DXARTS 481 A: Data-Driven Art I

Lynn Hershman Leeson, The Electronic Diaries of Lynn Hershman Leeson 1984–2019, 1984–2019, six-channel video installation, Courtesy: the artist; photograph: Alwin Lay.

DXARTS 481 A: Data-Driven Art I

Instructor: Laura Luna Castillo

Meeting Time: MW 12:30pm - 3:20pm

Location: RAI 205 and 132

Code: 801108

Credits: 5.0

Contact information: lcasti2@uw.edu / Raitt Hall 207/129

Office hours: TBA or by appointment

TA: Sadaf Sadri

Office hours: by appointment

Course overview

This is a theoretical-practical course exploring art made using information, algorithms, patterns, datasets, searches, and metadata. Media - images, video, sound, text - as data indexable, searchable, and part of larger systems. Implications and possibilities of artists using such systems, looking at dynamic, algorithmic based approaches to composing with highly distributed collections of data. 

This class combines technical instruction in Python with discussion of the ethical aesthetic and creative possibilities of data science. The format of the class is a mix of lectures and discussions overviewing concepts, theories and examples of data-driven artworks and processes. It also includes hands-on programming exercises, demos and archival research assignments. Students will work in small groups, pairs, or solo for in-class programming activities. 

No prior coding experience is necessary. Non-STEM and STEM students welcome. 

Learning objectives:

By the end of this course students will have an understanding of the technical and historical contexts for data-driven art whilst becoming familiar with contemporary data-driven art practices, and being able to critically engage with art projects made with data science tools. In combination with the technical skills required to work with data science tools, students will be well-positioned to make work that is in dialog with emerging contemporary art practices.

Skills and outcomes include:

• Collect, curate and creatively engage with archives and datasets.
• Discuss and evaluate ethical, social and hegemonic data-driven practices, explore datasets and its inherent biases. 
• Programmatically access, navigate and transform a variety of textual, visual and sonic datasets.
• Critically engage and understand the mechanisms behind generative AI and Machine Learning models.
• Construct a methodology that combines the artistic and scientific method as a hybrid practice, with quantitative and qualitative research methods applied to interdisciplinary data-driven projects.
• Engage with a deeper conceptual understanding on how algorithms and data-driven processes shape human experience and the way we tell stories with data.
• Develop methodologies to interface various media and data in a variety of interactive, generative and poetic processes.

Course structure and assignments:

• Most sessions consist of a 40 to 60 minute session at the beginning of each class, which can be a discussion on readings, case-studies relating to the topics from that week, invited artist talks, and lectures. The rest of the class will continue through hands-on practical experiments and demos. 

• The course is divided into 3 modules of 5 to 6 lectures/discussions plus practical demos, each relating to specific concepts and data processes. Each module culminates in a creative assignment. The modules are divided as follows:

1. 1. 1. Data-driven Arts: A Techno-conceptual Lexicon - An introduction to key concepts, theories, philosophies, vocabulary and artworks that illustrate the use of patterns, code and software as tools that transform archives, influence our everyday and hold memories, experiences and ideologies.
      2. Oracles and translators: encoders/decoders - Machine Learning, Artificial Intelligence and a variety of technological interfaces as poetic agents and conduits.  

1. 1. 1. Politics, Poetics and cybernetics - Exploring complex political, societal, ethical and creative aspects developing across an increasingly data-driven, everyday - AI landscape.

• The final assignment consists on creating a substantial creative and artistic project combining textual, visual and sonic archives with the data processes and concepts discussed throughout the course. The rubrics for this final assignment are based on the three modules, where each has specific conceptual prompts. The rubric should also be used as a from of inspiration and a guide to develop a critically complex creative project.

Grading structure:

• Each module is 20% of the grade, divided as follows:
  • • Archival assignments: 5%
    • Readings, technical experiments, diagrams: 5%
    • Module's final assignment: 10%

• The final project is 30%, consisting of two parts:
  • • Final project proposal: 5%
    • Final project development and presentation: 25%

• In-class participation is 10%, below are the kinds of participation we love:
  • • Being an active participant throughout in-class discussions and during critiques.
    • Help and collaborate with classmates, embrace the interdisciplinary and seek out connections with each other. 
    • Asking questions, asking for help with the code/technology, actively and consistently discussing your ideas and your creative process, this can be in class or approaching the instructor and/or TA during office hours or via email.
    • Sharing tools, artworks, books, movies, methodologies, code... anything that is exciting or relevant to the concepts we are exploring! 
    • Telling us about topics or tools you might be interested in testing and developing a project with. 
    • Fill-in awkward silences during discussions (this is extremely and utterly appreciated!)

Important notes, resources and software:

Lectures will take place between two classrooms, Raitt 205 for most lectures and discussions and Raitt 132 for labs and demos (also referred to as the computer lab), it is important to keep an eye on canvas announcements and emails.

