Course Description
AI tools can now write your essays, generate images from a sentence, and hold conversations that feel disturbingly human. You’ve probably already used them. But do you know how they actually work? Do you know who built them, what data they were trained on, and who benefits when you use them? This course takes AI seriously in three ways: technically, critically, and creatively.
On the technical side, you’ll learn what’s actually happening when a language model generates text or an image model turns a prompt into a picture. You don’t need to be a programmer to understand the basic architecture: training data, weights, attention, and the difference between what these systems do and what they appear to do. We’ll trace the history of AI from its origins in the 1950s, through symbolic AI and neural networks, to the large models that are available today.
On the critical side, you’ll wrestle with questions that don’t have easy answers. Who owns the output of a model trained on millions of people’s work? What does it mean when a hiring algorithm discriminates and no one can explain why? How should governments regulate systems that their own engineers don’t fully understand? How can we use AI in an ethical way?
On the creative side, you’ll learn to use these tools. Prompt engineering is a real skill, and the difference between a naïve prompt and a thoughtful one is enormous. You’ll work with language models and image generation tools, learning what they’re good at, where they fail, and how to push them toward results that are actually yours.
Learning Objectives
- Explain at a conceptual level how large language models and image generation models work, including the roles of training data, model weights, and attention.
- Trace the historical development of AI from its origins in the 1950s through symbolic AI, neural networks, and the emergence of large-scale models.
- Write effective prompts for both language and image generation models, and understand why prompt design matters.
- Analyze questions of bias, fairness, and accountability in AI systems, drawing on concrete cases rather than abstract principles alone.
- Evaluate the social and economic implications of AI, including its effects on labor, copyright, creative ownership, and the concentration of power.
- Engage seriously with multiple perspectives on AI policy and governance, understanding the interests and arguments of different stakeholders.
- Develop a position on what AI should and shouldn’t be used for, grounded in both technical understanding and ethical reasoning.
Readings
- Ted Chiang. “ChatGPT Is a Blurry JPEG of the Web.” The New Yorker, February 9, 2023. Online.
- Ursula M. Franklin. The Real World of Technology, rev. ed., Chapter 1. CBC Massey Lectures. Toronto: House of Anansi Press, 1999.
Evaluation
| Midterm |
40% |
Comprehensive |
| Final Examination |
40% |
Comprehensive |
| Simulation |
20% |
— |
- Midterm and Final Examination: Both tests will cover everything we’ve discussed in class and read about. You’ll see conceptual questions, and gobbits.
External Links
- Class Schedule : Schedule of activities
- Etherpad : Information sharing
Other Material
- 3Blue1Brown. But what is a neural network? | Deep Learning Chapter 1
- 3Blue1Brown. Transformers, the tech behind LLMs | Deep Learning Chapter 5
- TuringTest. Can you pass the Turing Test?
- SightEngine. Which image is AI?
Course Policies
- Outline: This outline is subject to arbitrary change. I shall announce any changes in class; if you are not present, you are still responsible for finding out what I announce.
- Missed Midterm: There is no makeup midterm. If you have a documented medical or compassionate reason for missing the midterm, the weight of your final examination will be prorated.
- Attendance: Come to class. Two unexcused absences is a marginal pass; three is a failure, even if your other work is fine. Excused absences don’t count.
- Laptops and Cell Phones: Don’t use mobile technology in class unless I specifically permit it.
- AI Tools: Don’t use generative AI for any assignments unless I explicitly allow it. If I do allow it, you must cite exactly how you used it according to Chicago style.
- Office hours: Office hours are first-come-first-serve, unless you have an appointment. If you want to schedule something, email me a few times that work for you.
- College Policies: Be familiar with the college rules on attendance and academic integrity as articulated in the Student Handbook
- Grading: If you opt for a letter-grade, your mark in the course will be translated according to the following scale: A+ (90–100), A (85–89), A– (80–84), B+ (77–79), B (73–76), B– (70–72), C+ (67–69), C (63–66), C– (60–62), D (50–59), F (0–49). If you do not opt for a letter-grade, the scale is as follows: Pass (65–100), Marginal Pass (50–62), Fail (0-49).
A Note on Liberal Arts Learning
An overarching objective of this course is to help you develop as a student of the liberal arts. True students of the liberal arts are able to reflect on the context in which they live, and reason about what it means to live a meaningful and happy life. Thus, they are able to be more than just children of their own time. But this means we must be willing to put our ideas to the test, see our own errors, and develop intellectual courage and humility. It also helps not to take ourselves too seriously.
