Neural Networks | Leonardo/ISASTwith Arizona State University

Neural Networks

Neural Networks

by Ranjodh Singh Dhaliwal, Théo Lepage-Richer, and
Lucy Suchman
University of Minnesota Press, Minneapolis, Minnesota, 2024
110 pp. Paper, $18.00
ISBN: 978-1-5179-16695.

Reviewed by: 
Robert Maddox-Harle
July 2024

This book was rather disappointing. The title Neural Networks does not accurately portray what the book is actually about. It would have been more helpful with a title such as, “Neural Networks: A Cultural, Historical, and Biological Appraisal”. The book is mainly concerned with the cultural and historical aspects of neural networks, especially from a media theory perspective. This is interesting and obliquely relevant to current issues with neural networks, but it is not really relevant to the current global surge of, and controversies concerning Deep Learning and Artificial Intelligence, with their underpinning structure of neural networks. The last chapter by Lucy Suchman saves the book in this regard.

My second criticism is that the graphical presentation of the book is ‘very ordinary’. Actually, it is more a booklet. At a mere 110 pages, it could have benefited from at least two more chapters to get right into the “heart” of the current (2024) neural network technical phenomenon. The type-font size is ridiculously small. Not all of us have perfect eyesight. This makes the book difficult reading. There is one black & white illustration!

Having said this, it should be stated that the book is well written, with excellent scholarship by the three authors, all are highly qualified in their respective fields. They show how the nature of neural networks has been “shaped and conditioned as much by cultural values as by biological/technical investigation.” It might be useful, then to think of neural networks not as being created, discovered, found, generated, or even studied. Rather, it may be fruitful to understand neural networks as being rendered. Notions of rendering offer an alternative to ideas about re-presentation” (p.13).

Attempts to study and find evidence for racial difference in neuronal cell biology to justify the atrocities by various country’s colonialisms is exposed. Further, the distorted analysis of these neuron cells by wartime psychiatry (early 19th century) is further discussed. “While they are currently known as a biologically inspired, statistical approach to machine intelligence, neural networks were first introduced as a neuroanatomical approach and later a psychiatric model, both of which actively enforced historical conceptions of racial and pathological difference” (p.14).

The book(let) has an Introduction – Rendering the Neural Network, followed by three chapters:

(1) Neural Media

(2) On Parascientific Mediations: Science Fictions, Educational Platforms, and Other Substrates That Think Neural Networks

(3) The Neural Network at its Limits

Each of the chapters contains “eye-opening” information not commonly known. This is fascinating though not entirely surprising because as most critical thinking discourse shows our so-called scientific facts and technological advances are always mediated by cultural, religious, and political influences.

The question of intelligence is obviously closely connected with brain and mind concepts. This however is also looked at critically in this book, as we see in the following quote: “When considered in relation to neural network’s trajectory across histology, psychiatry, and computer science, what neural media show is how the history of artificial intelligence is essentially that of mediating who or what is recognized as the bearer of intelligence” (p. 47).

Suchman’s chapter moves the discussion away from the media, historical approach to the “real issues” facing us in the “here and now”. She looks at the two approaches to computational neural networks, the cognitivist and the symbolic, outlining the approach of two giants in the computer networks field - Geoffrey Hinton and Gillian Einstein. Very briefly Hinton supports the “number crunching” approach whereas Einstein agues this approach will never imitate human intelligence because our intelligence is not just “weighted numbers” firing in lightning rapidity across our brains. Rather, it is a product of our whole body. The retina, brain stem, and heart are not technically part of the brain but are most certainly part of our minds. I have discussed this problem myself in two peer-reviewed published papers, (see reference below).

“Einstein conceptualizes the brain in terms of neural “circuitry”, but for her the connections not only are profoundly and complexly embodied but also change with experiences not reducible to weighted inputs.” (p. 101) [my emphasis].

So, the debate goes on regarding the nature of intelligence, the brain, the mind and just where the current artificial intelligence explosion is leading us.


"Disembodied Consciousness & The Transcendence of the Limitations of the Biological Body". Harle, R.F. 2007, Janus Head Journal. Special Issue - "Situated Body" (Winter 2006/07) Trivium Publications, New York.

"Cyborgs: Uploading & Immortality. Some Serious Concerns." Harle, R. F. 2001. Sophia International. vol. 41 no.2 November 2002. Ashgate Publishers.