AI Explorations

As AI becomes more integrated into the post-secondary experience, it’s important that instructors acquire knowledge of the capabilities of AI and skill in the available tools that could be utilized by students. Focusing on exploring, playing, sharing, asking questions, and learning about artificial intelligence and its impact on post-second education is key as you build a deeper understanding of this topic.

The following ‘Explorations’ will focus on teaching and learning components and challenge you with tasks to complete to become more familiar with specific tools. The last 5 explorations align with one of 6 AI laneways that are identified in the NIC AI Guidelines.

Video Explanation (6 mins)

Ethan Mollick’s book, “Co-Intelligence: Living and Working with AI” was published in April 2024 and has been respected as an excellent starter book in understanding AI. It was the chosen book back in May for Canadian higher ed educators and teaching leaders to read and explore via a pan-Canadian exploration group. The NIC library has hard copies in all of the campuses for signing out. This 4-5 hour breezy read is highly recommended. It isn’t education specific but has separate chapters on various potential uses of AI with examples he shares. He writes in very accessible language.

Ethan is often considered a balanced, but optimistic post-secondary educator who (along with his wife) are at the forefront of exploring how AI can impact education, work and life – along with the pitfalls, privacy, security and other uncertain components still being uncovered. He does a lot of daily work behind the scenes in exploring AI. He is a business professor at University of Pennsylvania. He writes a blog post on Substack, has a couple of websites sharing his work and often posts on LinkedIn.

His two websites: One Useful Thing and More Useful Things are excellent sites to read his latest blog post (One Useful Thing) or check out his Prompt Library and other resources (More Useful Things). You will be asked to subscribe to his Substack or upgrade to a paid version- you can agree to freely receive his posts in your email or ignore pop up box.

To kick start your AI Exploration read Ethan’s latest (100th) post on his Substack called, “Thinking Like an AI“- A little intuition can help (October 20, 2024). In simple language he describes how AI works. Length: 2500 words – may take you anywhere from 6 minutes to 16 minutes to read.

After you have read the article, try out one or two of his pre-written prompts: for Instructors or for Students. He indicates which of the most common GPTs (Generative Pre-trained Transformers) work best with each prompt (ChatGPT, Bing/Copilot, Gemini or Claude).

Ethan has essentially created ‘text’ for you to just copy and paste into a GenAI platform with all the key terms to have it turned into an assistant, a tutor, a quiz creator, a simulation creator, a lesson crafter etc. This is often where people struggle the most – writing good instructions for GPT models. There are full length courses just on how to ‘write a prompt’ but Ethan gives you quite a number that you can use and they work!

No need to purchase anything – free accounts will do for most of his prompts (paid versions have more features and capacity but let’s leave that for another day – they are often labelled as “Advanced”).

If you don’t wish to create a free account use Microsoft Copilot which you have with your NIC account or ChatGPT. Don’t put any personal information, student work, confidential information into GPTs!

But: sometime it might be good to get yourself a set of free accounts of the three other GPTs (chat environments) for our explorations – as each works differently and it might be good learning to see what one can and can’t do.

  • ChatGPT or GPT4 (from company OpenAI – free no account version, or look to top right and create a free account)
  • Claude (from company Anthropic – free with account you create with email)
  • Gemini (from Google, used to be called Bard – free version with your Gmail/Google account when you sign in).
  • Copilot (from company Microsoft, used to be called Bing Chat – free version with your NIC account when you are signed in)

So for this exploration let’s try this and follow these protocols:

  • Take 6-16 minutes to read: Thinking Like an AI
  • Explore some of the pre-written prompts (Instructor and Student ones) in Copilot or other GPTs from More Useful Things: Prompt Library – by copying the prompt into the first chat box, allowing it to think and then follow its responses as if you were using it as an instructor or student (this takes minutes to do but you may wish to follow the prompts and try out in a variety of GPTs or different prompts to see differences)

Many moons ago as a university professor, Liesel Knaack did research on “learning objects” (assessment of them, how to use them in learning process etc.) and taught an advanced instructional design elective building learning objects. Learning objects are interactive educational tools or collection of content/practice/assessment items that are freely accessible to anyone to share and reuse in many courses.

Her students would use Adobe Flash (a former multimedia platform) to build these learning objects/resources that they couldn’t find online or that the textbook or static materials prevented learners from fully understanding a concept. It took them nine weeks of class time and homework to build a learning object that functioned properly. Concepts such as balancing equations in chemistry or understanding how a dangerous simulation might unfold, how to prevent plagiarism or a physics law that required students to input various values to see the outcome.

Her students did such a great job that she took them to the MERLOT (multimedia sharing platform) conference in Ottawa one year and they presented their products – see some of their examples within slide deck. This was the beginning of her passion for creating shareable, reusable, remixable learning tools that eventually would expand into Open Educational Resources (OERs). 

Skip forward two decades and we see learning object examples such as simulations developed by the University of Colorado to help students understand physics – or students explore math calculations or engage in simulations and virtual labs.

