course
description

With his book Homo Deus: A Brief History of Tomorrow, Yuval Noah Harari popularized the idea of Dataism as a new religion rising to dominate global consciousness. Data as holy scripture, kept and interpreted by data scientists, worshiped by some and condemned by others. Yet statistics and mathematics are not built on faith. We don’t need to ask ourselves if we believe in data science, but rather how much we care to understand it. What do we as citizens need to know in order to avoid being manipulated or subjugated? What power do we gain by understanding data science?

Data science has been used to build robotic limbs and to spread disinformation. To protect the environment and to place entire populations under surveillance. It is a tool that can be made to serve the morals and desires of whatever entity wields it. Yet we tend to think of numbers and mathematics as neutral and objective. Where does the transition between numbers and morals occur?

We will dissect several examples of data-driven activism and develop a conceptual understanding of the underlying data science methods.

We will observe the degree to which morality can be encoded in mathematical modelling. Don’t worry! Friendliness with math won’t be expected, but mathematical ideas will be discussed—with the goal of empowering participants and illustrating how a deeper understanding of models and visualizations unlocks new perspectives and possibilities.

Each week we will analyze social and environmental problems from a data perspective. We will reflect on what data is needed and what is available, and we’ll discuss the various ways participants can delve deeper into data science and get involved in data-driven activism. In the latter half of the course, each participant will discuss and develop their own idea for a data project with the rest of the group.

course outline

Each class will consist of two parts:

Part 1: Case study. We discuss the problem, a given use-case, and how to think about it from a data perspective. We study the relevant data, how it was sourced, and the methods applied.

Part 2: Afterward we shift to group discussion in a broader context. What other applications use similar methods? What are some ethical implications? What could be done differently? How might you create or take part in a similar project?

Week 1: A moment to get to know each other.

This session is about getting to know each other and learning about everyone’s background and expectations. What brought you to this course? What is it that you hope to learn and practice? We’ll also give a general overview of the course and share ideas, examples, and resources.

Week 2: California police scorecard and designing a metric​.

While protest against excessive police brutality in the United States has spread nationwide, policing outcomes in most places have not changed. Campaign Zero set out to create a metric for evaluating police departments and hold them accountable using criteria based on violence and racism.

Questions: ​What kind of data is available? How do you quantify and compare violence and racism between different law enforcement agencies? What would happen if this metric were widely adopted? What ethical decisions are made in designing metrics?

Other applications: ​GDP, equality/diversity ratings, university rankings... Keywords:​ police brutality, metrics/rankings, qualitative to quantitative, statististics

Week 3: Dazzle camouflage and computer vision.

Companies like Clearview AI are helping law enforcement use face recognition algorithms to

identify people from photos or video. In protest against this form of constant surveillance, activists are bedazzling their faces with makeup patterns designed to disrupt face recognition algorithms.

Questions:​ How are facial recognition algorithms designed? How can they be broken? What do we gain or risk by exploiting the weaknesses in such algorithms?

Other applications: ​mapping garbage, helping the blind, identifying fake images...

Keywords:​ computer vision, neural networks, privacy, surveillance, crime deterrence, law enforcement

Week 4: Social networks and substance abuse prevention.

Researchers at the USC Center for Artificial Intelligence in Society are mapping drug use and social networks amongst homeless youth and using the data to create optimal peer-based intervention groups that maximize positive and minimize negative influence.

Questions: ​How are positive and negative influences in social networks measured? How do you measure the effectiveness of an AI-based intervention? How is that data for social influence algorithms gathered? How can such algorithms be abused?

Other applications: ​HIV prevention by influence maximization, violence minimization...

Keywords: ​social networks, influence, social psychology, substance abuse prevention

Week 5: Presentations

Students present a chosen data-driven project. This could be an existing project or a planned one, which is presented and discussed through a similar lens as the one used for our case studies.

Where to go from here? ​A focused discussion on the insights gained during the course, in addition to getting involved and sharing resources suited to class participants’ goals.

who is this class for?

Anyone with a desire to explore data science concepts in the context of data activism rather than dry theory. People looking to understand the algorithms that govern much of modern society. Artists, journalists or activists searching for ways to use data to explore an issue or further a cause.

about live classes

Classes are 'live' meaning that you can directly interact with the instructor as well as with the other participants from around the world. Classes will also be recorded for playback in case you are unable to attend for any reason. For specific questions, please email us and we'll get back to you as soon as we can.

about tickets

Tickets for this class are currently available via Eventbrite. If you would like to avoid Eventbrite fees, please email us for direct payment options.

about solidarity

We realise we're living in uncertain times. We are a small organisation with no outside funding and like many, we are also in survival mode. During this time, we are offering a limited number of pay-what-you-can solidarity tickets for this online class. Preference is given to women, POC, LGBTQ+ and persons from underrepresented communities in tech who would otherwise be unable to attend.

We have added a generous supporter ticket for anyone interested in helping to subsidize the cost of our solidarity tickets. We are greatly appreciative of your support.

meet the instructor