Leonhard Euler

In the [YOU] on AI Field Guide

The cycle that began with [YOU] on AI asks what it would mean to see the machine clearly—without hype's narcotic or fear's paralysis. Euler is the figure who supplies the substrate beneath everything the cycle examines: every neural network is a graph, and Euler founded the mathematics of graphs. His Seven Bridges proof is therefore not a historical curiosity but a live schematic of the thing now reshaping human life: an architecture of dots joined by lines, reasoning about what is connected to what, discarding everything else.

His lens cuts through the field's loudest confusion. We are constantly invited to be impressed by volume—the number of benchmarks a model passes, the words it generates per second, the tasks it performs. Euler is the human refutation: the most prodigious producer in the history of mathematics, whose output exceeded anything a machine yet matches in depth, and who could not be identified as a thinker by his volume alone. The machine can match the exhaust. What Euler's case demands we ask is whether comprehension rides behind it. Volume, he shows, proves nothing.

The book's deepest question concerns computation in the dark. Euler lost his sight in his fifties and produced a substantial fraction of his work completely blind, holding vast calculations in memory and dictating to assistants. He thus proves that genuine, understanding-laden thought can proceed without vision, by the pure manipulation of symbols—which is exactly what large language models do. This eliminates the easy answer. We cannot say the machine merely shuffles symbols without comprehension, because Euler shuffled symbols without sight and no one doubts he understood. What the machine lacks, if it lacks anything, is something harder to locate than perception.

Euler also supplies the exponential, and that gift is concrete. The scaling laws governing today's AI trace the exact curve he characterized when he placed e at the center of analysis: growth that compounds on itself, deceptive early and overwhelming late. To read Euler is to understand, from first principles and in a mathematician's own hand, why the present moment feels simultaneously incremental and vertiginous—and why the people inside an exponential are the last to see its shape.

Origin

Born in Basel in 1707, the son of a Calvinist pastor, Euler was sent to university at thirteen and placed under the informal supervision of Johann Bernoulli, one of the greatest mathematicians of the age. Bernoulli recognized the boy's gifts and made himself available for questions on Sundays; the arrangement proved sufficient. Euler earned his master's degree at sixteen, joined the newly founded Imperial Academy of Sciences in Saint Petersburg at nineteen, and never stopped. He published his first major results in his early twenties and thereafter produced at a rate that astonished contemporaries and still astonishes historians.

The Königsberg paper arrived in 1736 and is the seed of the graph-theoretic world we inhabit. Euler noticed that a puzzle the townspeople had failed to solve by trial was actually a question about structure—about which parcels of land were connected to which and how many bridges made each connection. He stripped the city to a diagram of four points and seven lines and proved, from the degree of each point alone, that the desired walk was impossible. It was the first result in what would become graph theory and, in the same stroke, a founding gesture of topology: the recognition that certain truths live in connectivity, not in geometry.

He spent most of his career at the imperial academies of Saint Petersburg and Berlin, moving between them as patronage and politics shifted. In his early thirties he lost sight in his right eye; decades later a cataract took the left, and a failed surgery left him almost entirely blind. His output did not slow. By some accounts it accelerated, freed from distraction. He died in 1783 in Saint Petersburg, reportedly mid-calculation, and the Academy continued publishing the backlog of his papers for more than forty years after his death.

Key Ideas

The graph as substrate of the connected world. Euler's reduction of Königsberg to four nodes and seven edges was the founding act of graph theory: the mathematics of which things are connected to which. Every neural network, every router, every social graph, every knowledge base is an instance of the object he isolated. The odd-degree criterion—a traversal crossing each edge exactly once requires zero or two vertices of odd degree—was the first theorem in a science that now underlies the architecture of the entire connected world.

The exponential and the number e. Euler gave e its symbol, its central role, and its defining property: the unique base for which the exponential function equals its own rate of change. This makes e the mathematics of unconstrained growth—a quantity that increases in proportion to its own size. The scaling laws of contemporary AI trace this curve. The softmax and cross-entropy that govern every model's training are saturated with e. Euler could not have anticipated the application; the mathematics is his, unchanged.

Abstraction that knows what to discard. In every major result, Euler performed the same operation: strip away everything inessential until only the structure remains. He discarded the city and kept the graph; he discarded the shape and kept the Euler characteristic. The critical difference from a machine's abstraction is that his deletions were licensed by proof—he could demonstrate that the discarded detail genuinely did not matter to the question. A model abstracts by gradient descent, deleting whatever did not reduce training error, with no certificate that the deleted feature was not the crux.

Notation as the substrate of thought. Euler standardized f(x), e, Σ, i, and the modern role of π—not as bookkeeping but as cognitive tools. Good notation offloads part of the reasoning into the form of the representation, making certain truths almost visible. The machine's embeddings, which encode meaning as position in high-dimensional space, are Euler's principle taken to its extreme: a representational system of enormous power built by gradient descent that no one designed and no one can fully read.

Computation in the dark. Euler's most productive years were his blind ones, holding entire architectures of calculation in memory. He proves that genuine understanding can proceed without sight, by symbol-manipulation alone—which is formally identical to what a computer does. This eliminates the easy objection to machine intelligence (it merely shuffles symbols in the dark) without settling the hard question: whether the understanding that indisputably accompanied Euler's blind computation also accompanies the machine's.