While the web now has many of these learning objects available, there are smaller interactive components that teachers look to include in their courses to help students learn or they just want something they can’t find online (a tool or a demo). Or teachers want some smaller components automated or created for engagement and practice by students.

Her interest in these interactive components remained evident as she shared how to use H5P (HTML 5 Package) in building interactive activities (quiz questions, matching games, content design, engaging presentations etc.) through various workshops held in 2024 at NIC. 

But recently Claude has the ability to create “artifacts”. An artifact can be more than a H5P interactive component – it could be document, snippets of code, a whole webpage (HTML code) or it can be an image or diagram – but what Liesel liked were the interactive “React” components like learning objects.

Here is a simple one she made to calculate reading time based on words. Kind of like a ‘calculator’ where it takes the research on reading time (which she confirmed is correct) and makes a little input device to approximate reading time for assignments or readings.

Here is another really quick (but kinda hard) interactive Claude Artifact she made – called a Generative AI Matching Game – you click on two cards and try to find the definition and term and match them. When you get them correct they stay turned over. You do the game until you match all terms and definitions. 

But how did she get there?

She asked (free) Claude with this prompt:
Hi Claude. Can you create a web tool that acts like a matching game of cards. There are 16 cards with 8 pairs that match. The topic is generative AI definitions. One of the cards in a pair is the word and another card is the definition. The player has to click and turn over only two cards at once and if they are correct pair (definition and word) and the player gets a point. If the cards are not a pair then they are turned back over in the same spot and the next player tries.

Claude responded with:
I’ll create an interactive card matching game focused on generative AI terms using React. This will include card flipping animations, score tracking, and the matching logic you described. And then in the window to the right it started doing all the coding. A minute later it built the matching game. I did nothing but watch!

Then she got this from Claude:
I’ve created a React component that implements the Generative AI Matching Game you described. Here’s a breakdown of the game features:

  1. The game consists of 16 cards (8 pairs) with generative AI terms and their definitions.
  2. Cards are shuffled at the start of the game and displayed in a 4×4 grid.
  3. Players can flip two cards at a time by clicking on them.
  4. If the flipped cards match (a term and its definition), they remain face up, and the player scores a point.
  5. If the cards don’t match, they flip back after a short delay.
  6. The game tracks the player’s score.
  7. There’s a “Shuffle and Restart” button to reset the game at any time.
  8. When all pairs are matched, a congratulatory message appears.

The component uses some UI components from the shadcn/ui library (Button and Card) for styling. It also uses the Shuffle icon from the lucide-react library for the reset button. To use this component, you’ll need to ensure that the necessary dependencies are installed and that you have the shadcn/ui components set up in your project. Would you like me to explain any part of the code in more detail or make any modifications to the game?

She saw the game, played with it a few times and then saw the ability to publish it – upon publishing she got a link to share with others. She could put it into her course or share it on a website.

It isn’t fancy – but it is a learning object she built with a simple prompt and no coding – in minutes! Remember: there could be errors or hallucinations!

TRY building an artifact in Claude – something more adventuresome than a matching card game please! Publish and share your link to your artifact so we can try it out. Share how you might find it helpful. Share what you learned? Here are some ideas.

NOTE: Free version of Claude restricts you to 7 prompts (I think) until it won’t do any more until you purchase next version. So think of your prompts or start a new conversation if you run out of “tokens” in a given conversation.

Resources

Google NotebookLM is an AI tool from Google (that was released over a year ago) – that can analyze up to 50 sources (website links, PDFs, YouTube videos) and extract themes, key messages and important points. It doesn’t do research on the web – it focuses solely on the sources you upload. It is an AI notebook – a handy dandy research assistant!

Just this past month a new feature was added to Google NotebookLM and puts this in a whole new category of research support. 

The biggest feature (that caused jaw dropping responses when it came out) is the creation of a PODCAST (with two realistic voices) chatting about the content you have uploaded.  Seriously. You can take exciting, dry or dense content and upload it and the podcast will generate (it takes a few mins to create podcast) a lively discussion about that content. You gotta try it! Time Magazine called it one of the best inventions of 2024. 

Other features include the ability to combine images, PDFs, and other content into instant research papers, curriculum for a course, etc. It also takes content in a variety of formats – and it references where the details came from along with table of contents, briefing documents and study guides! That’s the research assistant part – so if you can get past the podcast feature there is much more here to explore. 

Explore how it can build a research paper, organize it and build a final product. It is a time saving tool for busy academics!

Try it out: https://notebooklm.google/ (you need a Google account)

Recommended Skims

Your turn! Try out Google NotebookLM – and from an educational perspective of helping you with organizing papers, creating outlines, theming content from any source you upload (and the podcast feature) – HOW could students use this in an assignment (as an allowable tool to use) and how could YOU (as an educator, employee) use this for your work in preparing teaching and learning activities?

Well, we can’t go much further in our AI explorations without giving some airtime to the creative aspect of image generation (text to image) – but with a healthy dose of coming up against deepfakes!

This week take the deepfake quiz (to see if you can spot the real pictures!) and then go out exploring and create some realistic and/or fantastical images to use in your course.

