Objectian

Innovation and Its Waste

Every great invention is built from what someone else threw away. The greatest waste of all is the one we have not yet learned to see.

In the spring of 1856, during the Easter holiday, an eighteen-year-old chemistry student named William Henry Perkin set up a makeshift laboratory in his family’s home in the East End of London. His teacher, August Wilhelm von Hofmann, had suggested a grand ambition: to make quinine, the precious anti-malarial drug, from chemicals in coal tar. Coal tar was the black, sticky residue left over when Victorian gasworks turned coal into the gas that lit London’s streets. It was, by any reasonable definition, rubbish.

Perkin failed. He did not make quinine. What he made instead was a dark sludge, and when he dissolved it in alcohol, it turned a startling purple. Rather than discard it, he tested it on silk. The colour held. Within two years he had left his studies, persuaded his father to finance a factory, and launched the first synthetic dye. Mauve became the colour of an era, and Perkin’s accident opened the modern chemical industry. From the same waste product would come later dyes, medicines and materials.

The story is usually told as a parable about serendipity. It is better understood as a parable about waste. The essential ingredient was not luck. It was a young man who looked at what everyone else regarded as refuse and asked what it might contain.

Innovation has always worked this way. Progress is rarely a leap from nothing. It is a process of recovery: of noticing value in what has been discarded, and putting it to use. The waste of one generation becomes the raw material of the next.

Yet there is one form of waste so large, and so constant, that we have barely noticed it. It is the experience of physical things.

Consider what an ordinary object learns in its lifetime. A pump in a water-treatment works passes millions of litres, through floods and droughts, through cold winters and grit-laden summers. A battery in a delivery van is charged and drained a thousand times, in heat and frost, by careful and careless drivers. A batch of steel is rolled, welded, loaded and fatigued. Each of these things undergoes a vast, unrepeatable experiment. Each records the results in the only medium available to it: its own condition.

And then, almost always, the results are lost. The pump is replaced. The battery is recycled. The steel is melted down. A little of the experience survives in maintenance logs and warranty claims, in an engineer’s memory or a field report. Most of it vanishes. The designers of the next pump, the next battery and the next alloy begin again, guided by laboratory tests that can only approximate what the world will do.

We know what it costs when experience is recovered too late. In January 1954, a de Havilland Comet, the world’s first jet airliner and the pride of British engineering, broke apart in the sky near the island of Elba. Three months later another fell into the sea near Naples. The aircraft were grounded. Investigators recovered wreckage from the Mediterranean floor, and at the Royal Aircraft Establishment in Farnborough engineers sealed a Comet fuselage inside a vast water tank and pressurised it again and again, simulating flight after flight, until it failed. The culprit was metal fatigue, which had begun at the corner of an opening in the cabin roof.

The lesson was learned, and it was inherited. Every airliner built since has been designed and tested with the Comet’s experience in mind. But it was learned in the most expensive way imaginable: by tragedy, followed by a heroic act of recovery. The experience had been there all along, accumulating silently in the metal of every Comet in service. No one could read it.

The finest innovators have always understood that experience is an asset. In the 1770s Josiah Wedgwood, the potter who helped to industrialise Britain, spent years perfecting the fine stoneware he called jasper. He recorded thousands of trials in his experiment books, each one numbered, each one noted, the failures as carefully as the successes. Wedgwood knew that a failed experiment, properly recorded, is not waste. It is the most valuable thing a workshop can own.

Some of the most admired objects ever made are the products of exactly this kind of patient iteration. From the 1660s until his death in 1737, Antonio Stradivari refined the violin in his workshop in Cremona, adjusting its arching, its proportions and its varnish across decades of instruments. The Porsche 911, introduced in 1963, has been reworked generation after generation for more than sixty years while keeping its essential form. Each is a masterpiece of iterative design.

Yet neither ever had a common, whole dialogue with its own past. Stradivari’s knowledge lived in his hands and his workshop, and much of it died with him; makers have spent three centuries trying to recover it. Each new 911 carried lessons from its predecessors through engineers, racing teams and warranty files, but never through the cars themselves. What every individual instrument and every individual car endured, how it aged and what failed was scattered, partial or lost.

