course
description

As deepfake technology ripples ubiquitously through our newsfeed, we increasingly turn a skeptical eye to the media we are consuming online; from the physicality and likeness of celebrities, down to the voices which emanate from them, to the simulated crowds that surround them.

Pairing ‘fake everything’ with cinema becomes especially interesting as we consider that it is traditionally the place we go to purposely for engaging our suspension of disbelief.

Where do we draw the lines of the acceptable and unacceptable when it comes to what can and should be produced in this fake new world? Does the context really change everything?

In this experimental class we want to put machine learning technology into the hands of artists, filmmakers, and other creators in order to reclaim and reappropriate it as a tool for producing playful, thought-provoking, socio-political film clips and more. What is the potential for machine learning assisted filmmaking to become its own art form, providing an unexpected new language to poetically provocate the world around us?

In this program, we will start with filmmaking as a visual art form and political tool in its own

right. What aspects of a story make a film political? Is the personal always political? These discussions set the foundation and will carry on throughout our month together as we hear from filmmakers and technologists with their own unique perspectives and bodies of work.

Additionally, throughout the month you will learn a range of techniques involving generating and manipulating images of human bodies, including creating imaginary people and deepfakes, compositing them into other images or video, as well as making them talk with audio speech generation. In parallel to technical teaching, we will be hosting a guest lecture series, inviting filmmakers from around the world to share their experiences of using cinema as a political tool.

The goal is to provide the inspiration and tools for participants to begin telling their own stories, using machine learning assisted filmmaking to build connections and create communities in the face of rising injustices within the confines or our own fake “real” world. How can the empathy of make-believe splinter out of the machine and the screen and into social action? Join us this summer and let’s find out together.

in this course,
you’ll be introduced to…

• Creative and critical perspectives on filmmaking and machine learning

• Various tools and techniques for capturing footage to manipulate

• Creating lookbooks to craft a story

• Theory, tools and frameworks for using machine learning with moving images

• How to split, reassemble and organise videos and images using python

• How to generate images and voices using deepfakes, style transfer and image synthesis

• How to use face detection, segmentation and classification models to process images and extract information from video data.

• How to composite, edit and render videos using Blender

• The differences / value of using off the shelf tools such as RunwayML vs. creating your own from scratch

• An opportunity for introspection and self-reflection with others

• Professional development for artists and creatives

• Exciting humans working in the field

• Critical and conceptual development of projects to be exhibited in a final group showcase

• A variety of guest speakers with different perspectives and relevant practices

• An amazing network and community of like-minded creative beings and potential future collaborators

course outline

Week 1:
We will begin by getting to know each other more deeply, sharing who we are and what we care about as an act of generosity towards the community we are forming together over these four weeks. Guest speakers will introduce us to key themes and references, kickstarting the class with fresh perspectives on modern filmmaking, politics, and the potential and future of AI driven stories which will set the foundation for our work. Experiments with footage and content creation.

Week 2:
This week will feature a hands-on introduction to programming in Python. Participants will learn how to construct, deconstruct and process batches of images and videos. With this knowledge we will start to explore specific techniques for manipulating videos and images, including, creating imaginary people, deep fakes and style transfer.

The lecture series continues, supported by discursive sessions around the ethics of

deepfake technology, prompting a collective imagining of the technoscientific futures we want to be a part of.

We will continue discussions which began in week one about the craft of storytelling for film, beginning to construct narratives that will lead towards a final project. 

Week 3:
In week 3, we embark on collaborative work and begin capturing our stories. Additionally, participants will be introduced to using Blender as a tool to composite generated assets with other images or video. We will also explore making our images talk with audio speech generation.

Week 4:
This week sees the final development of your personal and collaborative projects, with the assistance of artists, peers, and mentors, culminating in a public screening at programs end to celebrate this month-long experimental journey and works produced during this time.

who is this program for?

This program invites artists, designers, creatives, technologists, filmmakers, writers, researchers, journalists, activists, and anyone with a curiosity to think critically about the ever-evolving relationship between filmmaking and machine learning to connect and collaborate with a group of like-minded creators. No prior experience needed.

meet the instructors