Why a 250-year-old book indirectly shapes the AI debate
On March 9, 1776, a Scottish moral philosopher published a long, dense book with a mundane academic title: The Wealth of Nations. Its author, Adam Smith, could hardly have imagined the technologies that would define the twenty-first century. Yet two hundred and fifty years later, his ideas remain uncannily relevant to one of the most pressing questions of our time: what happens when machines begin to perform the work that humans once did.
The debate surrounding artificial intelligence—whether it will enrich society, displace workers, concentrate power, or destabilize economies—turns out to be, in essence, a debate that Smith already framed. His central insight was revolutionary in its day. Wealth, he argued, does not come from hoarded treasure, imperial conquest, or favorable trade balances. It comes from productive capacity — from the ability of a society to produce goods and services efficiently through specialization, innovation and exchange.
His famous example of the pin factory illustrated the point: divide the work into smaller tasks, and output multiplies dramatically. He observed that 10 specialized workers could produce 48,000 pins daily — 4,800 each — compared to perhaps only one pin if working alone. Huzzah!
For two and a half centuries, technological progress has repeatedly vindicated Smith’s intuition. Mechanization replaced muscle. Electricity replaced steam. Computers replaced vast amounts of clerical work. Each wave of innovation initially displaced workers, but in the long run the economy adapted. New industries emerged, productivity rose, and living standards improved. The Luddites were angry about the textile mill? What fools! They were retrained, society did just fine, and everyone now has Zara or American Eagle.
Artificial intelligence appears to be the next stage in that story. For the first time, machines are not merely augmenting physical labor but increasingly performing cognitive tasks — writing, analyzing, diagnosing, even creating. In Smith’s terms, AI represents a dramatic extension of the division of labor, shifting entire categories of work from human minds to algorithms. For white collar types accustomed to the status that comes with doing these tasks, it is a nightmare.
If the historical pattern holds, this should ultimately be good news. Higher productivity means cheaper goods, new industries, and greater prosperity.
But theory is one thing, and human wiring is another.
Readers of AQL will recognize that we have wrestled with this question before. I come to it with a certain reluctance to sound alarmist. I hold two advanced degrees in computer science and once pursued a PhD in the field, teaching algorithms and theory along the way to unfortunate students. I am therefore not eager or comfortable presenting myself a latter-day Substack Luddite.
Still, anyone trained in mathematics develops a habit that can be difficult to suppress: the habit of considering edge cases.
What if the adjustment mechanism that Smith implicitly relied upon — technological change displacing some workers while creating opportunities for others — breaks down? What if automation does not arrive gradually, sector by sector, but rapidly and across much of the economy at once? What if large numbers of workers were displaced simultaneously?
Classical economics assumes that markets eventually rebalance. Labor moves into new industries, capital flows to new opportunities, and innovation creates jobs that were previously unimaginable. For most of modern history, this has proved broadly true. But artificial intelligence raises the unsettling possibility that machines might perform not just some human tasks, but many of them — perhaps even most economically valuable ones.
Would the economy still adjust smoothly under those conditions? Would workers be retrained quickly enough? Or would the pace of technological change outstrip society’s ability to absorb it? AI luminaries are themselves among the biggest of the alarmists. Geoffrey Hinton, winner of the 2024 Nobel Prize who is widely seen as one of the fathers of artificial intelligence, has argued he economy cannot make money from AI unless human labor is replaced at devastating scale.
We are at the point where the question moves beyond economics into psychology.
Smith understood that employment was a source of income, but he may not have fully appreciated how deeply it is also a source of identity. Work structures our days, anchors our social roles and confers dignity. Even in a wealthy society capable of distributing income through mechanisms such as universal basic income, the loss of meaningful work could create profound social tensions.
Human beings are not merely rational consumers of goods and services. They are status-seeking creatures who measure themselves against others and derive purpose from participation in productive life.
A world in which large numbers of people find themselves economically unnecessary could therefore become unstable. Without work, individuals —particularly young men, historically the most volatile demographic — may find their frustrations expressed in ways that are not especially constructive. Societies built on abundance alone may not prove as tranquil as economists sometimes assume. I am not eager to see what happens then,
It is here that the conversation about artificial intelligence begins to intersect with two of Smith’s most famous intellectual successors: Karl Marx and John Maynard Keynes.
Marx accepted Smith’s observation that capitalism was astonishingly productive, but he believed its benefits would ultimately accrue to those who owned capital rather than those who supplied labor. Artificial intelligence, owned and operated by a relatively small number of technology firms, makes that concern newly relevant. He would have a field day today. The unsympathetic likes of Peter Thiel, Elon Musk, Mark Zuckerberg et al would be manna for Marx.
Keynes, writing after the Great Depression, warned of “technological unemployment” — a condition in which machines replace workers faster than new jobs can emerge. Keynes remained optimistic that this would eventually yield a richer society with far shorter workweeks. But he certainly looks prescient.
Between Smith’s optimism, Marx’s skepticism and Keynes’s caution lies the unresolved question of the AI age. Right now, obviously, there is no definitive answer. Banning automation doesn’t seem the way to go, but politics may say otherwise. Some form of regulation may be in the office, to be sure.
Technological revolutions have repeatedly confounded both pessimists and optimists. The industrial revolution created wealth on an unprecedented scale, but also social dislocation. The digital revolution generated new industries that would have been unimaginable a generation earlier. Artificial intelligence may follow a similar pattern—or it may represent something fundamentally different.
What can be said with confidence is that Smith’s framework remains indispensable. He understood that markets can generate extraordinary prosperity, but only when certain conditions are preserved: competition, fairness, education and institutions capable of maintaining public trust.
Two hundred and fifty years after the publication of The Wealth of Nations, those conditions remain the foundation of modern economic life.
And on this anniversary it is worth pausing to recognize the extraordinary intellectual achievement of the man who first articulated them. If his ideas are any guide, technological progress ultimately expands prosperity rather than destroying it. That would mean that even in the age of artificial intelligence, the optimism of a Scot writing in 1776 may still be the best bet we have. At the end of the day, Adam Smith might tell us, things will probably turn out all right.
How charming that we remember him, a quarter-millennium later! If for nothing else, on this day of all days, his assurances are good enough for me.
