The child who had learned to read signs now had to understand how things held together.

The family home functioned as an informal laboratory. There was no program, no official curriculum, no clearly divided sequence of subjects. Books, maps, tools, vehicles, scientific conversations, legal reasoning, and technical objects coexisted. Knowledge was not organized into labeled disciplines. The world appeared as a set of nested systems.

Marcel Ehrmann explained electricity to his son with an image simple enough for a child to grasp and rigorous enough to remain useful: the mountain and the waterfall. Voltage could be compared to height. Current, to the quantity of water flowing. Resistance, to whatever slowed or constricted the passage.

A very high waterfall with only a trickle of water did not produce the same effect as a powerful mass of water falling from a lesser height. Height alone was not enough. The volume of water also mattered.

The invisible force was therefore not mysterious simply because it could not be seen. It obeyed rules that could be represented, measured, and tested.

In his father’s office stood a monumental globe, almost one meter in diameter in Thierry’s memory.

The object had a physical presence. You could turn it, bring a continent close, then step back and recover the whole planet.

This movement from detail to the whole became a way of thinking. A valley could be understood only in relation to a river, a river in relation to a mountain range, the mountain range in relation to a continent. Scale changed meaning.

There were also Michelin maps. Marcel had worked in the world surrounding André Citroën and knew the relationships between Citroën and Michelin.

Local maps, maps of France, Africa, and the world made relief, roads, densities, and distances readable.

Thierry could spend hours on them. He was not merely asking where a city was located, but why it was there.

Why did a road follow a valley rather than cross the mountain?

Why did a border follow a ridge, a river, or an artificial line?

Why did circulation concentrate in one place and leave another almost empty?

Why did people settle along certain shores?

Where did a river begin?

What power did it acquire before reaching the sea?

A map revealed relationships. A mountain was not simply a topographic feature. It could be protection, obstacle, water reserve, climatic barrier, logistical difficulty, or border.

A river was at once a means of transportation, hydraulic force, agricultural resource, military axis, and line of settlement.

The logic of the engine reappeared in territory: sources, flows, resistance, and transformations. Mechanics had been a first geography. Geography became an enlarged mechanics.

Books extended this world. After Jules Verne came works intended for adults.

Thierry read them even when he understood only part of them. An unknown word did not stop him. He approached it like a mechanism that had not yet been taken apart.

Among these books was, in his memory, a manual devoted to the tank, probably a military instruction manual.

A tank was a system of dependencies.

It replaced part of the cavalry, transformed the speed and geometry of war, and shifted balances of power without ever eliminating vulnerability. Every innovation generated its counter-technology.

Much later, drones would strike armored vehicles by targeting weak points that designers of earlier generations could not have imagined in the same way. No power is absolute.

Marcel was also a pilot. Aeronautical books entered the house. Lift, navigation, instruments, weather, and instrument flight rules became familiar subjects. Thierry asked why an airplane did not fall.

Instrument flight left him with a lasting image. The natural gaze can mislead. Clouds erase the horizon. The body can give false sensations. The pilot must then rely on a coherent set of indicators.

Trusting instruments does not mean abandoning judgment. It means distributing judgment among several sources of data.

Decades later, this principle would reappear in his work on weak signals, databases, redundancy, and control.

Rivers and dams introduced him to another form of power: hydraulics. Water appears soft, yet it can move masses, cut valleys, generate energy, and destroy structures.

Force is not necessarily rigid. It depends on volume, flow, pressure, duration, and geometry.

Observation came first. Scientific law came second.

Books on explorers and geology offered another scale. The maps were no longer complete. The ground itself became an archive.

Strata recorded durations, pressures, displacements, disappearances, and metamorphoses.

The world appeared as a superposition of layers, structures, and temporalities.

The house had become an informal laboratory.

Things had to be opened.

Thierry worked on mechanics and put his hands inside engines. That required accepting dependence on the manufacturer’s logic and, sometimes, on the tutelage of the garage technician who knew the machine better than the person who owned it.

Open. Dismantle. Observe. Identify the functions. Understand the articulations. Locate wear. Repair. Reassemble. Test. Dismantling was not destruction. It was a method of knowledge.

This distinction would remain essential. An engine can be deconstructed. So can a legal architecture, a database, an institution, a market, or a religion. Destruction erases. Deconstruction makes readable.

An engine taught interdependence. A symptom could appear in one place while its cause lay elsewhere.

The machine emitted signals: vibration, temperature, smell, voltage, leakage, resistance.

The notebooks began to gather these observations. Dismantle to see. Note in order to compare. Sketch to understand the arrangement. Date the observation. Measure. List. What had been experienced became data.

The first traces of a method were already there: observe, name, classify, connect, verify, transmit.

The laboratory notebook was both memory and protection against bias. Human memory reconstructs. It simplifies, exaggerates, and sometimes moves the order of events.

The present dialogue with AI is, in another form, a new laboratory notebook: a way of fixing data before time distorts it.

Writing preserves the state of an observation before the later result changes the memory of what had been perceived initially.

Much later, archives, databases, and the Art Market would impose the same rule: information must have provenance, date, structure, and the possibility of cross-checking.

The family laboratory had taught a broad conception of knowledge. Theory had to be tested. Observation had to be compared. Intuition had to take form. A machine had to be understood through the relationships among its parts.

At around fourteen and a half or fifteen, Thierry left this cocoon. He moved from the protected laboratory to a much larger terrain.

He carried with him maps, adult books, mechanics, doubt, rural learning, his father’s method, and the habit of questioning authority.

He also carried the wound of the expression “kaleidoscopic discourse.” He would have to turn multidirectional thought into an operational method and prove that linking disciplines did not amount to disorder.

Law, business, dematerialization, databases, networks, the Art Market, sculpture, and the Abode of Chaos would later become its testing grounds. None of them appeared suddenly.

Before the database, there was already the map.

Before the network, there was already the circuit.

Before the algorithm, there was already classification.

Before monumental sculpture, there was already matter dismantled, tested, and reassembled.

Before the author of the Abode of Chaos, there was already the child who refused to remain in front of the closed machine.

The end of childhood would therefore not mark the disappearance of the laboratory. It would mark its extension to the whole world.