Beyond Boundaries: A Feminist Manifesto for the Post-Anthropocene Epoch

Workshop Description:  This hands-on workshop provides participants with a two-fold approach to feminist manifesto-making, one that engages body/minds, and one that  engages AI as an art form, allowing a critique of the dangers and possibilities of unassisted and technology assisted intelligences.  Manifestos created in the workshop  will be published and exhibited in both digital and physical formats at the esteemed academic conference, EVA London 2023, and its associated art exhibition.

Introduction

Manifestos, historically potent mechanisms for endorsing transformation and articulating novel conceptions, constitute the foundation of our workshop. This interactive milieu invites attendees to navigate an unprecedented landscape wherein artificial intelligence and human acumen are inextricably intertwined. Each participant is to engage in a collaborative process with our unique artificial intelligence system, "AI Philosopher (AIP)," producing a few lines of manifesto with related writings that encapsulate their viewpoint on a salient feminist issue, particularly as it pertains to the Anthropocene epoch and the subsequent era that unfolds.

This innovative amalgamation of AI and human input ushers us towards a new realm of discourse that defies the conventional parameters of authorship and capitalises on the distinctive potential intrinsic to human-AI collaboration. Through this exploratory endeavour, we aspire to scrutinise the dynamic potentialities of feminism within a post-anthropocentric epoch.

The culmination of our workshop will be embodied in a collective volume, a compendium of the AI-enhanced manifestos and personal writings of each participant. This tangible evidence of our collective journey is to be published and exhibited in both digital and physical formats at the esteemed academic conference, EVA London 2023, and its associated art exhibition. This offering will stand as an enduring tribute to our expedition into envisioning a feminist future beyond the Anthropocene.

Objectives of the seminar

Students will learn about and practise the détournement techniques that Guy Debord and Gil J. Wolman, two French political theorists, invented in this workshop. The aim is to challenge and transcend the existing boundaries of commonly accepted signifiers and signifieds concerning objects, events, and others. Drawing inspiration from their seminal publication, A User's Guide to Détournement, students will be tasked with crafting a new dataset for AIP and establishing a fresh set of image-semantic labels. This endeavour is aimed at rerouting and estranging our conventional understanding of feminism and the conditions of humans and post-humans.

During the collaboration with AIP, students' primary focus should be on the logical construction of the content rather than the technical aspects of programming implementation. Essential aspects to consider include the selection of the target corpus snippets, the interrelationship among these snippets, as well as the choice and arrangement of the target images and their corresponding semantic labels. A critical aspect of this process involves understanding the implications and connections between the chosen signifiers and signifieds. It is worth noting that students will also be required to independently author their respective manifestos; while inspiration may be drawn from GPT4 or related large language models, the integration of AI-generated content into their written chapters is not permitted. This same principle applies to human intelligence's image-semantic labels of signifiers and signifieds.

For the purposes of this workshop, students will be assigned to one of four groups, each tasked with exploring a distinct topic. The collective efforts of these groups will result in the compilation of a comprehensive book. Students lacking suitable electronic devices, such as personal computers, laptops, or tablets, are encouraged to send their formatted content to explore@ivas.studio. Submissions may be in the form of a *.txt file (for the AI-human writing part), plain text, or a shared Google Doc or Microsoft Word file (hyperlink) within the body of the email. The same conditions apply for submissions of human intelligence's image-semantic labels of signifiers and signifieds. Please refer to the instructions detailed below for additional guidance.

Seminar Outcome: Book Structure and Exhibition

The resultant book of this workshop will be bifurcated into two significant segments: "Feminist Manifesto in the Anthropocene" and "Feminist Manifesto in the Post-Anthropocene." Unique in its construction, the book can be engaged from two distinct directions. The front section, which typically opens from right to left with the spine to the left, will only contain content that human intellect has created and orchestrated. Conversely, the rear part, unconventional with the spine to the right, opens from left to right and showcases content birthed through the collaboration between human minds and artificial intelligence.

It is paramount for students to focus solely on the production of content, with visual and typographic elements being the remit of the IVAS artists. The final embodiment of this collective endeavour will be a tangible book alongside its digital counterpart. This dual-format publication ensures accessibility and longevity, embodying the interplay between humans and artificial intelligence in the pursuit of a feminist future beyond the Anthropocene.

Operation Guide

The successful completion of this workshop involves careful adherence to the following set of instructions and operational tasks.

Text content

• Human intelligence production:

Each student will first engage in individual research and thought collection on their assigned feminist issue within the context of the Anthropocene or the Post-Anthropocene. Following this, they will independently craft a 5 to 7 line manifesto encapsulating their thoughts, perspectives, and arguments on the matter.

• Human-AI collaboration:

In this phase, students will collaborate with our AI system, AIP, to design a new dataset, incorporating the selection of the target corpus snippets and their interrelationships. Here, the focus should be on a meaningful coalescence of human intellect and AI in the creation of the manifesto content while ensuring there is no integration of AI-generated content into the human-written chapters.

Guidance

1. Selection of content materials

Students can choose the context readings from the seminar's recommended reading list or look for the ideal content in other publications like papers, articles, books, poetic works, etc.

1. Pre-trained dataset preparation

Students need to combine those text materials into a single *.txt file. Each selected text material should be adopted as a paragraph with 5 to 7 sentences as a standard unit and employ <|endoftext|> as a separator to divide the different selected materials. There is no strict requirement for the size of the pre-trained dataset. Students can make the decision based on their expectations for the final generated text. It is worth noting that the bigger the dataset, the better and more precise the performance of the AI-generated text.

1. Dataset training

Students are encouraged to send their formatted content to @ivas.studio. Submissions may be in the form of a *.txt file (for the AI-human writing part), plain text, or a shared Google Doc or Microsoft Word file (hyperlink) within the body of the email. Students with an interest in dipping their toes into the process of AI training can also interact with the Google Colab file that IVAS prepares for them. It's a simple operation to complete this process; students just need to click the "RUN" button or the "RUN ALL" button on the prepared Google Colab file to execute this task. However, students should be more focused on the logic of the content than the programming techniques.

Students should send out the pre-trained dataset to @ivas.studio by the end of July 6th.

1. Get AI-generated content

Once the AI training is completed, there are a few ways to interact with and get the AI-generated content: Send a list of questions or trigger sentences to @ivas.studio; send a request to @ivas.studio to get random AI-generated content; access the prepared Google Colab file through a browser to interact with AIP in real-time.

Students can get multiple AI-generated pieces of content and choose the ones they want to include in the book.

Image-context

• Human intelligence production:

For the image-semantic labels of signifiers and signifieds, students will need to select and organise relevant images that align with the target topics and their respective manifestos. They will also create related semantic labels to form a coherent image-context, thus aiding in the creation of a compelling and détourned visual narrative corresponding to their written manifestos.

Students will employ the artistic approach of collage on a shared Miro board  or physical paper to create sets of image-semantic labels of signifiers and signifieds between images. Students with no proper electronic devices, such as PCs, laptops, and tablets, can employ ready-made materials to draw the denotation and connotation maps on paper, take a picture, and then send it to explore@ivas.studio.

• Human-AI collaboration:

In collaboration with the AI system, students will embark on redefining and reinterpreting the semantic labels of their chosen images. The goal is to detour and estrange our conventional understanding of the images and their meanings, aligning this process with the détournement techniques referenced earlier. This human-AI collaboration will culminate in an avant-garde reinterpretation of the image context, which complements and enriches the written content.

By the end of this week, further introduction of the operation will be complete.