The 80th anniversary of ENIAC marks more than just the birthday of the first general-purpose digital computer; it serves as a reminder that the foundations of modern software were not merely engineered, but “woven.” While the Electronic Numerical Integrator and Computer was commissioned by the U.S. Army for the rigid precision of ballistics trajectory tables, its true legacy lies in its evolution into a narrative engine—a machine capable of predicting the chaos of weather and introducing the fundamental logic of the subroutine.
The Shift from Calculation to Prediction
For co-inventor John Mauchly, the military’s requirement for artillery firing tables was a starting point, not the destination. Mauchly’s long-term ambition was meteorology. He had spent years collecting rainfall data across the United States, driven by a desire to find patterns in storm systems. To Mauchly, weather was a system unfolding through time, and a model of a storm was essentially a story about how that system might behave.
This ambition shifted the conceptual utility of the computer. Rather than treating the machine as a static calculator for discrete sums, Mauchly viewed it as a tool to “narrate the chaos.” This vision was realized in 1950, when ENIAC was used to produce the world’s first computer-assisted weather forecast, a feat made possible by upgrades to the machine’s digital program memory by Klara von Neumann and Nick Metropolis, and operational code written by Norma Gilbarg, Ellen-Kristine Eliassen, and Margaret Smagorinsky.
Technical Context: The Original Six
While the hardware was designed by J. Presper Eckert and John Mauchly, the initial programming was handled by six women: Kathleen “Kay” McNulty, Betty Holberton, Ruth Teitelbaum, Frances Spence, Marlyn Meltzer, and Jean Bartik. These programmers worked without manuals, relying instead on blueprints to route electrical signals through the machine.
Logic Weaving: The Role of Kay McNulty
Kathleen “Kay” McNulty’s path to the Moore School of Electrical Engineering was shaped by a heritage of weaving and a mastery of mathematics. Born in 1921 in Creeslough, County Donegal, Ireland, she emigrated to Philadelphia at age four. By 1942, after graduating from Chestnut Hill College, she was recruited by the U.S. Army to compute artillery tables by hand before being selected as one of ENIAC’s original programmers.
The act of programming ENIAC was an embodied process. Without a formal manual, McNulty and her colleagues learned the machine’s quirks through touch and memory, routing threads of electricity into patterns. This intimate understanding allowed the programmers to locate failed vacuum tubes more efficiently than the technicians.
This “weaver’s” approach to logic led to one of the most significant breakthroughs in computer science: the subroutine. Credited to Mauchly and McNulty, the subroutine—a sequence of instructions that can be repeatedly recalled to perform a specific task—was not part of the original blueprints or funding proposal. It emerged as a creative extension of the machine’s capabilities, transforming how software is structured to this day.
The linguistic connection to this process is found in McNulty’s first language, Irish. The word ríomh can mean to compute, but it similarly means to weave, to narrate, or to compose a poem. The Irish word for computer, ríomhaire, describes someone who weaves, computes, and tells a story simultaneously.
From Vacuum Tubes to Neural Networks
Looking at ENIAC’s architecture—a room of panels, switchboards, and wires—it resembles a textile production house more than a modern server farm. This physical reality mirrors a deeper truth about computing: the most powerful properties of complex systems often emerge through use rather than specification.
Modern large language models and autonomous systems are the spiritual successors to this “loom” philosophy. Like the first weather models or the first subroutines, these systems are not merely calculators; they are narrative engines. They take raw inputs and produce accounts of how the world might unfold. The most critical capabilities of today’s AI are often emergent properties, discovered by the people who learn how to “weave” with the model’s affordances.
The history of ENIAC suggests that the future of technology will not be found solely in the blueprints of the engineers, but in the imagination of the programmers who treat the machine as a medium for storytelling and prediction.
As we move further into the era of generative AI, does the metaphor of the “weaver” provide a more accurate framework for prompt engineering and model tuning than the metaphor of the “coder”?
