This thesis describes my artistic practice as essentially project-based, site-responsive and hypermediating. Hypermediacy—i.e. the tendency of certain media or objects to keep their various constituents separate from their structure—is to be understood as opaque, juxtaposed and after a recurring contiguity with different kinds of interfaces. Accordingly, and within the context of the various projects that constitute this thesis, it is demonstrated how, in response to the particular places I work and to the various people I collaborate with, different kinds of materials and methodologies are incorporated in broader hybrids that are mediated (interfaced) in miscellaneous ways to this way result in original works of art. Materials and methodologies are shown to be intertwined and interdependent with each other as well as with the different ways in which they are interfaced, which accounts for an explicitly projectbased, rather than artwork-based, approach which, on its turn, de-emphasises the finished artefact in favour of process, performance, research and exploration. Projects are, then, shown to be explicitly site- or situation- responsive, as they are not implementations of preexistent ideas, but rather emerge as my original response to the particular sites, materials, people and the various other constituents that are involved in their very production. Interfaces to such hybrids as well as their very material and methodological elements are also shown to be hyper-mediated. It is finally argued that such an approach essentially accelerates multi-perspectivalism in that a project may spawn a number of diverse, typically medium-specific and/or site-specific, artworks that all exemplify different qualities which are congenital to the particular nature of each project.

Current approaches to generating rhyming English poetry with a neural network involve constraining output to enforce the condition of rhyme. We investigate whether this approach is necessary, or if recurrent neural networks can learn rhyme patterns on their own. We compile a new dataset of amateur poetry which allows rhyme to be learned without external constraints because of the dataset’s size and high frequency of rhymes. We then evaluate models trained on the new dataset using a novel framework that automatically measures the system’s knowledge of poetic form and generalizability. We find that our trained model is able to generalize the pattern of rhyme, generate rhymes unseen in the training data, and also that the learned word embeddings for rhyming sets of words are linearly separable. Our model generates a couplet which rhymes 68.15% of the time; this is the first time that a recurrent neural network has been shown to generate rhyming poetry a high percentage of the time. Additionally, we show that crowd-source workers can only distinguish between our generated couplets and couplets from our dataset 63.3% of the time, indicating that our model generates poetry with coherency, semantic meaning, and fluency comparable to couplets written by humans. / Graduate

Le projet artistique Poésie de l’ADN prend source dans une démarche interdisciplinaire, en associant la programmation, la biotechnologie, la création sonore et visuelle, l’ésotérisme ainsi que l’art participatif et performatif. Ce mémoire passe en revue les différentes étapes de réflexion et création autour de ce projet, tout d’abord en le replaçant dans son contexte historique et artistique, à la fois au niveau du courant bioart, puis plus précisément dans l’utilisation de l’acide désoxyribonucléique1 (ADN) en arts. Nous verrons comment s’inscrit Poésie de l’ADN aux côtés des oeuvres d’art génétiques visuelles et sonores, ainsi que de portraiture, tout en présentant les influences artistiques et esthétiques qui inspirent mon travail. Ensuite, ce mémoire aborde trois questionnements qui sont au coeur du développement conceptuel de ce projet : 1) les tensions entre l’aléatoire et la subjectivité, c’est-à-dire la constante négociation entre le contrôle de l’artiste et l’autonomie de la machine; 2) l’interprétation des données afin d’établir la signification que l’on reçoit et apporte à la matière; 3) ainsi que les considérations bioéthiques que soulève ce projet. Puis, le mémoire détaille le processus technique et les outils utilisés, ainsi que le processus créatif au sein duquel l’approche ésotérique est fondamentale. Enfin, nous verrons l’impact des critères ADN sur les oeuvres, ainsi que les portraits qui en découlent. Poésie de l’ADN est un projet artistique qui réunit à la fois le développement d’une application dont l’algorithme et les paramétrages permettent de générer des matières audio-visuelles à partir de données ADN, ainsi que la composition de portraits, à travers l’interprétation artistique, ésotérique et humaine de cette matière, pour créer des vidéomusiques individuelles. / DNA Poetry is an artistic project rooted in an interdisciplinary approach, combining programming, biotechnology, sound and visual creation, esotericism as well as participatory and performative art. This thesis reviews the different stages of reflection and creation around this project, first of all by placing it in its historical and artistic context, first off within the bioart artistic movement, then more precisely within the use of deoxyribonucleic acid (DNA) in the arts. We'll see how DNA Poetry fits alongside visual and audio genetic artworks, as well as portraiture, while showcasing the artistic and aesthetic influences that inspire my work. Then, this thesis addresses three questions which are at the heart of the conceptual development of this project: 1) the tensions between algorithm and subjectivity, that is to say the constant negotiation between the artist's control and autonomy of the machine; 2) data interpretation, in order to establish meaning in what is received and brought to the material; 3) and finally the bioethical considerations raised by this project. Then, the thesis details the technical process and the tools used, as well as the creative process in which the esoteric approach is fundamental. Finally, we will see the impact of DNA data on the audio and visual outcome, as well as the portraits that result from them. DNA Poetry is an artistic project that brings together both the development of an application whose algorithm and mappings make it possible to generate audiovisual material from DNA data, as well as the composition of portraits, through the human, esoteric and artistic interpretation of this material, in order to create individual videomusic art pieces.