AI & Society | Spring 2020 | Syllabus

Professor Molly Wright Steenson | Carnegie Mellon University School of Design

It’s not a class about penguin societies. But it makes you wonder what a society of AI penguins might be like. Nice picture, in any case. Photo by Jesse Chan/Unsplash

How does AI impact society?

How do artificial intelligence and algorithmic technologies impact society? In this class, we will will look at how AI and its related paradigms affect the way that we live. In this seven-week class, we’ll be reading, writing, watching, sharing, creating, debating, talking with experts, and more. We’ll explore a number of questions, such as: How do we understand intelligence? How does it impact labor and how we work? How is AI biased and how might it become more fair? Do we trust it? And what ethical considerations should we be taking?

The class will culminate by holding a fictitious future conference on AI & society that takes place in 2027, seven years from now, with the class curating and developing posters, panels, and keynotes. 2027 is far enough to think about future effects, but close enough that things may not be radically different. (Or will they?) We’ll finish the class by putting on a capsule version of that conference our final night.

Learning objectives

• Develop your critical acumen and your angles on issues related to AI, algorithms, and technology, and how they impact people, the public, and society
• Familiarize yourself with recent publications on critical technology questions
• Analyze the effects of technology in your weekly reflections and in class discussions
• Situate the work in your own interests by bringing in outside material that relates to the class topics
• Create and produce your own perspective on AI & society and the themes in class by contributing to the final conference evening

What you’ll be doing in this class

This class is set up to reward you if you give it an effort to engage critically and creatively. Here’s what you’ll be doing.

Readings/responses/signals: 30% (5 points per post)

Each week, there are readings for you to complete. They are listed here in the schedule, and are either on the Google Drive (CMU login required) for the class or are linked in the schedule. You will be expected to read them (although you’ll want to do this strategically: this piece on how to move through readings is helpful). You’ll have a reading response due each week on Canvas that has two parts to it:

• A response to a prompt or that I will provide each week. Your response should be approx. 250 words in length.
• A “signal” that that you will post with a paragraph of explanation that you write, approx. 100 words. A signal is a news item, research paper, photograph, video or other content that represents a direction of change or emerging trend. Each signal should contain a pointer to the document (a URL link, or APA-style citation) and tell us why it’s significant. This is a way to tie the class discussion back to your own interests, whatever they might be. It’s an opportunity to be creative.

These are due each week on Monday at 5 pm. These will be marked on completed/not completed; each one is worth 5 points. If completed according to directions, you’ll get the full points. If your post is late, it’ll go down a point per day. (Careful: this will add up.)

Guiding discussion: 10% (once, with a group of other students)

You, along with a group of students, will be responsible for helping to guide the discussion and bringing your signals to life in the class for one of the sessions. For the week that you and your colleagues are discussion leaders, you will set the key themes of discussion (and will want to email with each other beforehand to determine what they are). You will also share your signals that week and lead discussion in the class sharing their signals. If we have a guest, you might end up introducing the speaker, interviewing them, or helping the class to organize our conversation.

The “Conference” final project: 50%

The class will plan and hold a conference that takes place in 2027, and you will play a role and make a contribution to it. This might include a poster presentation, being a panelist or debater, being a speaker, putting together a Medium page that announces it, or something else. We will have time in the class to work on this project, and we will begin developing the conference on 2/5 (more information to follow), with a proposal due for your part by 2/10. Your creativity and criticality will matter here. In our final class, we will hold the conference with outside participants.

Participation: 10%

Be in class, be on time, be a part of small group and large group discussions, speak up and help your classmates to participate. Know when to step up and step back (please do speak in class, whether large and small group, and if you’re someone who tends to speak a lot, know when it’s good to step back and bring in others).

The schedule

Readings on Google Drive (CMU login required) or linked below.

1/15: What do we know? and the Useless AI Workshop

• Class intro, syllabus walkthrough
• Useless AI workshop & demos: (Thanks to Phil van Allen and his Art Center College of Design colleagues for originating this idea). We’re going to put our worst ideas about AI front and center by designing useless AI—something that seems like it could be sensible, but is useless, silly, or bad.

1/22: Intelligence & AI

We will explore what AI is so that we’re on the same page, and what intelligence is. How do you know when you see it?

Guest: Professor Wendy Ju, Cornell Tech, New York. Prof. Ju’s research is on interaction with autonomous vehicles & robots. She brought us a few projects to consider about our interactions with robots, such as a mechanical ottoman and a young trash can on the move.

