Rubber was useless.

A latticework reading of Veritasium's story of vulcanization — how one man's obsessive failure transformed a melting liability into the material foundation of the modern world.

▶ Play video

Photo: Veritasium

Why this story belongs in the latticework.

In the summer of 1834, a rubber warehouse in Woburn, Massachusetts opened its doors to reveal not inventory but catastrophe: every product had melted into a single foul-smelling goo. Natural rubber — the miracle material everyone was counting on for waterproof coats and shoes — had a fatal flaw. It melted in heat and shattered in cold. The company behind it, Roxbury India Rubber, was one failed season away from total ruin. This is the kind of failure the Farnam Street latticework was built to handle: a technology that failed not because the idea was wrong but because the material science was unfinished.

The story of Charles Goodyear and vulcanization is one of the cleanest illustrations in industrial history of feedback loops, activation energy, survivorship bias, and the dark side of obsession as a tool. It also introduces two models worth adding to the latticework: a pattern for identifying when a material barrier — not a conceptual one — is blocking an entire industry, and the phenomenon of accidental serendipity, in which the discovery happens to the person who stayed at the table long enough to notice it.

The video covers only the opening chapter of the Goodyear story, but the patterns it demonstrates are complete even in their compressed form. What follows reads those patterns against the canon.

Old models, clarified by melted rubber.

The most vivid reinforcement in the story is of activation energy. Natural rubber existed; everyone knew it had potential; coats and shoes were being manufactured and sold. The barrier was not awareness or capital or even demand — it was one unsolved material problem. In activation energy terms, the rubber industry had enough energy to sustain a reaction, but not enough to cross the transition state. The entire market was being held behind that single barrier, and all the investment in it was essentially stranded there. When Goodyear crossed the barrier with heat and sulfur, an enormous release followed — not because he discovered rubber, but because he removed the last constraint on using it.

Feedback loops — specifically negative ones — run through both the product failure and Goodyear's decade of experiments. The rubber melted; customers returned it; the company took losses; management buried the rotting inventory to hide the evidence. Each loop reinforced the previous signal: the problem was real, it was consistent, and it wasn't going away. For Goodyear, the loop ran differently: each failed formula confirmed that the barrier was real and that chemistry, applied persistently, was the right lever to pull. The loops provided the same information; the difference was whether you read failure as a dead end or as data.

Inversion applies to the problem definition itself. The rubber industry kept asking: how do we make better rubber products? Goodyear's implicit reframe — which took years to become explicit — was: how do we change the properties of the rubber itself? The product-level question was unanswerable with product-level tools. The material-level question was answerable, but only once someone moved down a level of abstraction.

Models the story bends.

The classic account of rational actor decision-making has trouble with Goodyear. He had gone bankrupt once before rubber entered his life; he spent years in poverty while pursuing the vulcanization problem; he used his family's kitchen as a lab; he pawned his children's schoolbooks. A rational actor would have stopped. The model of expected value — which says agents persist when the expected return justifies the cost — does not generate Goodyear's behaviour. What generates it is closer to an unconditional commitment: the problem became a terminal value, not an instrumental one. This is not a flaw in the person; it is precisely the feature that made the discovery possible. No ordinary cost-benefit calculation would have kept him in the search long enough.

The more comfortable model that gets bent is incremental iteration. The normal advice for product development is to ship early, get feedback, iterate. Goodyear couldn't ship. Every version of his rubber failed in the summer. The feedback loop was catastrophic rather than informative. The lesson isn't that iteration is wrong; it's that it requires a stable evaluation environment. When the environment itself is the source of failure — weather, temperature, time — the iteration model doesn't apply in its standard form. What replaces it is something closer to controlled experiment: isolating the variable (temperature stability) before shipping anything at all.

Survivorship bias bends in an interesting direction here. We remember Goodyear because he succeeded. But the Roxbury manager's instinct — to bury the evidence of failure and move on — was not irrational. Many investors and operators in the early rubber industry correctly concluded that the material science problem was not yet solved and cut their losses. They are not in the story because they didn't produce an outcome. The canvas doesn't show them. Yet their decision was defensible with the information they had.

New entries for the latticework.

The Goodyear story introduces what we might call the Material Barrier pattern: a situation where an entire industry or technology domain is blocked not by a conceptual or market problem but by a single physical or material constraint. The pattern is recognisable in retrospect — batteries for electric vehicles, protein structure for drug discovery, semiconductor lithography for computing. In each case, enormous value was stranded behind one material problem; the release, when it came, was disproportionate to the apparent scale of the solution. The latticework lesson: when you see a market with obvious demand, clear distribution, and existing investment that is nonetheless not reaching scale, ask whether there is a material constraint hiding below the product layer.

The second new model is Accidental Serendipity — distinct from pure luck. Goodyear's discovery of vulcanization reportedly came from accidentally dropping a sulfur-rubber mixture on a hot stove. But the accident only mattered because he was there, looking, with the problem loaded in his working memory, and with enough domain knowledge to recognise that what he saw was different. The structure is: prepared mind + anomaly + recognition. Luck provides the anomaly; the prepared mind is what converts it to discovery. This generalises: most breakthrough discoveries described as accidental share this three-part structure. The preparation is the variable the person controls; the anomaly is not.

The third is a refinement of obsession as a tool. The standard latticework warning about obsession is that it produces tunnel vision and sunk-cost blindness. The Goodyear story suggests a nuanced version: obsession with a well-specified problem is different from obsession with a particular solution. Goodyear stayed committed to the problem (temperature-stable rubber) while iterating freely across hundreds of solutions. That combination — terminal commitment to the problem, instrumental flexibility about the solution — is the productive form. Obsession becomes dangerous when the commitment transfers from problem to solution.

When to reach for which.

The latticework, after rubber.

The Goodyear story is useful precisely because it resists the clean arc of the innovation narrative. He did not have a vision and execute it. He had an obsession and survived long enough to notice an accident. The latticework lesson is less about entrepreneurship than about the structural conditions under which certain kinds of discovery become possible: a well-specified problem, an evaluation environment that provides real signal, a prepared mind, and enough time. All four have to be present. Remove any one and the accident doesn't convert.

Rubber had potential, but in its current form, it was just too problematic. If anyone could fix that, they'd stand to make a fortune. — Roxbury manager to Charles Goodyear, 1834

The practical residue: when you encounter something that is demonstrably useful in theory but catastrophically unreliable in practice, the instinct to write it off is correct — for now. The question worth holding is whether the failure is intrinsic (the idea is wrong) or material (the supporting technology hasn't arrived yet). That distinction is not always obvious in the moment. It almost never was for rubber either.