Class Meetings

Thursdays @ 2:20 PM - 5 PM
in the HCI Lab (SCI L120)

Prerequisites

CS204 or CS220 or CS230, or Permission of instructor.

Topics

• Human-Centered AI
• Research methods in HCI
• Interaction with large language models
• Mixed reality interaction
• Human-computer interaction with highly automated vehicles
• Designing and prototyping conversational agents
• Promise and Peril of AI and automation

Textbook, Readings, and Materials

We will be using videos, podcasts and research articles which are available online. Research articles are either linked from the course schedule or available in the CS366 e-reserve Google Drive.

We will also be using the books:

"Research Methods in Human-Computer Interaction" 2nd Edition by Jonathan Lazar, , Jinjuan Heidi Feng, and Harry Hochheiser.

"Human-Centerd AI" by Ben Shneiderman.

The books are available as an ebook in the library.

Policy on the Use of Generative AI

This course encourages students to explore responsible and critical engagement with generative artificial intelligence (GAI) tools such as ChatGPT. The use of GAI or AI powered tools is allowed on all assignments as long as students follow the requirements below:

Transparency and Attribution

Any use of a generative AI tool must be appropriately acknowledged - specify what AI tool was used and for what purpose.

For example, let's say a student used a generative AI language model to assist in writing a section of their research paper on the history of HCI. In this case, proper attribution would involve acknowledging the AI model used and providing details about it:
"In this section, we discuss the history of HCI. We employed the GPT-3.5 language model developed by OpenAI. The AI model was used to draft the initial paragraph, which was then reviewed and edited for accuracy and coherence by the authors."

Or, suppose a student used an AI powered tool for editing their blog post on the HCI for the future of remote work. In this case, proper attribution would involve acknowledging the AI tool used and describing its role in the editing process:
"In the process of refining this blog post, we utilized the ProWritingAid AI editing tool to identify and correct grammatical errors, improve sentence structure, and enhance overall readability. The edits suggested by the AI tool were reviewed and accepted by the author."

Failure to attribute content to an AI program is academic dishonesty.

Inaccuracy and Bias

Generative AI may generate incorrect information. Large Language Models such as GPT-4 and Bard might produce false "facts" and citations. Their output can contain inaccuracies, biases, or outdated information. Code generation models often produce inaccurate outputs, and image generation models can occasionally come up with highly offensive products or create images that are too similar or identical to existing images.

Blindly accepting AI-generated content without critical evaluation can be misleading and potentially harmful.

Students are responsible and accountable for the contents of any submitted work. Specifically students are responsible for any plagiarism, misrepresentation, offensive content, fabrication or falsification of content and/or references by AI tools. Violations of this policy will be considered Honor Code violations.

Note that in general, AI tools may be safely used to copy-edit student-generated content, but caution is advised in using them for generating new content.

Learning Together

The course will provide resources and support for students to better understand generative AI, its ethical implications, and responsible use. Students are expected to engaged with assign readings on this topic and to pursue and share additional relevant materials.
Students are encouraged to seek guidance from the instructor or teaching fellows as needed.

We draw your attention to the fact that different classes at Wellesley implement different AI policies, and it is the student’s responsibility to conform to expectations for each course. Our class policy is consistent with the ACM Publications Policy.

Course Requirements

The course requires active participation in class through discussion and hands-on activities. 
Students are expected to attend class regularly, arrive on time, and engage respectfully with peers and with class material.

Students are required to prepare for class by completing the reading and submitting their commentaries prior to lecture. Submit a commentary form for each requried reading by 11:59 PM the day before class.

There are two individual homework assignments in this course that explore research skills and methods of human-computer interaction research.

The course has a final project that requires students to work in teams of 2-3 students to research a topic related to human-computer interaction with AI and automation.

Grading Policy

Your grade will be computed as a weighted average of several components. The relative weight of each component is shown below:

• Homework assignments 20%
• Reading commentaries 20%
• Active participation and submission of in-classengagement tasks: 15%
• Group research project: 45%
• Total: 100%

Reading commentaries will be graded for submission - for example if you submitted 75% of the required commentaries your submission grade will be 75/100. In addition, a random sample of 4 commentaries will be graded on a check-minus/check/check-plus scale. These scores correspond to 3, 4, and 5 points. Throughout the semester, you may opt to pass on submitting four individual commentaries without detracting from your submission grade. There are no exemptions beyond this.

In-class activity grade will be determined based on completion; e.g., if you complete and submit 3 out of 4 tasks for a lab, you will receive a 3/4 or 75% on that lab.

Homework assignments and project milestones grades will be based on grading criteria specified in the description of the assignment.

The mapping from numerical score to letter grade looks like this: >=95 is an A, >=90 is an A-, >=86 is a B+, >=83 is a B, >80 is a B-, >=75 is a C+, >=73 is a C, >=70 is a C-, >=60 is a D, <60 is an F.