AI Software Design Studio

ID 20ST-09 Special Topic Design Studio, Fall 2024

RISD BFA Industrial Design

Software is now integrated into every aspect of our lives. The design of the digital products we use every day is more important than ever.

But phone and computer interfaces are now mature and optimized. Patterns are commoditized, processes are industrialized. Soon, generative AI and non-designers will be able to create live code for clean, standard interfaces — without needing professional designers to mock it up.

The industry does not need more product designers to reinvent the wheel. The job market reflects this.

At the same time, the arrival of generative AI models creates a new frontier for computing and opens up endless new creative possibilities. The way computers work is fundamentally changing — they can talk back to us now, and create artifacts of their own. Designers are sorely needed to shape this new material. Likewise, the processes and responsibilities of the designer will evolve beyond the old ways of static mocks and design systems.

Designers must adapt and shape the frontier of AI-driven computing — while navigating the opportunities, risks, and ethical responsibilities of working with this new technology.

In this new world, creation is cheap, craft is automatable, and everyone is a beginner. The ultimate differentiator will be the creator’s perspective, taste, and judgment. The software design education for our current moment must prioritize this above all else.

By course's end, students will have hands-on experience with an end-to-end digital product design process, culminating in a physical or digital product that takes advantage of the unique properties of generative AI models. Prior coding experience is not required, but students will learn using AI coding assistants like ChatGPT and Claude. The instructor's professional practice in digital product design, code-based prototyping, and experimental AI interfaces will guide the course.

To thrive in this new world, designers must be experimental, adaptable, and ambitious.

This class strives to embody these values as well. This is a highly experimental new curriculum in a rapidly-changing field. Whereas typical class curricula lag 4-8 years behind industry, in this class we will play with new tools and models as they are released, and hear from guests who worked on those new releases. Guest Q&As will show the wide range of perspectives on the future amongst industry practitioners working with AI.

This is a studio class where students will learn by making. In-class time will prioritize crit, higher-level lectures, and guest Q&As. Though students are expected to use tools like Figma and HTML/CSS/Javascript to complete their assignments, there are currently no plans to use lecture time to teach the minutiae of how to use these tools. There are already many resources elsewhere, e.g. via YouTube. Students are expected to consult these outside resources in order to complete their assignments. General in-class guidance related to these tools will focus on brief introductory demos, alongside broader techniques for problem-solving, prototyping, user testing, iterating, and debugging. Of course, students will get relevant, personalized guidance on tools and debugging during in-class work time.

This syllabus reflects a set of intentions, but it will not reflect the class that unfolds. Material will be adjusted for students’ skill level and interests. Students are encouraged to suggest topics and guests over the course of the semester. Did you see a cool new product? We can probably bring in a guest who worked on it.

Class materials will be open-sourced and recorded whenever possible, for the benefit of self-directed learners beyond this classroom. More information to come.

Sep 9, 2024

Sep 16, 2024

Sep 23, 2024

Sep 30, 2024

Oct 7, 2024

Oct 14, 2024

No class, but office hours by appointment during Monday 8am-1pm if you want feedback.

Continue working on  Bring people together - Week 5 of 5: Prepare presentation

Oct 21, 2024

Between class 7 & class 8: midterm project grades & what grade are you on track for in the overall class

Oct 28, 2024

Oct 30 — final day to withdraw from the class

Nov 4, 2024

Nov 11, 2024

Nov 18, 2024

Nov 25, 2024

Dec 2, 2024

Note: As this is a 3-credit course, our final reviews are during our final studio class on Dec 2. For 6-credit courses, your final reviews are during crit week (Thu Dec 5 - Wed Dec 11). Do not leave for the holidays before your 6-credit course’s final review!

Make a clock: 10%

Bring people together: 30%

Experimental AI interface: 40%

Conduct, attitude, professionalism, adaptability, rate of learning: 20%

Grading for projects in this class start from a baseline of C, which is the sufficient basic execution of all the assignment requirements. In the clock project for example, a project with 100 sketches of any quality, a Figma mockup and filled-out Figma template, and some version of a clock taken from Claude and copy-pasted into a Replit project, would result in a C.

B-grade projects show that effort and care has been put in to create an interesting output, even though the final product may not be at the level of polish of an A-grade project. In the case of this project, a “B” project could look like strong conceptual thinking paired with execution that could be improved. Or it could reflect a visually-polished project with a standard concept.

A-grade projects have excellence in execution and visual polish, supported by solid conceptual thinking. A-grade projects would feel at home in a professional designer’s portfolio. An “A” means that if I saw the final product in a student portfolio, I would be intrigued or impressed. To meet this bar, this usually means it needs to:

“A”s are rare.

+s and -s on the grade may be added based on effort and learning.

Not completing the assignment requirements results in an incomplete or D grade. Late assignment submissions that aren’t previously excused due to illness will be deducted with a 10% (single letter grade) penalty.

Giving “D”s for not completing assignment requirements is not meant to punish ambitious projects that just don’t get finished in time. Many students clearly go beyond their comfort zones and learn a lot about the process of building software on the way. However, it is a valuable skill to know how to manage your time and cut scope in order to take something across the finish line. Each assignment step is designed to help you learn something about the process of creating software. In the clock project for example, if you made a project that stayed within Claude and didn’t move code to Replit, it means that you lost the opportunity to edit code yourself to further control the output. If you made a good final clock website but didn’t put in much effort on the sketches, it means you lost the opportunity to push your creative brainstorming skills.

These links are optional for students interested in diving deeper into the materials covered in class. Relevant details will be mentioned in class.

Thanks to Barron Webster, Ryan Mather, Max Bittker, Andres Cwervo, Ben Spear, Jerry Hong, Jenna Meacham, Evangeline Stott, Kasey Klimes — and more names I’m sure I’m forgetting — for their feedback, advice, and help in planning this class.

Endless gratitude to guests who have taken time out of their day to share their knowledge with the next generation of designers: Greg Hochmuth, Oscar Dumlao, George Kedenburg III, Maddie Simens, James Buckhouse, Agatha Yu, Ryan Mather, Jason Yuan, Derrick Schultz, Willem van Lancker, and more to be announced.