AI Literacy and Generative AI for Educators

AI Literacy and Generative AI for Educators

Michael Myers | Digital Literacies Librarian | michael.myers1@senecapolytechnic.ca

Agenda

Concerns & Considerations

Sandbox

AI Tools

What is it?

Academic Integrity

Copilot

How does it work?

Evaluating Outputs

Resources

What is it?

The difference between GenAI and AI

Definition

Generative AI (GenAI) is an Artificial Intelligence (AI) technology that automatically generates content in response to prompts written in natural language conversational interfaces. Rather than simply curating existing webpages, by drawing on existing content, GenAI actually produces new content.GenAI is trained using data collected from webpages, social media conversations and other online media. It generates its content by statistically analysing the distributions of words, pixels or other elements in the data that it has ingested and identifying and repeating common patterns (for example, which words typically follow which other words).

UNESCO. Guidance for generative AI in education and research. https://unesdoc.unesco.org/ark:/48223/pf0000386693.locale=en

How does it work?

Algorithms, artificial neural networks (ANNs), Generative Pre-trained Transformers (GPTs), and Large Language Models (LLMs)

Machine Learning

Algorithms

Artificial Neural Networks (ANNs)

Generative Pre-trained Transformer (GPT)

Large Language Models (LLMs)

GPT Functions

Once the GPT has been trained, generating a text response to a prompt involves the following steps:

Tokens

Converted

Patterns

Guardrails

Repeat

Post-Processing

Concerns & Considerations

An ethical conversation

Opportunities

Challenges

• Will make up content and references that are not correct
• Often lacks the latest information and research
• Covers some topics more than others
• Does not cite where the information is coming from
• Can contain built in bias

• Brainstorming ideas
• Narrowing your search topic and identifying keywords
• Summarizing and outlining
• Translating text
• Writing computer codes
• Getting inspired
• Asking questions

More Challenges

Privacy

Hallucinations

Deepfakes

Copyright

Biases

Environmental

Academic Integrity

Citing and disclosing

FAQs

Can I use generative AI in my assignment?

Do I need to cite content generated by artificial intelligence?

What if I use AI to come up with ideas or an outline for my assignment?

Are there risks to using AI content as a source?

APA Citation

MLA Citation

Disclosure Statments

Tool

Purpose

Microsoft's Copilot was used to generate research ideas and questions for this assignment using the following prompt: "Study existing research papers, environmental data sets, and other academic research information. Identify emerging trends and areas of growth for the non-profit organizations in Canada. Suggest some research questions for non-profit admins to use who are looking to explore opportunities for their organization to grow."

Prompt

Evaluating Outputs

Using information literacy skills with AI tools

Currency Relevance Authority Accuracy Purpose

Things to look for

Redundancy

Devoid of specifics

Overuse of cliches

Extra "flowery" language

Overcomplicated wordchoice

Many many emojis

Tell Tale Signs

AI Tools

An exploration of a few tools

Napkin AI

Scribe

NotebookLM

Copilot

What is Copilot?

• Microsoft introduced Microsoft Copilot on February 7, 2023
• This chatbot, which is based on a large language model, has the ability to reference sources, compose poems, write songs, generate images and more
• It serves as the main replacement for the now-discontinued Cortana, and was initially launched as Bing Chat, it was integrated into Microsoft Bing and Microsoft Edge
• Copilot employs the Microsoft Prometheus model, which is built on OpenAI's GPT-4 foundational large language model

• The chatbot's conversational style is similar to that of ChatGPT, and it can communicate in a variety of languages and dialects

How do I access it?

