Debates on AI-generated music often treat recent systems as an unprecedented break, reviving anxieties about authorship, labour, and originality. This article offers a different perspective. Contemporary generative tools are understood here as the latest configuration in a long history of redistributing musical work across humans, devices, and formal procedures. Rather than presenting a comprehensive timeline, this article develops a genealogy through selected scenes, including musical automata, combinatorial paper algorithms, early rule-based computer composition, and machine learning. These cases are used to track how musical intelligence and creative agency have been repeatedly relocated, moving from mechanical performance to rule-following and from symbolic representations to statistical learning. This article argues that the distinctive shift introduced by prompt-driven systems lies less in claims of autonomous machine creativity than in the interface as a site of composition. Music-making becomes an iterative practice of specification, selection, and revision shaped by negotiation and curation under platform conditions. By treating AI music as a socio-technical reorganization of practice, this article contributes to the history of computing by using music as a case through which to examine changing relations among formalization, representation, interface, and human-machine collaboration.
This article examines the emergence of computer medicine in West Germany through the DOMINIG II project (1973–1980), the country’s largest-ever medical networking initiative, by examining the imaginaries, expectations, and automation anxieties through which computerization was rendered necessary within West German medicine. Drawing on the archived DOMINIG II application report and debates in the Deutsches Ärzteblatt (1968–1995), it traces how a concept of “Problembewusstsein” (“problem consciousness”) was developed to render (automated) computational systems the only rational response to perceived deficiencies in hospital communication. Against this backdrop, the article also reconstructs a counter-discourse within the medical profession at the time, centered on the metaphor of the “factory-like hospital,” which evolved from fears of physician replacement in the late 1960s to broader anxieties about an ethical erosion of medicine by the late 1970s. Situating the case within transnational entanglements, the article contributes to the broader historiography of medical computing and the history of technological change in human-centered workplaces.
This article examines Paysage, a computer-aided design (CAD) application developed by Atelier Kroll in the early 1980s to advance participatory architecture. While histories of architectural computing often focus on its earliest experiments in the 1970s or the proliferation of computing in the 1990s, the rise of personal computing in the 1980s remains overlooked. The development and use Paysage reveal how Kroll translated his call for an “industrial architecture without industrialists” to the digital realm, and used it to reconcile prefabrication with diversity, participation and adaptability. By foregrounding this overlooked experiment, the article argues for expanding architectural historiography to include custom software environments, demonstrating how digital tools embodied ideological positions and reshaped design practices during the formative years of personal computing. At the same time, it shows how software histories reveal professional networks and aesthetic regimes that would otherwise remain invisible to architectural history.
This article examines whether New Zealand’s geographic isolation, small domestic market, and post-colonial economic structure shaped the adoption of electronic computing between 1960 and 1975, and if this produced outcomes different from those in other countries. Drawing on archival sources, contemporary reports, and a compiled dataset of computer installations, we analyze patterns of growth, institutional development, and examples of local ingenuity. Import controls and limits on foreign exchange strongly influenced early adoption, encouraging shared use through service bureaus and slowing overall growth. We compared New Zealand with several other developed countries to establish whether there were any grounds for claims of exceptionalism during the early spread of modern computing. Our findings show that New Zealand’s experience largely followed that of other countries and, rather than being exceptional, tended to lag behind comparable economies. By placing New Zealand alongside these countries, this study contributes to a clearer comparative understanding of early computing history.
The DesCartes Program is a Franco-Singaporean effort advancing hybrid forms of artificial intelligence for critical urban infrastructure. Through ethnographic research, this paper examines the project’s wind map — a digital twin modeling air flows of Singapore’s Marina Bay — within the city-state’s territorial and digital ambitions. The wind map has been imagined as an instrument for urban governance, knowledge-making, and value creation. However, its development has been marked by frictions surrounding data access, securitization, and the sovereignty of urban information. These dynamics both complicate and illuminate Singapore’s aspirations for controllable smart city governance, suggesting that while the DesCartes Program has offered novel forms of urban legibility and management, it may also reproduce entrenched logics of territorialization, security, economics, politics, and technocracy.
Beginning in the late 1960s numerous mixed-use high-rise buildings such as Chicago’s John Hancock Center were constructed in US cities, employing novel externalized structural systems to create uninterrupted rentable spaces that could support a variety of interior configurations over time. Concurrently, large US architecture firms including Skidmore, Owings, & Merrill (SOM) and Caudill, Rowlett, Scott (CRS) began efforts to distill the design of tall office buildings into a series of related design problems that could drive computer simulations of a building’s performance as an investment. More than the individual computer programs or buildings, the most significant impact of these efforts was the impetus they provided for a transformation of design processes at large architectural firms, as well as the built environment generally, according to terms that might be amenable to computation. Drawing on primary sources from the period, this article outlines the foundations of this transformation in a network of collaborators and competitors, the legacies of this transformation in contemporary approaches to building information modeling, and how the practice of architecture and the urban fabric of the US have been shaped in the process.
