Critical Thinking with AI
Logistics
- Credit Hours: 3
- Meetings: Tuesday & Thursday 2–4:30 (usually ~4)
- Student Hours: Wednesday 10-12 @ Amaranth
- Amaranth Studio: 2068 Mesa Vista Hall
- History Office: 1077 Mesa Vista Hall
Getting Started
This is the home page of the course syllabus, which outlines all the instructions, logistics, and expectations for the course. The syllabus also has a schedule page, which details the reading and activities for each session.
Course Description
How do we know what we know—and what happens when machines start “knowing” things too? (and: do they?)
This course uses historical practices and narratives as a laboratory for exploring how AI can sharpen critical thinking and produce higher-quality intellectual work. Rather than treating AI as a shortcut or a threat, we treat it as a thinking partner: a tool that, when used deliberately, makes visible our own reasoning processes and exposes the assumptions we bring to interpretation, evidence, and argument.
Students will practice making sense of the past with the help of AI. The question we explore: how can it actually be helpful, as opposed to just stringing together words that sound authoritative? Some activities: finding sources, evaluating trustworthiness, synthesizing evidence, constructing arguments, identifying omissions. We constantly work alongside AI, learning when it helps, when it misleads, and what the difference reveals about how knowledge and expertise actually work.
Driving Question
How can AI help us think more critically, develop skills, and produce higher-quality work than we could without it?
Learning Objectives
- Understand how AI complicates ideas about thinking, knowledge, expertise
- Develop sophisticated prompting skills as a form of question design, critical inquiry, source evaluation
- Evaluate AI outputs for accuracy, bias, and omissions using lateral reading and groundtruthing techniques
- Identify characteristics of true expertise from beer-mat expertise
- Use AI as a research partner while maintaining academic honesty, intellectual agency and interpretive judgment
Instructor / Course presumptions
- This is not a credit recovery course
- You need a laptop or tablet you can bring to class for our workshop activities. You can check one out from the library if you need one.
- You actually want to learn how to learn with AI, not just complete assignments
- You will use AI as directed and refrain from using it when instructed (i.e., don’t bring a forklift to the weight room)
- You will turn in assignments on time not because there is a late penalty, but because it helps you do better on all the assignments and stay on track
Color Guide
Red boxes
These indicate something you have to DO or TURN IN.
Yellow Boxes
These indicate something you should be aware of—usually an upcoming assignment or a longer reading—but isn’t anything you need to immediately do.
Blue boxes
These indicate something that is important to know, but isn’t time sensitive.
Required Texts
All readings are available online or through Zotero. You never need to find anything!
We use a tool called Zotero to organize and provide access to all readings for the course. To get connected, carefully follow the getting started guide. If it doesn’t work for you, please follow the directions more carefully. They’ve worked for hundreds of students!
Zotero links
Remember that you can only access the PDFs in the library if you are a member of the course group!
Course Work
Math Fail?
Perhaps you've noticed these don't tally to 100%. The 10% buffer means that grading is less stressful for everyone. In other words, you get 10% of your grade for free to account for unclear directions, grumpy grading, or whatever.
Assessment Philosophy
You get graded on effort on this class, and the way you show effort is to show your work. It’s like math class, but without the math. I only care about the energy you put into an assignment and learning to use AI effectively, not what the final product is. It’s the process that’s important. You need to be sure I can see your process.
All work is graded on the following scale:
| Perceived Effort |
Grade |
Grade points |
| Very fine |
A |
4 |
| Fine |
B |
3 |
| Marginal |
C |
2 |
| Redo |
F |
0 |
Where's the D?
This also is on purpose. A lot of grading is about basic communication competence---like whether or not you can write a coherent paragraph. AI raises the bar for us. Notice you can't really fail, but you might need to redo the assignment if you want grade points. Courses that don’t challenge you with expectations are a waste of your time.
Final Grade Scale
Your final grade is simply the average of all your work (with assignment weights factored in), plus any extra credit from reading assignments and whether you’ve been a regular contributor to class discussions.
| Grade points |
Grade |
| 4.0 |
A |
| 3.7 |
A- |
| 3.3 |
B+ |
| 3.0 |
B |
| 2.7 |
B- |
| 2.3 |
C+ |
| 2.0 |
C |
| 1.7 |
C- |
| 1.3 |
D+ |
| 1.0 |
D |
| 0.0 |
F |
AI Policy & Academic Integrity
- AI is awesome! You may disagree, which is fine, but if you have objections to using it, this is not the course for you.
- Unlike many other courses concerned with what you produce for your assignment, I care only about HOW you produce your assignment and how much better than vanilla AI they are.
- AI should help make your reasoning MORE visible, not replace it.
- It is always incumbent on you to differentiate yourself from AI. If I can’t, you will need to redo the assignment.
- I can’t really prove you’re cheating if you do, but that’s OK. Eventually you’ll lose out to people who actually learned to use AI critically and productively instead of deceptively. That seems a sufficient enough consequence.
- ChatGPT, Claude, Gemini, etc. Free versions should be sufficient, but if you also use them for other things, you might need to pay for a month of one or two of these (~$20; way cheaper than books!).
- Google NotebookLM — Source-grounded AI for research with specific documents; free with Google account and we don’t need the pro version.
- Zotero Free bibliographic manager for distributing course readings. Way better than Canvas.
- GitHub — For hosting and collaborating our Disruptive Expertise site; free account required
Read to Engage, Not to Agree
For almost everything we read, we’re reading to ENGAGE with it, not because it’s right. There is a LOT to disagree with across the readings, and we don’t all have to agree on everything. The goal is to develop frameworks for thinking critically about AI, knowledge, and expertise.
Accessibility & Support
If life gets overwhelming during the course, please reach out. We can discuss formal or informal accommodations, deadline adjustments, or other support. The goal is maximizing learning under real life circumstances.