Social History of Artificial Intelligence

* Books: The Eye of the Master: A Social History of Artificial Intelligence. by Matteo Pasquinelli. Verso, 2023.

URL = https://www.versobooks.com/en-gb/products/735-the-eye-of-the-master

Description

"A “social” history of AI that finally reveals its roots in the spatial computation of industrial factories and the surveillance of collective behaviour.

What is AI? A dominant view describes it as the quest "to solve intelligence" - a solution supposedly to be found in the secret logic of the mind or in the deep physiology of the brain, such as in its complex neural networks. The Eye of the Master argues, to the contrary, that the inner code of AI is shaped not by the imitation of biological intelligence, but the intelligence of labour and social relations, as it is found in Babbage's "calculating engines" of the industrial age as well as in the recent algorithms for image recognition and surveillance.

The idea that AI may one day become autonomous (or "sentient", as someone thought of Google's LaMDA) is pure fantasy. Computer algorithms have always imitated the form of social relations and the organisation of labour in their own inner structure and their purpose remains blind automation. The Eye of the Master urges a new literacy on AI for scientists, journalists and new generations of activists, who should recognise that the "mystery" of AI is just the automation of labour at the highest degree, not intelligence per se."

Review

Simon Schaffer:

In this original and extremely timely book, Matteo Pasquinelli offers nothing less than a long-range history and critical analysis of a labour theory of automation and knowledge. He uses detailed studies both of the remarkable accounts of general intellect and the extractive and exploitative organisation of the industrial workplace produced in nineteenth-century British political economy and of the challenging developments of models of machine intelligence and computational systems developed in the mid-twentieth century United States to unlock the sources and meanings of the politics of artificial intelligence. The work shows how Marx's depiction of the development of the social individual under industrial capitalism provides indispensable resources for making sense now of what artificial intelligence means, and the forms of economic and political order that its embodiment of knowledge and control express. At a moment when apostles and prophets of machine intelligence proclaim both a utopian world of effortless control and a catastrophe of extinction, Pasquinelli's patient and clever work provides a crucial insight into the past and future of AI monopolies and their consequences."

(https://www.versobooks.com/en-gb/products/735-the-eye-of-the-master)

2. Rob Lucas:

"Matteo Pasquinelli’s recent book, The Eye of the Master: A Social History of Artificial Intelligence, is probably the most sophisticated attempt so far to construct a critical-theoretical response to these developments. Its title is somewhat inaccurate: there is not much social history here – not in the conventional sense. Indeed, as was the case with Joy Lisi Rankin’s A People’s History of Computing in the United States (2018), it would be hard to construct such a history for a technical realm that has long been largely tucked away in rarefied academic and research environments. The social enters here by way of a theoretical reinterpretation of capitalist history centred on Babbage’s and Marx’s analyses of the labour process, which identifies even in nineteenth century mechanization and division of labour a sort of estrangement of the human intellect. This then lays the basis for an account of the early history of connectionist AI. The ‘eye’ of the title links the automation of pattern recognition to the history of the supervision of work.

If barely a history, the book is structured around a few striking scholarly discoveries that merit serious attention."

..

His material provides the theoretical backdrop to a scholarly exploration of the origins of connectionist approaches to machine learning, first in the neuroscience and theories of self-organization of cybernetic thinkers like Warren McCulloch, Walter Pitts and Ross Ashby that formed in the midst of the Second World War and in the immediate post-war, and then in the late-50s emergence, at the Cornell Aeronautical Laboratory, of Frank Rosenblatt’s ‘perceptron’ – the earliest direct ancestor of contemporary machine learning models. Among the intellectual resources feeding into the development of the perceptron were a controversy between the cyberneticians and Gestalt psychologists on the question of Gestalt perception or pattern recognition; Hayek’s connectionist theory of mind – which he had begun to develop in a little-reported stint as a lab assistant to neuropathologist Constantin Monakow, and which paralleled his economic beliefs; and vectorization methods that had emerged from statistics and psychometrics, with their deep historical links to the eugenics movement. The latter connection has striking resonances in the context of much-publicized concerns over racial and other biases in contemporary AI.

Pasquinelli’s unusual strength here lies in combining a capacity to elaborate the detail of technical and intellectual developments in the early history of AI with an aspiration towards the construction of a broader social theory. Less well-developed is his attempt to tie the perceptron and all that has followed from it to the division of labour, via an emphasis on the automation not of intelligence in general, but of perception – linking this to the work of supervising production. But he may still have a point at the most abstract level, in attempting to ground the alienated intelligence that is currently bulldozing its way through digital media, education systems, healthcare and so on, in a deeper history of the machinic expropriation of an intellectuality that was previously embedded in labour processes from which head-work was an inextricable aspect.

(https://newleftreview.org/sidecar/posts/unlearning-machines?pc=1575)