Welcome to CSE190/291P
Hi!
If you're getting this, we've approved you to enroll in CSE190/291P, Generative AI and Programming, for spring 2026. We're excited to have you!
There are a few things we want you to be aware of before you enroll:
The course is new, and experimental. The field of generative AI and programming is large and moving fast; there's no way any of us can claim to be exhaustive experts. As a result, the course is going to involve learning from each other and experimenting with new tools just as much or more than Nadia and Joe telling you how things work or what to do. So expect project work, frequent demos, and peer feedback.
We don't have any guarantee of special funding for AI models, accounts, and usage for the course. We're assuming that students will be able to sign up for (potentially paid) accounts on services like Anthropic, OpenAI, Google Gemini, and so on in order to do work in the course. We don't expect it to cost more than about $100 of usage for the whole quarter, but this is an estimate, not something we can promise. Please reach out if this is a barrier for you (in particular if you would otherwise get things like textbooks paid for through a source like a scholarship).
We don't have a specific syllabus that we are committing to yet; you have to be a bit willing to figure it out along with us starting on day 1 of spring quarter 🙂. We do have a general plan for the broad themes and format of the projects:
A big theme for us is new programs we can build that would have been difficult or impossible without Generative AI. This means this course is not focused solely on using a coding agent like Claude to write existing programs faster than we could have before, or to write large programs as solo developers rather than with a team. Rather, we are interested in what new applications are possible: large-scale text processing, working with documents like PDFs or images or videos, presenting users with an agentic interface to existing but hard-to-use tools, etc. We will think about the interface we present to end users (who may or may not be software developers), and how to make generative AI systems work for them.
A general format will be open-ended projects with a round of demos, peer review, and responses to feedback. That is, you will not be submitting projects for the course staff to peruse and grade with a precise rubric for functionality. Rather, your goals will be to present and explain the work you've done, according to an assignment topic and theme, and then give and respond to feedback from your peers about the systems you all build.
In terms of preparation, we expect that the following kinds of tasks are reasonable to you, or you are confident in figuring them out on your own (potentially with the assistance of your favorite AI agents!):
- Using API keys (e.g. managing environment variables, configuration files, and source control)
- Using git and GitHub, making pull requests, creating and replying to issues
- Writing multi-file programs from scratch, including setting up tests
- Basic command-line user interface design
- Basic understanding of web servers and web pages: e.g. a web server is a program that stores some state and replies to requests with HTML pages
Feel free to reply with questions, but we may not be able to answer them all yet! If you know that you do not want to enroll, please let us know so we can admit additional students.
Best, Joe and Nadia