What Wedgwood did by hand for one product, and what generations of makers did imperfectly for the violin and the 911, Objectian is designed to do for the physical world. It is a common language in which an object, a material or a system can record what it has experienced, attribute that experience to its source, and pass it to its successors. The battery that aged well can teach the battery that follows it. The pipe that corroded early can warn the pipes laid after it. The crop that survived a dry year can inform the planting of the next.

This is not a promise that every object will become intelligent. It is something more modest, and more consequential. It is a way of ensuring that experience, which every physical thing already accumulates, is no longer thrown away.

Perkin found a fortune in the waste of the gasworks. Wedgwood found one in the waste of his kilns. The next great source of innovation may be hiding in plain sight, in the accumulated experience of everything we make, use and discard. The question, as it was for an eighteen-year-old in an East London kitchen, is whether we are willing to look at what we have been throwing away.

Inheritance Without Genes

Lamarck was wrong about giraffes. He may be right about machines.

In 1809 the French naturalist Jean-Baptiste Lamarck published a bold theory of how living things change. Creatures, he argued, acquire characteristics through effort and habit during their lives, and pass those characteristics to their offspring. The giraffe that stretched for the highest leaves would, in his view, bequeath a slightly longer neck to its young.

It was an elegant idea, and in biology it was wrong. Charles Darwin’s theory of natural selection, and later the science of genetics, showed that a parent’s experiences do not, in any simple way, rewrite the inheritance of its children. A blacksmith’s muscles do not pass to his son. Each generation of living things begins, in essence, with the same instructions as the last, and learns its own lessons from the beginning.

Human culture escaped this limitation. We invented language, writing and apprenticeship, and so we could pass on what we had learned rather than only what we had been born with. A master craftsman could teach an apprentice in years what had taken him decades to discover. Knowledge became cumulative. That, more than any single invention, is why the pace of human progress accelerated.

Our machines and materials, however, remain stubbornly pre-Lamarckian. A new generation of turbine blades inherits the design of the last, but not its experience. The specific blade that ran hot for ten thousand hours, the alloy that cracked under a particular rhythm of stress, the cell that lost capacity faster in a cold depot: these individual histories are almost never handed down. Each new object begins life knowing only what its designers could test in a laboratory.

Objectian proposes a different arrangement. It proposes that physical things should be able to inherit acquired experience, as Lamarck imagined, but by design rather than by biology. A predecessor’s recorded experience could be offered to its successor, so that the successor begins not from zero but from what came before.

Here, however, history offers a warning, and it is one that shapes everything about how Objectian works.

For centuries, physicians attempted to transfer blood from one body to another. The results were unpredictable and often fatal. Some patients recovered. Others died in agony, and no one could say why. The answer came in 1901, when the Austrian physician Karl Landsteiner showed that human blood falls into distinct groups, and that mixing incompatible groups causes a catastrophic reaction. The problem had never been transfusion itself. It was transfusion without a test of compatibility. Once blood could be matched, transfusion became one of the great life-saving practices of modern medicine, and Landsteiner was awarded the Nobel Prize.

Inherited experience is like blood. It can be lifesaving, and it can be lethal. Experience gathered by a battery cycled gently in a temperate laboratory may be dangerously misleading for a battery driven hard in a hot climate. Experience recorded by one alloy may not apply to another, however similar their labels. The greatest danger is not ignorance. It is inheriting the wrong knowledge with confidence.

That is why Objectian does not allow experience to be passed on simply because two objects are described as similar. Descriptions can be wrong. Labels can be mistaken. Records can be confused. Instead, before a successor inherits, its own physical evidence is used to test whether the inherited experience actually applies. If the predecessor’s experience predicts how the successor really behaves, it may be accepted. If it does not, it is refused, however compatible the paperwork may appear.

In our first computational tests, using published battery-ageing data, we deliberately gave one cell a false description. Methods that relied on descriptions accepted it and were misled. The recipient’s own measurements did not; they rejected the impostor. It is an early result, on a small scale, and it must be tested far more widely. But it points to the principle that matters most: inheritance is only as good as the test that governs it.

