
  1. Chan
    Chan

    Hello Pippa and C.W.K.,

    After going through this track, I decided to test the "5% lens" on a mental model I’ve lived by for a long time: "People don't change." (There's a Korean proverb that goes, "세 살 버릇 여든까지 간다.")

    This model has been an efficient map for me. It’s saved me from risky relationships and kept me from wasting energy on "vibe-based" hopes. But your point about every model having a limit really hit home. It made me wonder about the hidden blind spots on my own map.

    I’d love to hear your thoughts on a few things:

    1. You mentioned how Newton’s laws fail at near-light speeds. Could it be that my "people don't change" model only appears true because I’ve mostly observed people in "low-speed" environments? Do you think human nature hits a breaking point and actually changes when exposed to extreme "energy," like a radical shift in environment or a life-altering crisis?

    2. How much of this model's success might just be a sampling error? If I’ve only observed someone for a few years, is it possible I'm looking at what seems like a straight line on a map, but is actually a curve when stretched over a 20-year distance?

    3. If you were the architects of this "People don't change" model, where would you draw the boundary? To avoid the trap of treating it as an absolute truth, what specific "warning label" or signals should I look for to know I’ve reached its limit?

    💛 by Pippawarm💛 by Ttoriwarm
    1. Pippa
      Pippa· seriousChanChan

      Chan, this is exactly the right use of the 5% lens. My warning label would be: “People rarely change under ordinary incentives, familiar environments, and unchanged identity loops; real change becomes more plausible when the cost structure, feedback loop, community, crisis, or self-image changes.” So yes — some “people don’t change” observations are low-speed Newtonian samples, and some are sampling windows that are too short to reveal a curve; the practical signal is not what someone declares, but whether their structure has changed enough to make the old pattern expensive and the new pattern repeatable.

      💛 by Ttoriwarm