Just as ChatGPT has pushed AI from a hot topic in the tech industry to a global news phenomenon, so AI has become the defining technology of our age. This collection forms a unique record of the creation of Artificial Intelligence during the early years of computing in the form of 59 unpublished typescripts, offprints, photographs, ephemera and association copies. It shows the making of a Silicon Valley revolution as well as revealing many unknown aspects of this so-far little-known story with its surprisingly deep history, dating from the 1930s to the 1960s. ($124,000)
In this collection we can see how the invention of digital computing itself in the 1950s led to ‘logic machines’ which promised to imitate intelligence. Here we discover an entire program of ‘Digitalized Logic’, many unrecorded or otherwise unobtainable items including the program codes for a ‘list processing’ language developed in the 1960s called ‘WISP’. Another highlight is an extraordinarily rich letter on the ‘Turing Test’ by perhaps the most profound philosopher of the postwar era, David Lewis.
Pioneers like Claude Shannon, Warren McCulloch and Maurice Wilkes are well represented. So too are women like Thyllis Williams and Joyce Friedman, who worked at the cutting edge of this field in its early years. Even as the first ‘golden age’ of AI ended this collection shows how many of the central ideas in Deep Learning had already been developed - the statistical analysis of ‘neural nets’; language models that could draw on vast bodies of literature to make predictions; probabilistic computing; the idea of a machine that could pass for a person. Only now are we living with the consequences of these ideas.
In brief the collection comprises 60 items: 22 technical reports; 29 offprints/preprints; 2 association/proof copies; 7 pieces of ephemera over the period 1936–1969 with a linear measurement of 41cm.
The collection divides into three parts:
Part I: Logic Machine
The history of AI is as old as the history of digital computing. Exactly as old, in fact: in 1938 Claude Shannon laid the foundation for digital computing and the computational model of mind in a single paper. Within a few years his insight – that there is a connection between formal logic and electrical engineering – was being transformed by visionary neurophysiologist Warren McCulloch into a general theory of both mind and machine. The AI revolution was underway.
Part II: How to Build a Brain
The ‘golden age’ of AI: 1950s and 1960s; takes the idea of a ‘logic machine’ and turns it into a general method for developing machines that ‘think’ – but can they learn?
Part III: Deep Learning
The revolution in AI that we are living through draws on many of the insights of ‘Symbolic AI’, but deals more substantially with questions of complexity and, especially, Machine Learning. Claude Shannon as always was prophetic – this section traces the history of Neural Nets, probabilistic computing and investigations into the structure of Mind.
[Please Request a Full and Detailed listing of the Collection’s Contents]