Get to Know the AI Literacy Lab

Our Purpose: The University of Guelph-Humber’s AI Literacy Lab was established to explore the transformative potential of artificial intelligence (AI) in higher education. By undertaking small-scale experiments with AI tools, the lab aims to build out a database of practical applications and scenarios that can enhance pedagogy in teaching and learning, assessment and evaluation. We are here to serve the University’s mission to nurture any and all paths to student success.

We gratefully acknowledge financial support for the AI Literacy Lab, made possible in part by the University of Guelph’s University Teaching Leadership Fellowship and the University of Guelph-Humber’s Research Fund.

The Team at the AI Literacy Lab

Oluwafunmilayo (Funmi) Ogunbi, graduate student at the Lang School of Business and Economics at the University of Guelph

Ekaterina (Katerina) Kalantyrski, graduate student at the Lang School of Business and Economics at the University of Guelph

Mara Strazzeri, undergraduate student at the University of Guelph-Humber, Justice Studies

Ashnaa Narumathan, undergraduate student at the University of Guelph-Humber, Kinesiology

Julianne Uzun, undergraduate student at the University of Guelph-Humber, Community Social Services

Natasha Tat, undergraduate student at the University of Guelph-Humber, Psychology

Neil Andrews, undergraduate student at the University of Guelph-Humber, Business

Skyler Volanska Hutt, undergraduate student at the University of Guelph-Humber, Psychology

Matthew LaGrone, Program Head, Liberal Studies, University of Guelph-Humber; Director, AI Literacy Lab; Adjunct Professor, Department of History, University of Guelph. For any questions, please email [email protected]

All four above images were created using the same prompt on ChatGPT Plus Dall-E 3: ‘A child and AI robot sitting cross-legged, exchanging books’

Fine-Tuning (or: What we are working on right now)

• Faculty Workshops: We've run a series of workshops tailored for faculty. To access resources from our previous workshops, please visit the following link (Guelph/Guelph-Humber login credentials required): Workshop Resources

• AI Tool Reviews by Research Assistants: Each Research Assistant is reviewing various AI tools (for text, image, video, audio, etc.).

• Scholarly Resource Database: We will be going live this academic year with a comprehensive database of scholarly resources.

• Social Media Presence: To engage a wider audience and disseminate our findings, we are developing a robust social media presence, including a dedicated YouTube channel (Jan. 2024).

• Collaboration on Intercultural Communication: On a more focused note, we're collaborating with a professor at the University of Guelph-Humber to explore how generative AI can enhance intercultural communication (through voice cloning, translation and transcription services, and culturally appropriate avatars).

Human-in-the-Loop (or: How we can help!)

• Collaborative Projects: Our team is eager to work with instructors on classroom- or research-based (implementation or build) projects. Whether you have an idea you’d like to explore or a fully-defined project, we're here to offer our support and resources to grow your pedagogy.

• 1:1 Consultations: For a more personalized discussion or guidance, we are available for one-on-one consultations. Whether it's understanding the basics of AI, its potential implications in your field, or brainstorming innovative applications, we're here to assist. We will work with your schedule, and we are happy to meet in person or on Teams/Zoom.

• Upcoming Workshops: If we aren’t shipping and sharing knowledge, then we are failing in our responsibilities. To that end, we are planning a workshop next month. On Thursday, November 23 from 1pm-2pm, the AI Literacy Lab, in conjunction with the Associate Vice-Provost’s Office, is hosting “A Sharing Session of Student AI Use” in GH403 or streaming on MS Teams. Plus: Stay tuned for more workshop opportunities in Winter 2024!

• Contact us: [email protected]

AI in the Classroom: The 5 C’s of Prompting

Think of tools such as ChatGPT as a co-pilot; that is, you and the computer are working together.  

To enhance the quality of AI output, provide context—be as granular as possible, offer detail, ask for what you want—and, if possible, provide examples (this is called “few-shot prompting”). 

Part of dressing your prompts with useful context is the use of constraints: you can ask the AI to consider this and not that, to weigh this variable more than that variable. You can ask it to respond in a specific voice. In ChatGPT Plus (soon to come to the free version as well), users can add custom instructions that are applied to all future prompts. This function is of great utility for researchers and students. 

You are not simply feeding the AI a command with an output you control (in fact, you are not in control of the output, and this is the creative and perhaps emergent aspect of LLMs), but rather are starting a conversation where you can probe and poke the AI in your follow-up queries to the prompt output. Don’t forget to say “please” and “thank you”!  

Unlike human co-pilots or assistants, AI never gets tired. It is constant and can sustain conversations, though the quality degrades at a certain point and it becomes necessary to start a new chat. However, LLMs such as ChatGPT will be introducing ever larger context windows (and some, like Claude, already have quite large windows—though Claude is not available in Canada at the time of this newsletter) to allow a longer “memory” in a specific chat.  

Try prompting today with ChatGPT!

In Review: Sudowrite

Unleashing Creativity: My Journey with Sudowrite

Julianne Uzun

Writing has always been my solace, my window into unknown realms. When I came across Sudowrite, an AI-driven creative writing tool, it felt as if I had unearthed a hidden gem amidst the vast array of writing aids. My journey with Sudowrite was filled with peaks of creativity and valleys of challenges.

