The beginning of 2023 marked an extraordinary shift in the global technological landscape as ChatGPT became the fastest-growing application in history, reaching 100 million users in just two months. As Daniel Susskind detailed in his lecture series for Gresham College, this sudden rise of generative AI like Claude and Gemini is not an isolated event but a significant chapter in a decades-long story of machines assuming tasks once reserved for humanity. While early dreams of automation appeared in ancient Greek myths and 18th-century mechanical "automata," the serious scientific pursuit of artificial intelligence only began in the mid-20th century. At the 1956 Dartmouth Summer Research Project, pioneers established a "purist" philosophy, operating on the conjecture that every aspect of intelligence could be precisely described and simulated by a machine. These researchers believed that to build a capable machine, they had to observe and replicate human behavior, viewing human intelligence as the peak of a "capability mountain" that machines could only ascend by mimicking human thought and reason.
This purist approach dominated the field until the late 1990s, when a "Pragmatist Revolution" fundamentally changed the trajectory of AI. The watershed moment occurred in 1997 when IBM’s Deep Blue defeated world chess champion Gary Kasparov.
Related article - Uphorial Shopify
Unlike the expert systems of the 1980s that tried to follow human rules and logic, Deep Blue utilized brute-force processing power to calculate 330 million moves per second. It did not matter that Kasparov could not fully articulate his intuition; the machine outperformed him by functioning in a fundamentally different way. This shift—judging machines not by how they perform a task, but by how well they perform it—has since powered systems that beat quiz show champions and identify diseases more accurately than doctors. During his Gresham College presentation, Susskind identified the "AI fallacy" as the mistaken belief that machines must copy human methods to achieve human-level results, a misconception that has famously misled leading computer scientists and economists alike.
Economists were particularly unprepared for this shift because they had long classified tasks as either "routine" or "non-routine," assuming that tasks humans could not explain—such as making a medical diagnosis or driving a car—were safe from automation. Today, those boundaries have dissolved as algorithms use pattern recognition and massive datasets to perform these "non-routine" activities without any human-like understanding. This reality introduces the concept of "task encroachment," where AI gradually but relentlessly takes on more manual, cognitive, and affective tasks. While technology can complement human work by increasing demand for non-automated tasks, the pragmatist revolution has significantly bolstered the "substituting force" that displaces human workers. As the boundaries of machine capability continue to expand, the primary challenge for the coming decade is not necessarily mass unemployment, but the urgent need for individuals and governments to intervene as traditional work increasingly sits out of human reach. The future of work will be defined by machines that, as the philosopher Douglas Hofstadter eventually realized, can "fly without flapping their wings".
