---
title: Extended Intelligences II
page_type: course
track: Explore
course_type: Workshop
feature_img: /assets/images/2024-25/year-1/t-2/extended-intelligences-ii.jpg
faculty:
    - christian-ernst
    - pietro-rustici
ects: 3
---

{{ insert_banner() }}


## Syllabus

This course offers designers and makers a hands-on introduction to generative AI through physical computing. Participants will learn to build agentic systems that bridge the digital and physical worlds by connecting sensors and actuators to AI agents.

Course Contents:
1. Showcase of Salient Projects: The instructors will demonstrate projects that combine generative AI with physical computing and interactive systems.

2. Introduction to Generative AI and MCP: Participants will learn the fundamentals of generative AI and the Model Context Protocol (MCP), understanding how AI agents can interact with the physical world through connected devices.

3. Building with Embedded Systems: Students will work with wifi-enabled microcontrollers (such as Raspberry Pi Pico W) to create local MCP servers that expose sensors (temperature, light, sound, humidity) and actuators (LEDs, motors) to AI agents.

4. Prototyping Agentic Systems: Using Langflow, participants will design and prototype AI agents that can sense and respond to their environment, creating interactive, intelligent physical systems.

5. Design Considerations and Ethics: The course addresses responsible AI usage, privacy concerns in sensor data collection, and ethical principles for designing AI systems that interact with physical spaces.

By the end of this course, participants will have built their own AI-powered physical computing project, gaining practical skills to create intelligent, responsive systems that blend generative AI with the physical world.


**Keywords: Generative Artificial Intelligence, Agentic Systems, Rapid Prototyping**

### Learning Objectives

Learning objectives:

1. Gain a clear understanding of the concept of generative AI, its principles, and its applications.
2. Learn about algorithms, models, and techniques used in generative AI.
3. Develop practical skills in agentic systems and tools via MCP for generative AI projects.
4. Understand the ethical considerations and responsible usage of generative AI.
5. Develop a solid foundation in generative AI for future design and making endeavors.



### Structure and Phases (Schedule)

The course unfolds across three intensive 3-hour sessions. 

**Day 1** 
- Introduction and projects showcase 
- Introduction to the class assignment
- Starting group making and hardware setup

**Day 2** 
- The evolution of LLMs and agentic systems
- Introduction to the anatomy of a MCP server and AI tool usage
- Follow-up support for the class assignment

**Day 3**
- Project Execution in Group Work
- Students presentation


## Methodological Strategies

- Introductory lectures to build an understanding of the problem space
- Group project execution phase to apply learnings on a chosen topic 
- Academic understanding
- Hands-on/ tactile experience
- Learning by application
- Collaborative project execution
- Iterative Design, Design Thinking


## Deliverables

A fully functional physical computing prototype that connects sensors and actuators to an AI agent through MCP, demonstrating meaningful interaction between the digital and physical worlds. 15-minute presentations showcasing the conceptual foundation and live demonstration of the working system. Course documentation on students' blogs summarizing the project outcome, technical implementation, and personal reflection on designing agentic systems for physical spaces.

## Grading Method

| Percentage  | Description                         |
| ----------- | ------------------------------------|
| 20%         | Participation                    |
| 30%         | Prototype and Conceptual Quality   |
| 30%         | Presentation                    |
| 20%         | Reflection                  |



!!! ects "European Credit Transfer and Accumulation System (ECTS)"

    {{ ects }} ECTS 

## Additional Resources

- [Dottod I: gallery of cybernetic interpretations](https://www.dottod.net/) - Project
- [Dottod II: Icon's replicants](https://furnitures.dottod.net/) - Project
- [Del Complex / Del Complex Incident Report September 2023](https://xiosky.com/delmedia/media/archival_media/Del_Complex_Incident_Report_7_September_2023.pdf) - Project
- [Communicative Agents for Software Development](https://arxiv.org/pdf/2307.07924.pdf) - Paper
- [The Reversal Curse: LLMs trained on "A is B" fail to learn "B is A"](https://arxiv.org/abs/2309.12288) - Paper
- [The Zizi Show](https://www.jakeelwes.com/project-zizi-show.html) - Project
- [Large Language Models as Optimizers](https://arxiv.org/abs/2309.03409) - Paper
- [Decomposing Language Models Into Understandable Components](https://www.anthropic.com/index/decomposing-language-models-into-understandable-components) - Article
- [Infinite Images and the latent camera](https://mirror.xyz/herndondryhurst.eth/eZG6mucl9fqU897XvJs0vUUMnm5OITpSWN8S-6KWamY) - Article

## Materials Needs 

Student’s own computer with a web browser and a python development environment. 


## Faculty

{{ insert_faculty() }}