Readings:

• Meredith Broussard, pp. 30–39, (Ch. 3, “Hello AI” ). [Drive]
• Janelle Shane, ch. 3, “Does It Actually Learn?” You Are a Thing and I Love You, 2019. Janelle Shane writes the hilarious AI Weirdness blog, where she documents how she trains neural nets to do funny things—and in the process, she shows how AI works. This chapter is her description of how machine learning works. The reading is illustrated, really funny, and useful.[Drive]
• Alan Turing, “Computing Machinery and Intelligence, Mind, Vol. 59, №236. (Oct., 1950), pp. 433–460. This is the piece that introduces the idea of the “imitation game” or “Turing Test:” convincing a human that a machine is a human. This article is about the question of whether a machine can think. This piece is not as easy going as the other two (it was published 70 years ago and the language is more formal)—but we will be working through his argument in class, especially the way that he imagines objections to the question of whether machines can think. [Drive]

And in class, we’ll spend a few minutes watching a snippet of this famous Star Trek Next Generation episode, “The Measure of a Man.”

1/29: People, Labor & Work

AI is about people, and AI affects the work that people do. It requires human labor. In this session, we’ll look at the effect of AI on labor in a few different areas of focus (trucking, content moderation, Mechanical Turk and other piecework—and what your signals bring to the discussion.)

Readings:

• Cathy O’Neill: ch. 6, “Ineligible to serve: Getting a Job,” Weapons of Math Destruction, 2016 [Drive]
• Lilly Irani & M. Six Silberman, “Stories We Tell About Labor: Turkopticon and the Trouble with ‘Design’” [Drive] —this will be a critical look at questions of design and the impact it has on labor. (You’ll see these threads come together at the end of the piece.)

You will read one of the following (to be assigned in class). Discussion leader group will be distributed among these pieces.

• Mary Gray & Siddharth Suri, “Algorithmic Cruelty & the Hidden Costs of Ghost Work,” Ghost Work: How to Stop Silicon Valley from Building a New Global Underclass, 2019 [Drive] (Pardon the formatting of the PDF, it’s not terrific.)
• Sarah Roberts, “Social Media’s Silent Filter,” The Atlantic, March 8, 2017.

• Karen Levy, “The Contexts of Control: Information, Power, and Truck Driving Work,” The Information Society 31 (2015): 160–174.

Here’s the video of Dr. Levy’s 2017 lightning talk at AI Now. She refers to the “integration of the human into the machine” in some ways that are compelling, scary, and real—it’s worth a watch even if this isn’t what you’re choosing to read. (~5 minutes)

2/5: Fairness and Bias

Algorithms and by extension, AI, often reinforce negative biases—to negative effects. In this class, we’ll explore the ways that this happens and possibilities for change. We will also start working on the Conference in this class.

• Ruha Benjamin, Race after Technology, excerpts: (Ch. 1) pp. 49–76 , Excerpts from Ch. 2 (pp. 90–96) & Ch. 5 (pp. 174–180) [Drive]
• Multiple authors, “Roles for Computing in Social Change” FAT* ’20, January 27–30, 2020, Barcelona, Spain.[Drive]

2/12: Trust

(By 2/10 at 5 pm, you will also submit a one-page proposal of your project for the Conference part of the class, due on Canvas.)

Many say that we’re in a time where our trust is deteriorating. It’s especially apparent in big tech and in politics, and often, AI and algorithms are to blame. We’ll talk about the trust issue in this class. Do we trust? Can we trust? And we’ll look turn our attention to the question of trust in medicine, with a possible guest who suggests that trust might just have a way forward.

• Ethan Zuckerman, “The Economics of Mistrust,” McSweeneys, p. 95–109 (I’ve provided the whole publication, but only Zuckerman’s piece is assigned) — [Drive]
• Multiple authors, “‘The Human Body is a Black Box‘”’: Supporting Clinical Decision-Making with Deep Learning,” FAT* FAT* ’20, January 27–30, 2020, Barcelona, Spain. [Drive]

2/19: Ethics

In the last two years, there’s been an upsurge about ethics and AI. There’s also critique about these approaches. How do we approach technology ethically, while not viewing it as a checklist or way to dodge regulation? We’ll have Professor David Danks (Head of the Department of Philosophy at CMU) and Carol Smith (CMU SEI) as guests in this class.