There is one further safeguard that biology never needed. Inherited experience in Objectian is never merged invisibly into the successor. It is kept separate, attributed to its source, and weighed against the successor’s own experience as that experience accumulates. A child may forget where it learned a lesson. An object speaking Objectian does not.

Lamarck imagined a world in which the striving of one generation became the starting point of the next. Nature declined to work that way. Engineering need not. If we can give physical things a language in which to inherit experience, and a test that keeps that inheritance honest, then every generation of what we build can begin where the last one ended.

The Right to Be Forgotten, for Things

A memory that cannot be corrected is not knowledge. It is a liability.

In December 1912, a solicitor and amateur antiquarian named Charles Dawson stood before the Geological Society in London and presented a remarkable discovery. From a gravel pit near the village of Piltdown in Sussex he had recovered fragments of a human-like skull and an ape-like jaw. Together, they seemed to show the long-sought missing link between apes and humans, and it had been found, as many in Britain were pleased to note, in England.

For four decades Piltdown Man took its place in the story of human origins. Textbooks described it. Scholars built theories around it. And its influence spread in ways that were less visible. When the anatomist Raymond Dart announced in 1925 that he had found a genuine early human ancestor in South Africa, the Taung Child, many experts were sceptical, in part because his fossil did not fit the pattern Piltdown appeared to establish. A false ancestor was distorting the reception of a true one.

In 1953 scientists at the Natural History Museum, among them Kenneth Oakley, Joseph Weiner and Wilfrid Le Gros Clark, subjected the remains to new chemical tests. Piltdown Man was a forgery. The skull was human, the jaw came from an orangutan, and the bones had been stained to appear ancient. The exposure was swift and decisive. The harder work came afterwards: tracing every argument, every interpretation and every assumption that had rested on the fraud, and setting them right.

Piltdown is usually remembered as a story about deception. It is more usefully remembered as a story about dependency. The forgery was one object. The damage lay in everything that had been built upon it.

Every system that accumulates knowledge faces the same problem. Science has its retractions; medicine, its withdrawn treatments; industry, its recalls. In each case the difficult question is not only what was wrong, but what else was affected by it. Which conclusions relied on the flawed data? Which designs incorporated the faulty part? Which decisions should now be reconsidered? Too often, the honest answer is that no one knows.

For a language of inherited experience, this question is not a footnote. It is fundamental. If physical things are to pass experience to their successors, then some of that experience will one day prove to be wrong. A sensor will be found to have drifted. A test rig will turn out to have been miscalibrated. A record will be discovered to have been confused with another. When that happens, the error will not sit quietly where it began. It will have been inherited, and perhaps inherited again.

Objectian is therefore designed to forget as carefully as it remembers. Every inherited experience carries the identity of its source and a record of every successor that relied upon it. When a source is discredited, the language does not simply delete it and hope for the best. It follows the line of inheritance forward, identifies every descendant whose knowledge depended on the faulty evidence, and marks each one for correction. Where a decision about a physical operation rested on that knowledge, the decision is reconsidered.

In our early tests we deliberately corrupted the experience of a first-generation battery cell and allowed it to pass to two later generations. Tracing the damage precisely, and correcting only what was affected, restored the decisions as well as starting again from scratch, at around half the effort. Blocking everything connected to the faulty source would have been safer in appearance but crippling in practice. The tests also taught us something humbling: correcting bad evidence does not make every decision good. Governance can remove a known error. It cannot supply knowledge that was never there. We regard that as a principle, not a weakness.

There is a phrase in European data protection law, the right to be forgotten, that describes a person’s ability to have certain information about them erased. Objectian applies a version of that idea to things. An object has a right, and its owner a duty, to have false experience withdrawn from its record and from the records of everything that inherited it.

This matters for more than accuracy. It is the foundation of trust. No manufacturer, regulator or researcher should rely on inherited experience unless they can be confident that it will be corrected when it proves untrue. A system that can only accumulate will eventually be poisoned by its own mistakes. A system that can trace and withdraw them can grow more reliable with age.

It took forty-one years to expose Piltdown Man, and longer still to undo its influence. The physical world deserves better. If objects are to remember, they must also be able to forget, precisely, accountably and in time.