Geoffrey Hinton, the father of modern neural networks, and a 2024 Physics Nobel laureate, sees the future of AI in stark terms: Big Tech can only justify its massive investments if algorithms replace workers. “They’re spending $420 billion on AI. That money only makes sense if they lay people off,” he told Bloomberg TV, as reported in this Medium article.
Anthropic CEO Dario Amodei shares Hinton’s bleak prediction: “80% of jobs could be radically transformed or disappear in the next decade,” highlighting the unprecedented speed of AI’s disruption. Both messages sound apocalyptic, but we’ve been here before: every great technological revolution begins by destroying jobs, and then creates new ones.
During the first industrial revolution, steam engines destroyed artisanal trades and family workshops. In the early 19th century, fearing for their livelihoods, the Luddites burned the new, steam-powered mechanical looms. Within a few decades, total employment had not only recovered: it had multiplied. The new jobs were different, concentrated in cities, more technical and more specialized, but also better paid and more stable.
Electrification repeated the process, and so did computers in the 1970s, and the internet in the 1990s. In each case, technology did not destroy human labor, only mechanical labor, freeing up time and capacity for more complex and creative tasks. The problem was never the machine, but the transition period between one technology and the next.
What’s different now is the speed of technological change. AI does not require building factories or transporting raw materials: it is digital infrastructure, governed only by the speed we can build new data centers, which means that its impact is faster and more visible. But even so, to think that the end result will be a world without work is as reductionist as assuming that the automobile destroyed jobs related to horse-drawn vehicles without spawning new tertiary industries.
Recent economic evidence suggests we are living through an early stage of what some analysts are calling a technology shock. A European Central Bank study highlights the rapid adoption of AI throughout Europe, although regular use remains low and its effects are not yet fully reflected in aggregate productivity. Something similar is observed by economists at the National Bureau of Economic Research (NBER), outlined in “Artificial Intelligence and the modern productivity paradox”: the promise of efficiency is real, but takes time to materialize. A third of my doctoral dissertation uses the productivity paradox and invalidates it in the case of small and medium-sized firms, and I can attest that this is exactly what happens: productivity gains are not easy to measure, but they exist.
In the microeconomic field, MIT has shown that the adoption of AI in industry initially reduces productivity, and is then followed by a rebound when organizations learn to integrate it structurally. A recent OECD report concludes that the macroeconomic benefits will be gradual and will depend on factors such as institutional support, investment in human capital and the adaptability of the productive fabric.
In other words, history is repeating itself. First comes disruption, then readjustment, and finally, expansion. Efficiency is not the enemy of work, but its evolution. Each technological leap has displaced trades, yes, but it has also expanded the limits of what we understand by work. In 1850 it took eight hours to make a shirt, today that same human energy is used to design fashion, coordinate global chains or develop software that predicts trends.
The anxiety generated by AI is understandable and occurred with each of the previous disruptions as well, but forgetting the lesson of history would be a mistake. Technology destroys specific jobs, but not employment. It sends certain professionals into unemployment at an individual level, but generates an increase in employment and wealth in the aggregate. What changes are the skills needed and the speed of change. Instead of fearing automation, we should be concerned with how to prepare people for what’s coming, as well as to provide them with ways to survive and stay above the poverty line as they make the transition.
Geoffrey Hinton and Dario Amodei may be right about one thing: the changes are going to be radical and affect many people. We are already seeing it. It’s little consolation for those who lose their jobs, but that doesn’t make AI an inevitable tragedy for humanity: it is only the repetition of a cycle that has been transforming the expression of our productivity for two centuries, if not more. And if history proves anything, it is that innovation has never left humanity with less work; it has only demanded a higher level of intelligence.
(En español, aquí)
—
This post was previously published on MEDIUM.COM.
—
If you believe in the work we are doing here at The Good Men Project, please join us as a Premium Member today.
All Premium Members get to view The Good Men Project with NO ADS.
Need more info? A complete list of benefits is here.
—
Photo credit: iStock.com