Embedded in Microsoft Edge

Available as a website

Available as an app

Q&A

Thank you very much!

library.senecapolytechnic.ca/sandbox/

michael.myers1@senecapolytechnic.ca

Resources

For deeper learning

Samokishyn, M. (2023). Algorithmic awareness toolkit: Teaching algorithmic literacy in academic libraries and beyond. Instructors guide. https://pressbooks.bccampus.ca/algorithmicawarenesstoolkit/

United Nations Educational, Scientific and Cultural Organization. (2023). Guidance for generative AI in education and research. UNESCO. Retrieved April 8th, 2023 from https://unesdoc.unesco.org/ark:/48223/pf0000386693.locale=en

Museum of Science. (2022, March 29). What is AI? [video]. YouTube. https://www.youtube.com/watch?v=NbEbs6I3eLw

KI-Campus. (2023, September 25). Generative AI Explained in 2 Minutes[video]. YouTube. https://www.youtube.com/watch?v=rwF-X5STYks&list=TLGG-6RYWXR09JIyOTA0MjAyNA&t=6s

Privacy International. (2015, May 12). Big Data | Privacy International [video]. YouTube. https://www.youtube.com/watch?v=HOoKhnvoYkU

BBC News. (2023, June 18). Can artists protect their work from AI? – BBC News [video]. YouTube. https://www.youtube.com/watch?v=DUwDqSY8StE

TED. (2023, May 19). The Incredible Creativity of Deepfakes — and the Worrying Future of AI | Tom Graham [video]. YouTube. https://www.youtube.com/watch?v=SHSmo72oVao

IBM Technology. (2023, April 20). Why Large Language Models Hallucinate [video]. YouTube. https://www.youtube.com/watch?v=cfqtFvWOfg0

ILLC. (2023, November 20). Ethical issues with Generative AI [video]. YouTube. https://www.youtube.com/watch?v=d5uuuW31-AQ

CrashCourse. (2019, December 13). Algorithmic Bias and Fairness: Crash Course AI #18 [video]. YouTube. https://www.youtube.com/watch?v=gV0_raKR2UQ

Federal Court of Canada. (2023, December 20). NOTICE TO THE PARTIES AND THE PROFESSION The Use of Artificial Intelligence in Court Proceedings. https://www.fct-cf.gc.ca/Content/assets/pdf/base/2023-12-20-notice-use-of-ai-in-court-proceedings.pdf

UNESCO. (2023, February 6). Ethics of AI: Challenges and Governance [video]. YouTube. https://www.youtube.com/watch?v=VqFqWIqOB1g

TED. (2023, November 6). AI Is Dangerous, but Not for the Reasons You Think | Sasha Luccioni [video]. YouTube. https://www.youtube.com/watch?v=eXdVDhOGqoE

Peters, M. (2023). Utilisation transparente de l’intelligence artificielle. Martine Peters, chercheuse et professeure en science de l'éducation Directrice du Partenariat universitaire sur la prévention du plagiat. https://mpeters.uqo.ca/logos-ia-fr-peters-2023/

Relevance

• Does the information meet your needs?
• Does it relate directly to your research question or topic?
• Who is the intended audience?
• Is the depth of information appropriate (not too basic or too advanced)?

Ask yourself: “Is this source useful for my purpose or assignment?”

YES Remember to include a citation whenever you incorporate content generated by AI in your assignment. For example, if you paraphrased or quoted text generated by ChatGPT, cite the source both in text and in your references list or works cited. Neglecting to cite this material may constitute an academic integrity offence. Since your reader can't access the exact AI-generated text, you may also consider including a copy of the text as an appendix, or as part of supplemental materials.

As part of using artificial intelligence responsibly and ethically, it is important to be transparent and acknowledge how you have used AI in your work. If you used AI to generate ideas or an outline for your assignment, include a statement in your work indicating which AI tools you used and how you used them. Neglecting to include this statement may constitute an academic integrity offence.

Large language models (LLMs) like chatGPT can generate authoritative-sounding prose on many topics and domains, they are also prone to just "make stuff up". Literally plausible sounding nonsense! In this video, Martin Keen explains the different types of "LLMs hallucinations", why they happen, and ends with recommending steps that you, as a LLM user, can take to minimize their occurrence.

Find Better Resources

Generative AI content has its benefits: it is great for brainstorming some ideas, crafting research questions, identifying keywords for searching, creating a possible outline, etc. However, Generative AI should not be used as a source for research. Generative AI will make information up and will not offer its references or search strategies, so if the generated content offers stats, figures, facts, or analysis be sure to investigate them as outlined in step two. If during your investigation you determine that claims from generated AI content are correct, be sure to use the verified peer reviewed research you found to reinforce these facts and claims rather than the generated AI content claims. Discover a world of verified research and content through the Seneca Libraries!