Pros:

• Boosted Creativity: With Sudowrite by my side, I could effortlessly brainstorm characters, plots, and entire worlds, pushing away writer's block with each interaction.

• Visual Canvas: Initially daunting, the canvas became a haven where my scattered ideas formed coherence.

• AI Transformations: The platform's capacity to turn ordinary descriptions into vivid, sensory-rich narratives amazed me.

• Twist-generating Feature: This tool added unexpected twists to my narrative, keeping both me and my readers engaged and excited.

• Time-saving Tools: AI-generated book summaries and critiques freed up my time, allowing me to concentrate on my true passion - writing.

• Live Help - Classes & Author Hours: This new feature became my support system, providing immediate responses to my queries and expert advice for my concerns. The promise of weekly sessions, all included in my membership, felt like a saving grace.

Cons:

• Overwhelming Features: The vast array of features, while impressive, was sometimes too much to take in.

• Disruptive Interface: The interface occasionally hindered my writing flow, making me grapple with numerous clicks and choices.

• Canvas Personalization: The Canvas, although full of potential, lacked the tailored structure I desired.

• No Plagiarism Checker: This missing feature left me doubting the originality of my content.

In conclusion, Sudowrite proved to be both an inspirational guide and a challenging enigma. It undeniably elevated my creativity, but its intricacies sometimes acted as obstacles. Still, it forced me to think differently, enhancing my writing with every session. I'm optimistic about its future upgrades, hoping for a more user-friendly interface and bespoke organizational tools. For now, I journey with Sudowrite, accepting its imperfections and rejoicing in its successes, all with the aim of crafting narratives that resonate deeply.

Tech Trials: TextFX

We used Googles new free to use TextFX Software with the same prompt for each tool: “AI in Education”

Simile (Create a simile about a thing or phrase)

• AI in education is like a flashlight in a dark room, illuminating the path to new learning opportunities.

• AI in education is like a new tool in a teacher's toolbox, providing new ways to engage students and personalize learning.

• AI in education is like a personal tutor who can adapt to the individual needs of each student, providing the perfect level of support and guidance.

Unexpect (Make a scene more unexpected or imaginative)

• AI in education that is powered by a tiny hamster wheel

• An AI program that helps children solve math problems using interpretive dance

• AI in education that is delivered via a sentient Roomba vacuum cleaner

POV (Evaluate a topic through different points of view)

• AI in education will create a new class of digital teaching assistants that will help students learn at their own pace.

• AI in education will help to identify and close learning gaps, ensuring that all students have the opportunity to succeed.

• AI in education will create a more equitable learning environment, by providing all students with access to the same high-quality educational resources.

Explore the many other tools within this platform such as acronyms, alliterations, story scenes, and more!

Deep Learning (or: Assorted Links)

• Visit the University of Guelph’s OTL site for all things AI and teaching and learning: Teaching in the Context of AI | Office of Teaching and Learning (uoguelph.ca) 

• The AI Ate My Homework, and My Schoolwork Too 

• One Useful Thing | Ethan Mollick | Substack - if you ignore everything else, don’t ignore this. The OG of practical generative AI guidance in the university classroom, Ethan Mollick, at the Wharton School at Penn, writes a regular (and free) newsletter stuffed with interesting and innovative ways of deploying AI tools in the service of elevating higher education. Mollick’s work has the highest learning per minute rate in this space. 

• How colleges will change exams and assessments to stop use of AI to cheat | Independent.ie 

• Generative AI exists because of the transformer - a really fantastic explainer from the Financial Times’ Madhumita Murgia, fantastic visuals

• AI data training companies like Scale AI are hiring poets - Rest of World 

• GitHub - microsoft/prompts-for-edu - very useful GitHub repository from Microsoft for prompts 

• Announcing ‘Be My AI,’ Soon Available for Hundreds of Thousands of Be My Eyes Users 

• Spotify’s AI Voice Translation Pilot Means Your Favorite Podcasters Might Be Heard in Your Native Language — Spotify

• What Builders Talk About When They Talk About AI | Andreessen Horowitz (a16z.com) 

• Simulating History with ChatGPT - by Benjamin Breen (substack.com) 

• Welcome to State of AI Report 2023 

Future State Estimation (or: What’s next for the AI Literacy Lab)

• Podcasts: If there can be, to date, over 300 episodes dedicated to Whatever Happened to Pizza at McDonald's, then surely there is room for a podcast on the practical application of generative AI in higher education. Slated to launch in Fall 2024, this interview series will feature experts, educators, and AI researchers discussing the practical intersections of AI and education.

• Publications: We will pulling together our research, findings, and innovations from Fall 2023, and we're eager to share these with the broader academic community. We will share some of these in our next newsletter, set to be released in early 2024.

• Website Launch: To offer a centralized platform for our initiatives, resources, and updates, we are in the process of designing a dedicated website. Set to go live in Winter 2024, this platform will serve as a hub for all things AI in academia.