From the estate of computer pioneer C. Gordon Bell (1934-2024), considered the “father of the minicomputer.” He was an early employee of Digital Computer Corporation (DEC) where he was the architect of the PDP-4 and PDP-6 and a major contributor to the PDP-1, PDP-5, and PDP-11. He was in charge of the successful 32-bit VAX computer (the rival computer to Data General’s Eagle Project featured in Tracy Kidder’s Pulitzer Prize-winning book The Soul of a New Machine)𓆏. Bell went on to found Encore Computer and la﷽ter Ardent Computer and spent the final years of his long career at Microsoft in a research capacity. Bell is also notable as a co-founder of The Computer Museum in Boston and was a founding board member of its successor, the Computer History Museum in Mountain View, CA. 

Cohen, Harold. “What is an Image?” From: Proceedings of IJCAI-6, Tokyo: 1979.

Cohen, Harold. “How to Draw Three People in a Botanical Garden.” From: Association for the Advancement of Artificial Intelligence Proceedings, St.Paul, 1988.

[ARTIFICIAL INTELLIGENCE & ART.]

EARLY AI / CYBERNETIC ART MILESTONE.

Harold Cohen / AARON.

Harold Cohen (1928-2016) graduated from the University of London’s Slade School of Art in 1950. He became a part of the London art scene and exhibited his paintings internationally although he began to look to other methods of expression by the late 1960s, eventually taking a position at the University of California, San Diego where he developed his interest in computer technology. It was there that he met fellow professor, and future originator of the A🐟pple Macintosh project, Jef Raski🥂n who introduced Cohen to the CDC 3200 mainframe computer as well as programming in FORTRAN.

Cohen began to devise methods to approach computing and art but became frustrated with the batch processing required on a mainframe and eventually acquired a Data General Nova minicomꦑputer and later DEC PDP-11s and VAXs so that he could operate the computer 🌱directly. He also transitioned to programming in C.

His early computer painting system attracted some interest and he was invited as a Visiting Scholar to Stanford’s Artificial Intelligence Lab (SAIL), home of AI discipline founder John McCarthy, where he spent two years on a paint system. Cohen’s initia๊l program was built upon a small defined set of rules and forms the computer translated to drawings on paper using a small, two-wheeled robot equipped with a marker that Cohen called a “turtle.” The process bears a similarity to cybernetic entities like W. Grey Walter’s robot tortoises Elmer and Elsie, Claude Shannon’s robotic mouse Theseus, or even W. Ross Ashby’s Homeostat. Cohen had encountered the field of cybernetics in the early 1960s and his 1963 Robert Fraser Gallery exhibition statement references cybernetic principles: “[Painting] is self-analytical, self-critical and possessed of this finely balanced feedback system, self-controlling.”

Cohen was quick to🃏 point out from an earlyꦍ stage that “AARON is not a transformational device. There is no input, no data, upon which transformations could be done: in fact it has no data at all which it does not generate for itself in making its drawings. It is a complete and functionally independent entity, capable of generating autonomously an endless succession of different drawings” (Cohen, p3, 1979).

Cohen, in his 1988 paper “How to Draw Three People in a Botanical Garden,” defined AARON as “a program designed to investigate the cognitive principles underlying visual representation.” He went on to clarify that the early, pre-1980 versions of AARON “dealt exclusively with internal aspects of human cognition. It was intended to identify the functional primitives and differentiations used in the building of mental images and, consequently, in the making of drawings and paintings. The program was able to differentiate, for example, between figure and ground and inside and outside, and to function in terms of similarity, division and repetition. Without any object-specific knowledge of the external world, AARON constituted a severely limited model of human cognition, yet the few primitives it embodied proved to be remarkaℱbly powerful in generating highly evocative images: images, that is, that suggested, without describing, an external world” (Cohen p848).

The initial work of Cohen’s paint system consisted of primitive line drawings which Cohen would color by hand. AARON’s evolution was similar to that of a child, but with Cohen as its sole influence. Scribbles led to drawings that resembled figures and then began to look more and more recognizable. “From the program’s inception around 1973, I had been convinced that AARON wou🙈ld need to be built upon a convincing simulation of freehand drawing, and gave much attention to modeling the feedback-dependent nature of human drawing behavior.” Cohen deliberately programmed AARON to simulate the drawing behavior of young children at🃏 the post-scribbling stage in order to increase the range and diversity of its output.

Cohen debuted a more compl🎃ex version of AARON programmed in LISP in 1995. It could now also color the forms that it drew. Cohen continued to develop his program into his 80s and exhibited his work🥂 in major museums throughout the world.

The present work was created during an exhibition at The Computer Museum in Boston and demonstrates AARON at an important stage in its development - the figures were beginning to take recognizable form as human representations with faces and bodies among terrain. Cohen had considered 1984, the year of the present work, as a major transitional year for AARON for this reason. “By 1984 the earlier ‘rock-art’ pictorial paradigm had given🐎 way entirely. The pressure to provide real-world knowledge of what AARON’s ne꧙w visual space contained became inescapable and the first of several knowledge-based versions of the program was constructed” (Cohen p 850).