In 1959, Cornell University undergraduate Thomas Osborn worked with a small team on the assembly and training of the Mark I Perceptron at Cornell Aeronautical Laboratory. This account records his experience as a “worker bee” on the first hardware implementation of a machine learning program.
Quantum computing (QC) is often framed as a technological race defined by qubit counts, hardware breakthroughs, and error-correction milestones. Between roughly 2014 and 2025, however, a distinct quantum software layer began to emerge alongside advances in physical devices. Drawing on indicators such as investment patterns, patent activity, the institutionalization of university programs, and emerging measurement standardization efforts, this essay suggests that the mid-2020s may represent a formative period in the consolidation of quantum software as a recognizable field. The question is not simply whether it is premature to speak of “quantum software” but what historians can observe when a discipline begins to take shape ahead of its anticipated underlying hardware.
Drawing upon accounts of early coding exploits and press coverage of the 1960s computer dating fad, this article argues that the history of learning to code in the United States is one of casual sexism, exemplified by stories of young men attempting to use coding and access to computers to improve their social lives or, in some cases, as entrepreneurial opportunities for computer matchmaking businesses. The history of young men attempting to computationally improve the odds of meeting young women they might date is also part of the history of social media, as seen in Mark Zuckerberg’s Facemash project and Ben Mezrich’s The Accidental Billionaires, a non-fiction account of Facebook’s founding. Rather than accept the media image of coders simply as socially inept computer nerds, I show how computers and coding, particularly at elite US universities, became conflated with sociability and romantic desire.
Jean-Raymond Abrial is one of the central figures in the development of formal methods for software and systems engineering. Over a career spanning more than five decades, he has played a decisive role in the creation of the Z specification notation, the B-Method, and Event-B and in demonstrating their applicability to large-scale industrial systems. This article presents a scholarly biographical account of Abrial’s life and work, tracing the evolution of his ideas from early work on real-time languages and databases, through foundational contributions to formal specification, refinement, and proof, to the development of industrial-strength tool support such as the Atelier B and the Rodin platform. The article situates Abrial’s contributions within their historical, intellectual, and industrial contexts and assesses their lasting impact on software engineering and formal reasoning about programs.
The National Academy of Engineering classified widespread electrification as the greatest engineering accomplishment of the 20th century. Innovation in computing technologies proved essential to this achievement. Additionally, as engineers adapted computer hardware and software for grid control, they contributed to advances in computation. The traditional story of electrification follows steady expansion from small central station power systems to continental-scale interconnected grids. Yet, transitions triggered by blackouts, evolving energy resource mixes and patterns of use, regulatory restructuring, and technological innovation punctuated this history. System operators faced a consistent challenge—keeping the lights on—as transitions threatened to undermine grid stability and efficiency. Today we are in the midst of another, perhaps more profound, and certainly more rapid transition. Additionally, computation itself poses both challenges and solutions for grid stability. This paper analyzes how 20th-century engineers and system operators developed the algorithms that provide optimization and control of the electric power grid.
This essay sketches out what I call “weird AI”: probabilistic, associative systems that behave as if they feel. It shows how midcentury architects of artificial neural networks deliberately courted the uncanny as they engineered space for intuition, emotion, and nonrational thought, even as standard histories cast computing as an Enlightenment project of calculation and control. To make sense of artificial intelligence’s past and present, historians must move beyond an information-centric framework and reckon with the affective undercurrents that have shaped the field from its start.
The Center for Research in Management Science (CRMS) at the University of California, Berkeley, pioneered the use of interactive, computerized, multiperson simulation experiments for social science research. Conceived by Prof. Austin Hoggatt, the center's Management Science Laboratory included a computer system with linked DEC PDP-5 and PDP-8 computers. An advanced APL-based version was designed and built during 1972-1976. In January 1974 it could support one experimenter developing and running an experiment connected to multiple terminals. By 1976 the system was complete, with an easy-to-use, interactive, APL subsystem and a multiuser, time-shared, operating system. The project was a technical success, making it possible to write experiments, debug them, run them, and analyze the results. It established the Management Science Laboratory as the prototype for computerized modeling in economics and behavioral science. However, the lab failed to jell as a community of experimenters and support personnel, and without continued institutional support the hardware and software quickly became obsolete.
