Welcome to CS193T!
We're so glad you're interested in the course. Below is an overview; you can also browse the syllabus or topics schedule. If you have questions about which AI course(s) offered in CS would be right for you, check our course placement guide.
Auditors
Posted by Cynthia on 09/21/26
We welcome all interested Stanford affiliates (students, staff, and faculty with an active Stanford email) to join us in auditing this course. Due to space and staffing constraints, auditors do not attend in person or receive support/advice from course staff, but are welcome to watch lecture videos on Canvas and self-study the tutorial content and assignments on this website. To be added to Canvas as an auditor, email the Head TA Piper. Once on Canvas, the videos are under "Panopto Course Videos".
For non-Stanford-affiliates, this course will be available for remote study as a credit-bearing course (i.e., you get a real Stanford transcript and transferrable credits for it) for a tuition fee, through Stanford Online. More information here.
Course Overview
Whether you’re excited, curious, skeptical, or just unsure where to start with AI, this course meets you there.
- Weekly hands-on tutorials, where you build with a dedicated TA who follows your work all quarter.
- Assignments are intended to be customized to the work you actually need to get done in your major, your career, your life.
- Guest speakers showcase use cases across the humanities, sciences, law, business, and more.
- Term project: Create a custom AI system that aids the work you’re doing now, and is portfolio-ready for your job search.
You’ll learn not just AI tool how-to, but when to trust, when to push back, and how to keep your own thinking and learning sharp along the way. We want to give you practical, immediately-applicable skills to get the most out of AI, but we’re not purveyors of mindless pro-AI hype. Our exploration of AI will take seriously societal concerns about bias, copyright, labor impacts, and environmental costs, as well as potential individual risks like emotional manipulation and impaired learning.
Topics
- How AI really works
- Advanced prompting skills
- Probing models’ strengths and weaknesses
- “Agentic” AI and complex reusable workflows
- Guest panel: research/industry use cases
- Educational AI tools, reflective self-governance of AI use
- Societal impacts
- Building audit and accountability systems
- Efficiency, tokens, and budget management
- Evaluating models
Tools
- Anthropic: Claude Chat, Claude Cowork, Claude Code
- Google: Gemini Enterprise, Notebook Gemini (AKA NotebookLM), Deep Research, Data Insights, Code Assist
- OpenAI: ChatGPT, ChatGPT Work, ChatGPTEdu, Codex, ChatGPT Deep Research
- Free, open-source models (Deepseek, Mistral, Llama)
- Connectors from AI models to your Google Drive files, email, Slack, and more
No Prerequisites!
There are no prerequisites for this course. In fact, we’re hoping to create a comfortable place for all kinds of folks to learn and grow!
Units and Grading basis
2 units, S/NC. Our focus is on helping you deploy AI in your work, not points.
Lecture
Lectures will take place on Tuesdays at 1:30-2:50PM in NVIDIA Aud (Huang Building). There will be video recordings for auditors and excused absences.
Tutorials, Workload, Grading
Starting in week 2, you’ll meet weekly for an 80-minute tutorial in Coda, B80: 20 minutes on that week’s tool, then a full hour building with it yourself while TAs circulate to help. You’ll be placed in one block — Tuesday, Wednesday, Thursday, or Friday — and stay in it, with the same TA and the same table group, all quarter.
Each week your TA pulls you aside for a ten-minute check-in on the previous week’s assignment, tailored to your actual use cases, so you’ll always know whether you’re on track for passing expectations. Tutorial attendance is mandatory. Read more about tutorial.
Office hours are folded into tutorial this quarter: drop in on any tutorial block and a TA on floater duty will help you.
Assignments are given weekly and assessed during tutorial. Most assignments are just asking you to deploy AI on the actual work you need to do anyway, so it is our expectation that the course will not add significant workload. We do want you to add layers of reflection, analysis, and comparison to your AI practice (e.g., If you attempt the same task using several different AI models what differences do you observe? Is there a better way you could have structured your use of AI for the task that would reduce token consumption and therefore environmental impact? What safeguards could you put in place to ensure the result is accurate?).
The assignments will culminate in a complex, custom AI system that will be showcased at an end of quarter poster session (no final exam).