Will Machinic Art Lay Beyond Our Ability to Understand It?

I’ll present a paper with Rui Penha at the Conference on Artificial Intelligence and Information.

In this communication, we will argue that artistic creations made by artificial minds will most likely lay beyond our ability to understand them, perhaps even beyond our ability to recognize them at all as artistic creations. We will assume that the emergence of consciousness in artificial minds is possible: the fact that the belief in that hypothesis is not consensual is irrelevant to the argument being proposed. Additionally, we will assume that the artistic creation we are referring to is one made by the artificial mind’s own volition and not the one that—either procedurally, generatively or interactively—is made based on (or directed towards) an anthropocentric view of art.

It is inevitable that we start by facing the problem of the definition of art. A definition that circumvents the ontological question will not be very useful in this context, since a definition such as e.g. the institutional theory of art allows any artefact or proposition to be considered as art solely by being recognised as such by someone acting on behalf of the artworld, not establishing, however, any criteria for such recognition. We will therefore start from the definition of art as embodied meaning, and from its relation with embodied cognition, to show that there is a binding of artistic creation to the subjective experience of existing in a natural and cultural world through a human body that is born with a foretold death.

We will furthermore look at the reasons that justify the difficulty to understand artistic manifestations that are not inscribed in our own cultural matrix or previous experiences. Finally, we will try to demonstrate that the best we can aim at, as human beings standing by an artistic creation by another species, is to an understanding of what could have motivated another human being to create such a work. As such, we shouldn’t be able to understand an artistic creation originating by an artificial mind with a physical experience of the world that differs from our own, even if they have a privileged access to our culture. The boundaries for this incomprehension, that we believe to be inevitable, are therefore those of the human mind.

Proceedings of xCoAx 2017


The proceedings of xCoAx 2017, the fifth conference on Computation, Communication, Aesthetics, and X, are now available at 2017.xcoax.org. Edited by Luísa Ribas, André Rangel, Mario Verdicchio & Miguel Carvalhais. Designed by Luísa Silva Gomes. ISBN 978-989-746-128-6. 389 pages.

On Procedural Dissemination and Artificial Aesthetics (notes Towards a Philosophy of Computational Media)

At Digicom, I’ll present a paper written with Pedro Cardoso.

Procedural media possess traits that stand them apart from classical media, and lead them to foster a very peculiar set of relationships with their readers. Chief among these is how their procedural layer is discovered through a process of virtuosic interpretation, and how this allows for empathy and for the transference of their core mechanics from the computational system to the reader’s mind. This paper focuses on how this process is developed, and how these media conduce to the dissemination of a type of aesthetic experience that is unique to them, but remarkably similar to cognitive processes that humans develop towards each other.

On Emptiness (or, On Finishing with a Blank Canvas)

A paper presented at the xCoAx 2017 conference, “On Emptiness (or, On Finishing with a Blank Canvas)”:

Starting from a review of monochromatic art, minimalism, and conceptualism, this paper attempts to understand the significance of these approaches to processor-based arts, looking at a dominant strategy in these works, the deployment of destructive processes. The paper then studies a series of recent works by Austrian software artist Lia, that are closer to the roots of monochromatic painting, suprematism, and constructivism, and are based in a deeply procedural and conceptual approach that we describe as a constructive process.

Patterns for Serendipity in Interaction Design

At xCoAx 2017, a paper by Ricardo Melo and myself, “Patterns for Serendipity in Interaction Design”:

With today’s filtering and personalisation of digital content, there is a growing need for systems that actively promote novel interactions and that allow the user to discover new, unsought information. As such, this paper starts by addressing the need for these serendipitous systems and how one can design for serendipity considering its unpredictable nature. We then propose a series of user patterns that define the mental model that is more conducive to serendipitous experiences, derived from our revision of the literature as well as our observations. Finally, and through an analysis of the state of the art, we propose a tentative series of design patterns at both the implementation and interaction level, which constitute a framework for the design of interactive systems that afford the experience of serendipity.

Interaction under Interference

At xCoAx 2017, a paper by Luís Eustáquio and myself, “Interaction under Interference”:

With today’s filtering and personalisation of digital content, there is a growing need for systems that actively promote novel interactions and that allow the user to discover new, unsought information. As such, this paper starts by addressing the need for these serendipitous systems and how one can design for serendipity considering its unpredictable nature. We then propose a series of user patterns that define the mental model that is more conducive to serendipitous experiences, derived from our revision of the literature as well as our observations. Finally, and through an analysis of the state of the art, we propose a tentative series of design patterns at both the implementation and interaction level, which constitute a framework for the design of interactive systems that afford the experience of serendipity.

Creation of Meaning in Processor-based Artefacts

At the ISEA2017 proceedings, a paper with Pedro Cardoso:

Processor-based artefacts are often created following conventions inherited from analogue media forms, allowing the development of experiences that, in spite of the new platforms, are not fundamentally different from those that were already possible in the previous contexts. But contemporary media and arts often use processor-based artefacts focusing on conceptual and mechanical principles that do not attempt to simulate earlier forms but rather explore their computational nature. These systems bring about new modes of reading and new challenges, to both readers and artists or designers. In order to optimize the usage of processor- based media, creators need to understand how these artefacts are interpreted and how readers develop processes of creation of meaning in procedural contexts. This will allow authors to ground their practices on procedurality rather than only on surface con- tents, and to make a constructive use of contingent behaviour, learning, adaptation, selection, and other traits of these systems, not being limited to the emulation of well-established media forms. This paper outlines some of these challenges and proposes designing for the meaningful interpretation of computational artefacts.

Valuably Unsought: Systems for Digital Serendipity

At the ISEA2017 proceedings, a paper by Ricardo Melo and myself:

Contemporary interaction with media is mediated through a plethora of digital systems, conditioning said interaction to the experiences that these systems anticipate and limiting the potential of the medium for surprise and serendipity. Through a literature-review and system analysis, we assert the value of serendipity in our digital interactions, arguing the necessity of a distinction between Natural and Artificial Serendipity, while establishing key areas of action of serendipitous systems: Information Encountering, Experience, Collaboration, Creativity and X. We identify specific systems within each of these key areas, as well as their methods and mechanics for achieving Artificial Serendipity in the Digital Medium.

Artificial Aesthetics reviewed by Neural

Neural issue 55 / Autumn 2016 has published a review of Artificial Aesthetics.

Over the years, Miguel Carvalhais has developed a consistent practice producing music compositions, graphic design, live audiovisual performances and sound art installations. Furthermore, this practice is accompanied by academic work teaching and publishing. Artificial Aesthetics manifests all of the aforementioned work in a single, coherent articulation around the fulcrum of his research: the aesthetic qualities of programmable media. The autonomous quality of programmable media emerges as a key factor in Artificial Aesthetics and the author goes through an attentive historical analysis of the processual systems for its creation, from the recombination of elements to various degrees of artificial intelligence. Here Carvalhais is developing an analytical model for the analysis of procedural systems. This is arguably a new methodology, but one with the concrete potential to become a standard. One of the most valuable aspects of this book is how every single chapter includes relevant content emerging across specific visual, audio and textual artworks, all of which are drawn together in a coherent presentation. The “artificiality” of processes is constantly referred to when the author qualifies elements with or without acknowledgeable human qualities. As Carvalhais brilliantly defines it, we are still in the historical phase of “computational incunabula”, already with a substantial history behind us. This book can potentially become a classic, but it already looks it could become an essential reference for digital art and culture in the future.