chapter 4 (Can Computers Be Creative?’: A Misguided Question

Style transfer

The dominant mode of AI art production where AI learns and replicates the visual style of existing artists. The author criticises this as mere mimicry rather than genuine creative engagement.

Crowdsourced beauty

The author's critical term for the aesthetic standard used to judge most AI art. Refers to beauty derived from aggregating what already exists and what people already recognise — things that look symmetrical, mesmerising, or familiar — rather than genuinely new expression.

Mimesis

Aristotle's concept of art as creative imitation of nature. Crucially, mimesis proceeds by addition and remediation — it transforms what it imitates and adds something new. It was a form of genuine creative engagement, even before humanist notions of originality existed.

Mimicry

The author's term for what style transfer actually does. A "belaboured resemblance" that pretends to be something it is not. Unlike mimesis, mimicry adds nothing new — it simply reproduces the surface features of existing works as convincingly as possible.

Remediation

A concept from Bolter and Grusin referring to how new media forms transform and absorb older ones. Used by the author to explain what made mimesis genuinely creative — it remediated nature rather than simply copying it.

The Next Rembrandt

A 2016 Microsoft-led project where a deep learning algorithm analysed over 300 Rembrandt scans to identify his most characteristic features, generating a new image in his style that was then 3D-printed with ink simulating oil paint. Used by the author as a prime example of superficial AI imitation art.

"Can it pass?" question

The public fascination with whether experts can tell the difference between AI-generated art and work by real masters. The author sees this as a populist guessing game rather than genuine artistic engagement, though it does raise interesting questions about authorship and expertise.

Joseph Rabie

Artist and critic who argued that The Next Rembrandt is not art but "the empirical science of perception being modelled and applied at a high level." Used by the author to support the critique of AI imitation work.

Vilém Flusser

Philosopher of technology whose concept of "programmed freedom" is central to the chapter. Argued that humans and their apparatuses merge into a unity, enabling new kinds of collaborative creativity that challenge humanist notions of individual genius.

Apparatus

Flusser's term for modern technological systems — not old-style tools like hammers or paintbrushes, but complex machines, software, and wider infrastructures that enact both symbolic and material transformations.

Programmed freedom

Flusser's concept describing the relationship between humans and their apparatuses. Humans are neither fully free nor fully determined — they make creative selections from the range of options the machine's algorithm allows. The camera's programme shapes the photographer's intention even as the photographer directs the camera.

Human-apparatus unity

Flusser's idea that in modern society, humans and their machines merge into a new kind of collaborative unity. Neither the human nor the apparatus is simply in control — they operate together, enabling new forms of creative action.

Post-humanist view of art

The author's proposed alternative framework for understanding creativity. All artworks — from cave paintings to AI experiments — are understood as produced by humans in assembly with nonhuman agents

Assembly

The post-humanist idea that artworks are never produced by humans alone but always in combination with a plethora of nonhuman agents. Challenges the humanist notion of the lone creative genius.

Humanist notion of creativity

The traditional view that places the individual human — with their unique genius, originality, and free imagination — at the centre of the creative process. The author challenges this by arguing human creativity has always been entangled with nonhuman systems.

"Can computers be creative?" — a misguided question

The author's central claim. The question presupposes a sharp division between human and machine creativity that does not actually exist on the post-humanist view. Rather than asking whether computers can be creative, we should reconsider what creativity and the human itself actually are.

Biological programme

The author's extension of Flusser's argument to the biological level — human life itself operates according to a programme (DNA sequences). Used to argue that the entanglement of humans with nonhuman systems is not unique to the post-industrial era but foundational to what it means to be human.

Post-humanist art history

The new paradigm for understanding art that the author calls for. Would see the human as part of a machine, dispositive or technical system — not its inventor, owner, and ruler — and would recognise all artworks as produced in assembly with nonhuman agents across history.

Contrast with Goetze

Where Goetze maintains a sharp distinction between human creative labour and AI processing to argue AI training is theft, this chapter blurs that boundary — arguing human creativity has always been partly computational and entangled with nonhuman systems. However Goetze's argument about consent and economic harm does not necessarily depend on this distinction alone.

"I know what I like"

A phrase the author uses to characterise the shallow aesthetic standard behind crowdsourced beauty — art judged purely on whether it appeals to existing tastes rather than whether it offers genuine new expression or intervenes meaningfully in the world.