Mercadente, Artificial Intelligence and the Wealth of Nations

Bryan Mercadente, “Artificial Intelligence and the Wealth of Nations,” Libertarian Alliance (UK) (24 July, 2026)

Submitted to and actually published in Bryan’s school magazine

Artificial intelligence has become the latest object of fashionable anxiety. Politicians warn of disappearing jobs. Journalists speculate about permanent unemployment. Every week produces another prediction that algorithms will soon replace lawyers, accountants, teachers, computer programmers and anyone else whose work consists largely of processing information. The implication is that this must be a disaster. An economy in which fewer people are employed, we are invited to believe, is necessarily an economy in decline.

This is one of the most absurd assumptions of a political discourse that has been generally absurd for as long as anyone can remember. It mistakes a means for an end. The purpose of an economy is not to maximise employment. It is to maximise production. Employment is simply one of the costs of production. Like every other cost, it is something that prosperous societies have sought to reduce rather than increase. The wealthiest civilisation is not the one that requires the greatest quantity of labour to produce a loaf of bread. It is the one that requires the least.

This truth seems impossible for educated people to grasp. We congratulate governments for creating jobs without asking whether those jobs create anything of value. We speak as if the existence of work were itself a public good, irrespective of what that work consists of. Yet if employment alone were the measure of prosperity, Britain would become richer by abolishing tractors and returning to horse-drawn ploughs. We could ban washing machines and create millions of jobs in domestic laundry. We could outlaw computers and re-employ armies of clerks armed with ledgers and carbon paper. Nobody proposes this because everyone instinctively understands that labour is a scarce resource. The less of it required to produce a given result, the richer society becomes.

Imagine two islands, each inhabited by one hundred people. On the first island, every inhabitant spends sixteen hours each day cultivating enough food to avoid starvation. Nobody becomes an architect, composer, physician or scientist because nobody possesses the leisure to do anything except survive. On the second island, technological innovation enables ten people to produce enough food for everyone else. Ninety people are now free to pursue different activities. Some establish businesses. Some conduct scientific research. Some become artists. Some undoubtedly waste their lives. The point remains that the second island is incomparably wealthier than the first, not because more people are employed, but because fewer people are required to perform the same productive task. Falling labour requirements are not the consequence of prosperity. They are one of its principal causes.

Economic history is little more than the repeated confirmation of this principle. The Industrial Revolution did not make Britain rich because it created factories. It made Britain rich because factories enabled vastly greater quantities of cloth, iron and manufactured goods to be produced with fewer hours of human labour. Agriculture followed the same pattern. Before mechanisation, a large majority of the population worked on the land. Today, a tiny fraction produces enough food not merely for Britain but for export as well. The workers released from agriculture did not disappear. They entered manufacturing. Later, as manufacturing itself became more productive, labour moved into services. At every stage, the same gloomy predictions accompanied the transition. At every stage, they proved mistaken.

Artificial intelligence belongs within this historical sequence. The steam engine reduced the cost of moving physical objects. Electricity reduced the cost of manufacturing. The computer reduced the cost of calculation. Artificial intelligence reduces the cost of manipulating information. It is not a different phenomenon. It is the application of the same economic principle to a different category of human activity.

This point deserves emphasis because modern economies increasingly revolve around information rather than physical production. Drafting reports, analysing data, reviewing contracts, translating documents, writing software, processing insurance claims, preparing financial forecasts and answering routine enquiries, all consume enormous quantities of skilled labour. None of these activities is valueless. The question is whether they require as much human effort as they currently consume. Artificial intelligence increasingly suggests that they do not.

The obvious beneficiaries are consumers. If legal research becomes cheaper, legal advice becomes cheaper. If software development becomes more productive, software becomes cheaper. If architects can explore hundreds of possible designs in the time previously required for twenty, buildings become cheaper to design. The same principle applies throughout the economy. Falling production costs are not an unfortunate side effect of technological progress. They are its entire purpose.

This is why I find the current panic so curious. We do not normally mourn the disappearance of expensive production methods. Nobody regrets that lifting heavy loads no longer requires teams of horses. Nobody campaigns for the restoration of manual telephone exchanges. Nobody insists that accountants should abandon spreadsheets and return to handwritten ledgers. Once a superior technology appears, the previous method rapidly acquires the status of quaint historical curiosity.

