The Artist in the Machine

The Artist in the Machine
by Arthur I. Miller

The MIT Press, Cambridge, MA, 2019
432 pp. Trade, $29.95
ISBN: 978-0-262-04285-7.

Reviewed by
Robert Maddox-Harle
February 2020

This is a very well researched and highly readable book. The theme is the investigation of the creative potential of machines to produce art, literature, and music.

I was interested to read in Miller's concluding chapter that; "At present, AI is more like a child that needs to be painstakingly taught before it can produce the desired image or sound, let alone create of its own accord" (p. 304). I wrote a short introductory paper some 20 years ago, discussing the possibility of creating a conscious machine stating precisely this! [1].

The Artist in the Machine is an excellent introduction, and primer, to the whole field of AI, machine learning, and machine creativity. Readers new to this discipline will be very well informed after reading this book, and AI experts will be sure to come across some hitherto unknown facts to perhaps inspire their future research. The computer research and recent advancements in AI described by Miller most certainly are part of the "Brave New World" scenario. He methodically explores this idea without the usual hysterical, dystopian/utopian/Silicon Valley hype associated with the future of AI. Although at times the book is a little like a Who's Who of the glitterati of Google, Silicon Valley, and the most prestigious universities!

Arthur I. Miller is Emeritus Professor at University College, London, UK. His website www.artistinthemachine.net brings to life the URLs cited and enables us to listen to the music and view the images discussed throughout the book. After his Introduction, Miller's next four chapters explore what constitutes creativity, outlining seven hallmarks of creativity and two of genius. These chapters are essential, for if we are to understand machine creativity we first need to understand human creativity.

The following 40 chapters (some are rather short) discuss AI from its birth, all the way through to 2019. These chapters cover topics such as computers with feelings and emotions, the nature and possibility of conscious computers with many practical examples from the "coal face", so to speak, of computers making original, unique art, writing poems and stories, and even producing theatrical shows such as Staged by Android Lloyd Webber and Friends (p. 249).

Miller increasingly sees a new breed of artist, "…who is computer scientist rolled together with artist, musician, or writer. "The question of whether computers can be creative is no longer a matter purely of speculation, philosophy, metaphysics, or morals. We now have hard evidence” (p. 261). I believe it might be more accurate to state that since the advent of the powerful, personal computer and unlimited access to the ubiquitous Internet, an 'extra' breed of creatives is evolving alongside traditional artists and writers. All expressions of artistic creativity are still valid, which until the advent of AI were the sole province of humans. If you doubt this, please see my paper, The Biobehavioural Basis of Art [2].

I also think that rather than ask can machines be creative and produce art we should ask as Artistic Director of Ars Electronica in Linz, Gerfried Stocker, does. "Rather than asking whether machines can be creative and produce art, the question should be, "Can we appreciate art we know has been made by a machine?" (p. 263). That is, the question of the possibility of machine creativity is somewhat passé! The real question, which Miller discusses in chapter 38, A Glimpse of the Future concerns computer awareness and intentionality, of which Searle and Dennet have much to say elsewhere.

As previously mentioned this book is very well researched; however, I am astounded to find that Jon McCormack is not even mentioned, let alone his seminal work discussed, (which I have also reviewed for Leonardo) Computers and Creativity. [3] Further, another minor criticism is that Harold Cohen’s work, especially with Aaron for decades, gets one paragraph? Cohen was a pioneer in machine artistic creativity and had much to add, from a different slant, to his work in the last few years of his life! This ‘different slant’ is highly relevant to future AI artistic machine interaction. [4]

These minor criticisms aside this book will become a standard reference resource for students and scholars in the fields of machine learning and creativity and also be a fascinating read for the lay person interested the our global computer/robotics co-habitation society.

[1] "The Possibility of Creating a Conscious Entity: A Humanities Perspective." Harle. R.F. 2014 Sept. Modern Research Studies: International Journal of Humanities and Social Sciences vol. 1, Issue # 2, Argartala, India. (First version unpublished 1999)

[2] "Biobehavioural Basis of Art." Harle, R.F. (Revised version 2008) Technoetic Arts: A Journal of Speculative Research. University of Plymouth, UK.  vol. 6 no.3, 2008  http://www.intellectbooks.co.uk/

[3] Computers and Creativity. Edited by Jon McCormack and Mark d'Inverno. Springer-Verlag, Berlin Heidelberg, 2012. 430 pp., illus. Trade, ebook. ISBN: 978-3-642-31726-2; ISBN: 978-3-642-31727-9

[4] “Harold Cohen (1928-2016) The Last Six Years of a Creative Life,” by Louise Sundararajan. http://www.setumag.com/2019/08/harold-cohen-1928-2016-last-six-years.html.