Editor’s note: The first essay The Argument ever published asked, “How do we live with each other?” To mark our first anniversary on August 17, we’re returning to that founding question with a series about what we owe people whose beliefs and choices we may find difficult to accept or even understand. During this series, familiar voices at The Argument and new voices alike will approach that question from different angles.
Today, Derek Thompson asks what society owes the people who bear the costs of technological change. Behind the bloodless phrase “transition costs” are people whose lives and livelihoods really are at stake, even if humanity is better off. But, as Thompson argues, people and institutions can shape what happens next.
Two popular claims about artificial intelligence seem to be at odds. One side argues that AI might be profoundly dangerous. Between the cybersecurity risk, the bio-security risks, and the bit about ending humanity, there are a lot of reasons for concern. Another side argues that intellectuals have a tendency to neurotically worry about every new idea and that the alarm surrounding AI is another in a succession of technological moral panics.
But looking back at the early critics of technologies we now consider indispensable, it’s striking how much reason they had to fear — and how much work it took to address those dangers.
Even if the tech accelerationists are right and the existential freak-out over AI is completely misguided, that doesn’t translate exactly to “calm down.” Even “normal” general-purpose technologies are initially volatile and dangerous, even those we come to think of as unalloyed goods.
The printing press brought wars and witch panics to Europe before it gave us literary fiction and calculus. The Industrial Revolution destroyed millions of lives before its growth saved billions of lives. In the early years of electricity, people and horses were electrocuted with some regularity, and power utilities collapsed after the market crash of 1929.
Practically every major technological revolution has been, at least partly and briefly, an economic, social, or political catastrophe. From this reading of history, it would be odd if AI arrived without significantly destabilizing some part of our world for the worse.
Many of the boosters of artificial intelligence, by insisting that it is only a normally powerful technology,1 forget just how wrenching most normally powerful technologies can be. Let’s start with a machine that now seems so benevolent that its demise is often held up as a cultural tragedy: the printing press.
1. The printing press
Imagine the year is 1489. You are an intelligent and progressively minded cobbler living somewhere near modern-day Berlin. A neighbor swings by your home to ask you a question that’s been on his mind recently. This new thing, the printing press. Is it … good?
Not a chance, you reply. For one, it’s an economic boondoggle. Its inventor, Johannes Gutenberg, died broke a few years back after being sued by his financial backer. In the “print rush” of the moment, too many copies of too many titles have landed with a thud before an illiterate continent, and print shops are failing left and right. A letter-making machine for a world that doesn’t know what letters are! What a joke.
And it’s not just a bubble, you point out. In the few cases where people can read, they’re reading things that are turning them into homicidal monsters. A “witch craze” has gripped Central Europe, with tens of thousands of people being executed for witchery. (A 2024 study in the journal Theory and Society found that each new edition of the European treatise Malleus Maleficarum (Hammer of Evil-Doers) — issued more than a dozen times between the 1480s and 1520s — triggered an increase in witch trials.)