Why, then, does artificial intelligence provoke such alarm? Part of the answer lies in the identity of those whose occupations now face disruption. Previous technological revolutions largely affected manual labour. Artificial intelligence strikes most directly at occupations involving routine information processing. This includes many of the professions that have enjoyed increasing prestige during the past half-century. Lawyers, accountants, consultants, policy analysts, communications specialists, financial advisers and administrators all perform tasks that involve structured information processing. Artificial intelligence does not eliminate the need for expertise in these fields. It does reduce the quantity of routine intellectual labour required to exercise that expertise effectively.

The distinction is important. Artificial intelligence rewards genuine competence while reducing the value of routine competence. An excellent engineer becomes more productive because AI enables him to test designs more rapidly and analyse larger quantities of data. A mediocre engineer gains much less because the technology cannot compensate for poor judgement. Likewise, an experienced doctor may use AI to identify unusual diagnoses more quickly, whereas someone lacking medical knowledge cannot rely upon it safely. Expertise is amplified rather than replaced.

Routine administration is different. If the principal function of an occupation consists of moving information between systems, summarising documents, preparing standardised reports or completing established procedures, the economic value of that activity inevitably falls once machines perform much of it automatically. This should surprise nobody. It reflects the declining cost of information processing.

The implications extend beyond economics into culture. Modern Britain has attached high prestige to occupations associated with paperwork, management and administration. Universities have expanded on the assumption that ever larger numbers of graduates would spend their lives processing information inside large organisations. Technical competence became subordinate to credentialism. Practical achievement yielded status to institutional position. The engineer designing a bridge often found himself subordinate to someone whose principal contribution consisted of producing compliance documentation about the bridge.

There has always been something absurd about this hierarchy. Civilisations ultimately depend on those who discover, invent, design, manufacture, transport and maintain the material foundations on which everything else rests. Administrative systems exist to facilitate these activities, not to replace them. Yet developed societies have often behaved as though supervision were more valuable than production, documentation more valuable than discovery, and regulation more valuable than invention. Artificial intelligence may help restore a healthier balance.

If information processing becomes dramatically cheaper, organisations will require fewer people devoted exclusively to information processing. Relative prestige may shift back towards those occupations grounded in scientific understanding, technical expertise and entrepreneurial judgement. That would not diminish the importance of knowledge. On the contrary, it would distinguish more clearly between genuine knowledge and the routine administration of information.

This also explains why education remains indispensable. Some commentators imagine that AI makes learning unnecessary because facts are now available instantly. The opposite is true. The more powerful these systems become, the greater the advantage enjoyed by those capable of questioning them intelligently. Artificial intelligence does not abolish the need for knowledge. It increases the returns to possessing genuine knowledge rather than superficial familiarity.

There is an important distinction between information and understanding. Information can be stored externally. Understanding cannot. A student who possesses no grasp of economics will ask poor questions of an AI system and accept poor answers. A student who understands economics will identify weaknesses, recognise assumptions and employ the technology as a powerful research assistant. The difference lies not in access to information but in the quality of judgement brought to its use.

The same principle applies across every intellectual discipline. Artificial intelligence is unlikely to replace the scientist who understands physics, the historian who can evaluate conflicting evidence or the entrepreneur who recognises commercial opportunities before others do. It will, however, make each of them considerably more productive.

There will undoubtedly be disruption. Some occupations will disappear. Others will change almost beyond recognition. New professions will emerge whose names we cannot presently imagine. Nobody in 1900 foresaw cybersecurity specialists or aerospace engineers. Nobody in 1975 anticipated app developers or cloud architects. The future habitually refuses to conform to the occupational categories of the present. But we should know that societies that embrace productivity-enhancing technologies become richer than societies that resist them. Every previous industrial revolution has increased living standards precisely because it enabled fewer people to produce more. Artificial intelligence continues this tradition. It is not an enemy of civilisation but one of its characteristic achievements.

The real danger therefore lies elsewhere. It is not that machines become too capable. It is that societies become too timid to exploit their capabilities. Britain became the first industrial nation because it repeatedly embraced innovations that many contemporaries regarded as alarming. It prospered because it welcomed productivity rather than fearing it.

Artificial intelligence deserves to be viewed in the same spirit. It will undoubtedly eliminate some familiar ways of working. It will almost certainly create others that we cannot yet imagine. More importantly, it will reduce the cost of producing one of the most valuable commodities in the modern economy: organised information. If previous generations had rejected the steam engine because it threatened coachmen, or electricity because it threatened candle makers, modern civilisation would scarcely exist.

The appropriate response to artificial intelligence is therefore neither panic nor nostalgia. It is curiosity. The technologies that make labour more productive have consistently made mankind richer. There is little reason to suppose that this particular revolution will prove the first great exception to a rule that has governed economic progress for more than two centuries.