{"id":13336,"date":"2025-07-18T10:27:21","date_gmt":"2025-07-18T14:27:21","guid":{"rendered":"https:\/\/itp.nyu.edu\/physcomp\/?page_id=13336"},"modified":"2025-12-03T20:56:24","modified_gmt":"2025-12-04T01:56:24","slug":"jeff-feddersen","status":"publish","type":"page","link":"https:\/\/itp.nyu.edu\/physcomp\/itp\/sections\/jeff-feddersen\/","title":{"rendered":"Jeff Feddersen"},"content":{"rendered":"\n

Fall 2025<\/p>\n\n\n\n

Tuesdays, 12:20pm to 2:50pm in 370 Jay St, Room 413<\/p>\n\n\n\n

Important Scheduling Note: <\/strong>Due to a conflict, Class 4, scheduled for September 23rd, will be rescheduled to MONDAY September 22nd. Exact time TBD. <\/strong><\/p>\n\n\n\n

<\/span>Welcome<\/span><\/h2>\n\n\n\n

Welcome! I\u2019m very excited to be teaching Physical Computing again this year. It\u2019s material I love<\/a> and use a lot<\/a>. I created and teach two other courses in the Physical Computing Area: Energy<\/a> and Time<\/a>. I\u2019m curious about every aspect of how things are made \u2013 how materials can be shaped into useful and beautiful forms, how sensors can perceive the world, how code can be crafted to run on processors and affect the world.<\/p>\n\n\n

\n
\"Gif<\/a>
Me in my workshop surrounded by pcomp stuff.<\/figcaption><\/figure><\/div>\n\n\n

Recently there\u2019s been an explosion in tools available to beginners for embedding computation into just about any project imaginable. It can be a bit overwhelming \u2013 electronics! programming! so many boards to choose from! \u2013 but this course lays a foundation for you to build amazing things now and continue developing creative and technical skills over a lifetime.<\/p>\n\n\n\n

<\/span>AI<\/span><\/h2>\n\n\n\n

Another very new and rapidly changing factor in our field is Artificial Intelligence. I began studying Computer Science during the \u201cSecond AI Winter\u201d<\/a>, when progress in the field seemed to stall. But we are all familiar with the unprecedented pace of AI today, and recent (as of 2025) model advances have made LLMs<\/strong> amazing tools tool for writing code, and even designing physical objects<\/strong>. (Pause a moment to consider that these are emergent abilities, often discovered after the fact<\/a>, not “built-in” by design.) AI can be an powerful learning guide and creative tool, and at the same time it can create very difficult to solve problems when learning a new field and trying to build things that work in the real world. Everyone – you as students, me, industry pros, everyone – is learning right now<\/em> how to harness these new capabilities and handle these problems. <\/p>\n\n\n\n

My plan this year is to remain open to AI<\/strong>. It’s not going away. There will be good and bad ways to use it. In preparation for this year\u2019s class, I\u2019ve been trying to put myself in the mindset of a student<\/strong> to see how tools like Gemini<\/a>, Claude<\/a>, and ChatGPT<\/a> can help understand the material of this course. I\u2019ve asked models basic questions about electronics, requested explanations of code, and even given open-ended prompts to build passable midterms or finals for this class. The results are fascinating!<\/strong> Often helpful, sometimes astoundingly so; occasionally confusing and downright frustrating. Even sometimes – dangerous. But with enough potential that I\u2019ll continue to explore and share the results as we work together this semester.<\/p>\n\n\n\n

\n
\"Google's<\/figure>\n\n\n\n
\"Anthorpic's<\/figure>\n\n\n\n
\"The<\/figure>\n
Some potential tools… <\/figcaption><\/figure>\n\n\n\n

I also have so many things I still want to learn<\/strong> – new boards I haven\u2019t had time to program yet, new IDEs and development tools I want to add to my skill set, let alone vague project ideas I want help moving forward – and I\u2019m seeing how AI can assist, very much as you might during your time at ITP. <\/p>\n\n\n\n

If you are coming to this class as a first-time programmer, I\u2019ve always felt that PCOMP is a great way to learn coding<\/strong>. It\u2019s so fun to see what you write come to life in a real object that is running just your code. I think PCOMP will be a great way to engage AI, too, because there is an immediate, testable ground truth – does the code run? Does the circuit work? <\/p>\n\n\n\n

AI can too easily replace deep study with easy answers<\/strong>, short circuiting our learning<\/a>. And I’m wary – and puzzled, excited, thrilled – about what it will mean to \u201cbe creative\u201d in an age where machines can do things that mimic art. I\u2019m so glad to be working with you this semester to find ways to accelerate our learning and deepen our understanding. Our task is to learn and build while preserving the core, human part of our creative practice, using tools that will no doubt be different and more powerful at the end of our 14 weeks together than at the start. This is an amazing time to be studying creative technology! <\/strong><\/p>\n\n\n\n

3 AI rules for this class:<\/strong><\/p>\n\n\n\n