We Built a World We No Longer Understand: Why NASA Can't Replicate Its Own Engine

Try a thought experiment: picture a bicycle and sketch its mechanical diagram from memory. Chances are, your sketch will be wrong. This demonstrates what psychologists call the illusion of explanatory depth — the confident belief that we understand things we actually know only superficially.

Rebecca Lawson's Experiment

In 2006, psychologist Rebecca Lawson conducted an experiment with 175 ordinary people and 68 expert cyclists. The task: complete a bicycle diagram or select the correct version from several options. Despite their confidence, 50% of participants made errors when drawing the chain mechanism.

Our brains perceive the world through "affordances" — action possibilities — collapsing complex mechanisms into convenient black boxes. You don't need to understand gear ratios to ride a bike. You just pedal.

Modern Technology as Magic

GPS navigation requires accounting for both of Einstein's theories of relativity. An error of 38 microseconds leads to a drift of 11 kilometers per day. MP3 compression uses psychoacoustics — it strips away sounds that humans can't perceive. Arthur C. Clarke formulated it perfectly: "Any sufficiently advanced technology is indistinguishable from magic." The problem is that critical thinking shuts off when technology becomes magic.

Bone Conduction and Mentalists

Mentalists use speakers that transmit vibrations through the bones of the skull. The occlusion effect makes the voice louder when ears are covered. For an unprepared person, this looks genuinely supernatural — a perfect demonstration of how ignorance of mechanism breeds belief in magic.

The ADE-651 Scandal

In the late 2000s, Iraq purchased over $85 million worth of "miracle devices" called ADE-651 for detecting explosives. The device looked sophisticated: a black handle, a telescoping antenna, swappable cards with markings. Price tag: from $8,000 per unit.

When experts from Cambridge disassembled the device, they found: the handle was empty, the antenna was an ordinary telescoping rod, the cards contained anti-theft tags from retail stores, and there was no reader whatsoever. The operating principle was ideomotor effect — the operator's micro-muscular contractions amplified tiny movements of the antenna.

Jim McCormick had simply bought a batch of cheap novelty "golf ball finders" at $20 apiece, swapped the labels, and marked up the price a thousandfold. Trucks loaded with explosives passed through checkpoints undetected. The death toll runs into hundreds, if not thousands. McCormick was sentenced to 10 years in prison in 2013.

The Tower of Abstractions in Software

Software development is built on layers of abstraction: semiconductor physics → microarchitecture → machine code → C++ → Python → JavaScript → visual editors. The higher the level, the less developers understand about what's happening below.

Joel Spolsky's "Law of Leaky Abstractions" states that every abstraction is leaky. When your application freezes, the abstraction has leaked — resources have been exhausted at the lower levels.

New software demands more resources for the same functionality. Slack takes 10 seconds to open and requires a gigabyte of memory, even though modern processors are thousands of times more powerful than those of the Apollo era.

The F-1 Engine and Lost Knowledge

NASA planned to resume production of the powerful F-1 rocket engine from the Apollo program. The agency had complete blueprints, but the attempt to reproduce the engine hit an unexpected wall: physical specimens didn't match the schematics.

In the 1960s, engineers refined parts "on the spot" — they'd see a vibration, drill an extra hole, and everything worked. Changes were rarely documented: "Why write down the obvious?" Those people retired and took their knowledge with them.

Tacit knowledge — experience impossible to convey through instructions. It's "the master's fingertips." Reproducing the F-1 turned out to be impossible: it would require rebuilding entire industries, retraining people, and restoring supply chains.

The takeaway: technology is people, manufacturing culture, and supply chains — not just blueprints. Break the chain for one generation, and knowledge becomes an artifact.

Solutions

Cross-Pollination of Knowledge

Great Ormond Street Hospital faced high mortality rates during patient handovers. The doctors realized the problem was logistical and invited Ferrari F1 pit crew technicians. They implemented pit-stop protocols: a single leader gives commands, silence during critical moments, rigid checklists. The result: technical errors dropped by 42%, and information transfer errors fell by 49%.

Openness and Documentation

Makani (Google X) spent 13 years developing flying energy kites. The project was shut down in 2020 due to lack of competitiveness. Instead of archiving the data, the company published its source code, blueprints, flight logs, crash analyses, and a documentary film.

NASA created the NASA Lessons Learned Information System portal, where engineers describe problems and the logic behind their solutions — an attempt to preserve wisdom that was previously passed on only informally.

The Technosphere or Magic?

No single person can explain the complete manufacturing chain of a smartphone — from rare earth metal extraction to 5G protocols. Humanity has become a collection of narrow specialists, each maintaining fragments of a giant mosaic.

Generative neural networks add another layer of opacity. Even the creators of ChatGPT, Claude, or Gemini face the interpretability problem. Engineers don't know why the model produced a specific output — "knowledge is smeared across billions of weights."

"Prompt engineering" is essentially the art of incantations. If the developer doesn't understand the code (generated by AI) and the user doesn't understand the device, technology becomes a matter of faith.

We are already cyborgs — our implants are in our pockets, not embedded in our brains. The symbiosis is fragile. The risk is not a machine uprising, but a competence collapse — when we become button operators incapable of repairing the system.

Recommendations

• Open the hood. Take things apart, study source code. Curiosity is immunity against magic.
• Demand evidence. If someone responds to questions about a miracle technology with "it's too complicated" — you're being deceived.
• Preserve knowledge. Support Open Source, write documentation, share your failures.
• Learn to draw the bicycle. Train your brain to understand mechanical and logical connections.

Technology is crystallized human labor, intellect, and mistakes. Let's not forget that.