Most sessions will begin in 205 and then move downstairs to 132, unless directly specified in the calendar or via Canvas announcement a few days prior.

• In this course we will be using ANACONDA,  which is an open-source distribution of the Python and R programming languages that's used for data science, machine learning, and artificial intelligence. This software will allow us to experiment with certain Machine Learning algorithms and interface them with real-time, creative applications. Students should install Anaconda Software in their computers prior to the second week of the quarter. It can be downloaded here (for both Mac and Windows): Download it here
• We will also be using Geany, which is a lightweight text editor that allows us to work and test our own Python code. Download it from here
• In this course, we will also be using TOUCHDESIGNER, a visual programming software based in Python, which allows us to create and activate all kinds of data in the form of interactive 2D and 3D applications in real time.
  • Download the software here: https://derivative.ca/download   
  • If you would like to know more about this software please visit: https://learn.derivative.ca/courses/100-fundamentals/ 
• We will interface some archives and datasets with other technologies through the deployment of a website, the platform we will use is called Glitch.com, a free, open web platform. It is advised to make your own account as this allows to save and test a variety of tools, apps and implementations. You can access it here.

Machine Learning and AI labs:

Most of these algorithms and models require to be run in a computer with GPU, we will be training small models from scratch and fine-tuning models which, albeit small, still require substantial computing power and memory. The DXARTS computing lab, Raitt 129, has 10 iMacs with the M3 chips which can run these algorithms. If the need arises, we also have 5 Alienware Desktop computers which can be set-up ona case by case basis. If you have your own computer and you know it has the specs, please support us by using it during our labs to free up a lab computer for those who might not have the right specs in their personal machines. If you are unsure, please reach out to the instructor or TA. 

DXARTs server: We will be using the server to train models with larger datasets and for longer periods of time, we will go over this processes together during one of our labs. 

Using Google Colab, ChatGPT, other Generative AI tools that are out there and commercially available: 

As we will see throughout the quarter, the majority of these services collect user's data one way or another, and make use of it at different stages and for a variety of applications. This becomes particularly relevant if students are working and developing personal archives or working with archives which are not part of the Creative Commons, or those for which they do not own any rights, for example content from UW libraries or datasets and archives from vulnerable and/or historically oppressed communities. 

Currently, the only UW-supported generative AI tool is Microsoft Copilot. The UW agreement with Microsoft provides added protections, including enhanced protection of user data and privacy, for all UW users. 

Google Colab is not appropriate for confidential or sensitive data, and therefore our course will not require the use of Colab. Yet, we acknowledge that a variety of interesting tools are best accessed and run through the Colab service, for this reason it is up to students to choose to use it as an alternative for course work and as a tool to explore course concepts and creative ideas for your projects. For the most part, this course tries to provide non-colab options. Terms of Use are standard Google terms, not available under UW’s negotiated contract. 

Note regarding Multiple Submissions:

Students are not allowed to submit one project for credit in two different classes (see UW Policy on Academic Misconduct), without prior discussion with and permission from BOTH class professors. Students under these conditions will have to develop a substantial project that is deemed to be complex enough to earn credits for both classes.

Religious accommodation  

Washington state law requires that UW develop a policy for accommodation of student absences or significant hardship due to reasons of faith or conscience, or for organized religious activities. The UW’s policy, including more information about how to request an accommodation, is available at Religious Accommodations Policy (https://registrar.washington.edu/staffandfaculty/religious-accommodations-policy/)Links to an external site.. Accommodations must be requested within the first two weeks of this course using the Religious Accommodations Request form (https://registrar.washington.edu/students/religious-accommodations-request/)Links to an external site..

DEI+A statement

DXARTS strives to create a safe, affirming, and welcoming space for all cultures, races, nationalities, sexes, gender identities, ages, religions, and economic statuses. We are committed to ensuring that the University of Washington’s DEI+A values are reflected through our diverse community of students, faculty and staff, our research and public-facing events, our course curriculum, our actions, and interactions within the department. 

For more departmental and campus-wide resources, please visit the DXARTS DEI+A page.Links to an external site.

COVID-related policies

As per UW policy, masks/face coverings are recommended inside campus buildings - see UW policy hereLinks to an external site.. As this class is conducted in-person, students are expected to participate in class to fully benefit from course activities and meet the course’s learning objectives. Students should only register for this class if they are able to attend in-person. To protect their fellow students, faculty, and staff, students who feel ill or exhibit possible COVID symptoms should not come to class. When absent, it is the responsibility of the student to inform the instructor in advance (or as close to the class period as possible in the case of an unexpected absence), and to request appropriate make-up work as per policies established in the syllabus. What make-up work is possible, or how assignments or course grading might be modified to accommodate missed work, is the prerogative of the instructor. For chronic absences, the instructor may negotiate an incomplete grade after the 8 th week, or recommend the student contact their academic adviser to consider a hardship withdrawal (known as a Registrar Drop).