• danah boyd & MC Elish, “Don’t believe every AI you see,” The Ethical Machine, Shorenstein Center on Media, Politics, and Public Policy
  at Harvard Kennedy School, November 13, 2018
• Jacob Metcalf, Emmanuel Moss, danah boyd, Owning Ethics: Corporate Logics, Silicon Valley, and the Institutionalization of Ethics,” social research (86:2), Summer 2019
• Rodrigo Ochigame, “The Invention of Ethical AI:How Big Tech Manipulates Academia to Avoid Regulation,” The Intercept, https://theintercept.com/2019/12/20/mit-ethical-ai-artificial-intelligence, December 20, 2019

2/26: 2027: The Conference!

It’s 2027, and you’ve curated a conference on AI & society. We will hold a micro version of the conference in this class—more information to come.

Attendance policy

Absences of any kind are strongly discouraged as your learning and work will be adversely affected by the information and activities you miss. We only have 7 meetings and that’s it. If you’re out, you will miss significant content.

Be punctual, arriving just before the class start time so we can begin sessions promptly, and stay for the duration of each class. Come to class prepared. But if you’ve been up all night and you’re stressed and you didn’t complete the readings, please don’t cut the class: it’s better for you to be here. Bring yourself.

A few attendance policy items from the School of Design:

• If you are five minutes late or leave class early you will be marked as absent. Three absences may cause your final grade to drop a letter. Six absences may earn you a failing grade for the course.
• Please schedule doctor’s appointments, interviews, etc. for times other than class sessions.
• In the event that you encounter a health or life issue that requires you to miss class (such as a physician providing you with instructions that necessitate your quarantine) please notify me as soon as possible to provide an idea of the severity of your illness/issue and the length of time needed for recovery.
• If you attend a conference or job interview, you are responsible for getting all information you may have missed from your classmates or teammates.You may be marked down.

Keep in mind, you are responsible for information you miss through absences or lateness. (Note: If your illness/issue requires recovery time that exceeds the absence policy for a passing grade, a leave of absence may need to be considered. If this becomes the case consultation with university resources on how best to support you may be necessary.)

Late reading responses will be marked down by a point a day. Late conference proposals will decrease your Conference grade by a point a day. Failure to attend the final class and participate in this session may cause your final grade to drop a letter. Please bring academic timing conflicts to my attention as soon as possible and do not make travel plans before verifying the date of the event with me.

Academic integrity

The point of this class is to develop and situate your own ideas in a broader discourse — and in order to do that properly, you need to cite your work. No form of academic dishonesty will be tolerated. When you use words, images, videos, or even ideas and thoughts that are not yours and that you do not credit or properly cite, you are guilty of plagiarism. It is better to ask for more time on a deadline than to plagiarize. If you have any questions, ask. Do not cut and paste from other sources, even into your own notes, without keeping some system that tells you exactly where your work came from. Get in the habit now of taking good notes. If you quote the readings, put the quoted parts in double quotation marks and use APA style to cite.

If you are caught plagiarizing, you will receive a zero on the assignment. But that does not rule out further ramifications depending on the infraction, in accordance with CMU’s academic integrity policies.

Please take care of yourself

Remember that we — your professors and your classmates alike — want you to succeed and thrive.

Please take care of yourself. CMU is an intense place and can make emotions more vivid. Stress is real. Emotions are real. Depression is real. The effects of the changing seasons are real, especially if you’ve not lived somewhere that gets dark or snowy.

Some things to remember:

• Getting enough sleep is important. It can be hard to come by when you’re under a lot of pressure, but it’s one of the best things that can help to protect you when you’re feeling like you’re under stress.
• Move your body and get exercise. Walking, running, and biking in this city are a pleasure and even on campus, you can find some lovely places for greenery and reflection.
• Eat well. Make sure you have access to healthy snacks at school and take time for meals.
• Engage mindfulness practices like breathing and yoga. CMU offers a mindfulness room that offers meditation sittings. There are yoga classes at school, around Pittsburgh, and online. CMU offers a free Headspace account.
• Alcohol is a depressant and it’s easy to build up a tolerance to it. Take care around your use of it.
• Make time to relax. Grad school can absorb you 24/7 if you let it. Remember the things and people who are outside the studio world and engage with them.

If you are struggling or feeling down, please know that you are not alone. All of us benefit from support during times of struggle. There are many helpful resources available on campus and asking for support sooner rather than later is often helpful.

If you or anyone you know experiences any academic stress, difficult life events, or feelings like anxiety or depression, we strongly encourage you to seek support. Counseling and Psychological Services (CaPS) is here to help: call 412–268–2922 and visit their website at http://www.cmu.edu/counseling/. Consider reaching out to a friend, faculty or family member you trust to get connected to the support that can help. Again, you’re not alone.