Review: Peer Reviewed Articles

There are five common types of algorithmic bias we should pay attention to: data that reflects existing biases, unbalanced classes in training data, data that doesn't capture the right value, data that is amplified by feedback loops, and malicious data.

Caution

The overuse of images can dilute a message. When too many visuals compete for attention, they can overwhelm the viewer and obscure the narrative. This is particularly problematic in business communication, where clarity and conciseness are paramount. To avoid this, images should be used strategically—supporting rather than dominating the message.

Accuracy

• Is the information reliable and correct?
• Is it supported by evidence?
• Are there spelling, grammar, or factual errors?
• Can you verify the claims with other sources?

Ask yourself: “Is this information backed up by trustworthy sources?”

AI, Machine Learning, Deep Learning and Generative AI Explained IBM Technology

AI-generated media that looks and sounds exactly like the real world will soon permeate our lives. How should we prepare for it? AI developer Tom Graham discusses the extraordinary power of this rapidly advancing technology, demoing cutting-edge examples -- including real-time face swaps and voice cloning -- live from the TED stage.

Trace to the Source

It is very important to get the whole picture and understand the context of which the information and/or media was originally created in. When you trace the claim, quote, or media back to the source, you can see it in it’s original context and get a sense if the version you saw in your generated AI content was accurately presented. Be sure to use your critical thinking and evaluation skills to judge if a piece of information is being misrepresented to support a point

Review: Critical Thinking Readiness

Investigate

If the generated content includes specific claims, data, or facts, always fact-check them using external sources. Verify the accuracy of the information generated by the AI model.Use trusted fact-checking websites, reference materials, or subject matter experts to validate the claims made in the generated content. Evaluate the quality and relevance of the generated content. Does it align with your expectations and objectives? If you don't have prior information about the generative AI model used or its training data, take steps to gather more information. Research the model's training data sources, the organization behind it, and any available performance metrics. Determine if the AI model has been fine-tuned for a specific domain or task. Models fine-tuned on reliable datasets for specific applications may be more trustworthy.

Library Resources

Be aware that there is a possibility that the content AI generates is biased, outdated, or false. Always evaluate the information generated by AI tools. Students are responsible for the accuracy, factuality, and truthfulness of artificial intelligence content used in assignments.

STOP!

The first move is the simplest. STOP reminds you of two things. First, when you generate content using a generative AI model, take a moment to STOP and assess the quality and credibility of the output. Second, ask yourself whether you know the model or source of the information. Is this AI model reputable, and has it been trained on reliable data sources? Consider the reputation of the AI model itself. Is it known for producing accurate and coherent text, or is it prone to generating misleading or nonsensical information

The output of AI systems is built upon the output of creators, AIs are trained on millions of human-made images. BBC Click’s Ben Derico investigates how artists can protect their work from the growing power of generative AI.

Always check with your instructor first! Before starting your assignment, confirm with your instructor that you are permitted to use generative AI tools such as ChatGPT. Unauthorized use of generative software may constitute an academic integrity offence.

Purpose

• Why was this information created?
• Is it to inform, teach, sell, entertain, or persuade?
• Is there bias or a particular agenda?
• Are there advertisements or sponsored content?

Ask yourself: “Is the source trying to inform me or influence me?”

Currency

• Is the information up to date?
• When was it published or last updated?
• Is the topic time-sensitive (e.g., medical, tech, news)?
• Are the links functional and current?

Ask yourself: “Is this the most recent information available on this topic?”

As the demand for artificial intelligence grows, so does its thirst for one of our planet’s most precious resources. Andrew Chang explores how the supercomputers that run AI stay cool — by using massive amounts of fresh, clean drinking water.

There are many facets of privacy issues associated with use of AI and how Big Data and algorithms shape our information landscape. Explore and learn more about the privacy concerns around Big Data and AI.

Authority

• Who is the author or source?
• What are the author’s credentials or affiliations?
• Is the publisher reputable?
• Is there contact information or an “About” page?

Ask yourself: “Can I trust the person or organization behind this information?”