Remember: All free accounts will count “tokens” (processing power to build images) and after a couple of pictures may cut you off until 24 hours have passed. So, think about your prompt before you go mucking about. You can get a few decent images on all free accounts with the platforms below.

FIRST – TAKE QUIZ:

Take the deepfake 10 question quiz to see if you can spot the deepfakes (real looking pictures that were computer generated) at Leon Furze’s site.

SECOND – SKIM ARTICLE:

This article (with accordion expanding showing same images tested in each platform) article from ZDNET on their reviews of image generation GPTs – within each review of the platform there is a link to another article with steps on how to use the platform.

THIRD – EXPLORE:

AI Image Generators (some of them)

Open AI’s Dall-E 3 – Website

  • CNET Review Page
  • Cost: $20 a month with ChatGPT Plus Subscription
  • ChatGPT Free Version: 2 images in a 24-hour period

Google’s ImageFX – Website

  • Through Google Account
  • ZDNET Review and Instructions to Use
  • Can access this through Gemini chatbot

Leonardo AI – Website

  • CNET Review of Leonardo AI
  • Cost: Free and premium plans
  • 150 tokens that resets daily
  • Options for size and number of images per each generation

Microsoft Designer’s AI Image Generator – Website

  • Cost: Free version with NIC Microsoft license
  • ZDNET Review and instructions

Adobe Firefly – Website

Midjourney – Website

  • No free account – starts at 10$ US/month
  • Very high-quality images

Canva – Website

  • Cost: Free and paid account
  • Good basic image generate for beginners

Video and audio is an untapped potential in post-secondary education but has been challenging to access due to the learning curve and the need for sophisticated software and hardware resources required to engage students with the platforms and their outputs. NO MORE!

Liesel Knaack was so inspired by these platforms that she took a year off work (eons ago) to complete a post-graduate certificate in multi-media at college. She got an immersive learning experience shooting video with production level cameras, doing audio production in a studio and creating a couple of interactive educational resources for learners. 

In her early days of post-secondary teaching she had her students making videos, creating podcasts and developing interactive assignments – always full of creativity and fun in learning. In mid 2000’s she got into Jing (now called TechSmith Capture) – a lightweight (and free) video recording tool (like Kaltura Capture which you have access in Brightspace and NIC MediaSpace)  to record summaries for her classes, create instructional videos and bring content to life. Here are other uses of video.

She sees many uses of text to video generation for instructors and especially for those who create training materials, presentations, introductory videos – as you can create studio-quality videos with AI avatars and voiceovers in many languages – and as easy as PowerPoint. If you don’t have a video-quality voice or image or don’t speak many languages or don’t want to waste hours cutting videos – then check this out!

Here is what Liesel made in 20 minutes with Synthesia: MOST of the text was provided but she zhuzhed it up a bit! (or accessed from host)

She set up a free account with Synthesia (easier to use than other popular product, HeyGen). Upon account set up, she shared what she was wanting to do (create training videos, promote a topic). 

Then for the project topic it asked her what she wanted to create. She said “create a video outlining the possibilities with generative AI text to video for teaching and learning”. 

Within seconds it created the slides, the text, the images, the builds, the transitions and the default avatar. She changed up the text, added another slide, repositioned the avatar, previewed what each narration would sound like, took off the logo and generated the 2.5 minute video in 9 minutes. She downloaded and uploaded to Kaltura Video. You don’t need to do that – you can just “publish” it and it will store on their site with a URL.

With paid accounts, you could use over 180 avatars, have up to 360 minutes of video per year – and create your own digital twin (your own avatar of you!).The free account is only 3 minutes in total per month with 9 AI avatar choices and 2 stock personal avatars – but you can access the 140 languages and voices!

Overview for training and learnin – Give it a try!

If you are not into the video aspect but want to try text to audio – Eleven Labs is a great starting spot with a free account to create something like this short audio clip or this sound effect Liesel made within minutes! Eleven Labs is a voice generator that can produce content in 32 languages – great for any voice overs, creating an audio book or producing content for a course that is in a format that students can listen to – like a module audio book!

Explore
1. Create a video from text with avatar: Use Synthesia (or HeyGen or other platforms) and/or
2. Create an audio clip from text: Use Eleven Labs or any other suggested platforms like Speechify, Altered and Murf
3. Reflect on how you might use the end product in teaching and learning situations

As the enormity of AI integration in our everyday lives becomes more prominent and accessible we need to consider how it can be tailored to benefit our students as they progress through higher education studies. Through the adoption of a 6-Lane AI Assessment Scale here at NIC we can see many areas where strategic, intentional integration of AI tools can assist learners.

This week’s post and challenge will focus on Lane 2: AI as a Study Tool. There are many ways that AI can be used as a study tool, specifically this week we are going to explore the practice of tutoring.

Who has not benefited from tutoring in their own personal learning paths? Many breakthrough moments of understanding are in deep conversation with experts on a particular topic.

How often do your students have these opportunities? It’s safe to guess that an expert is not readily available to interact with each of your students to provide unending attention and patience in coaching them along their path of understanding. Most likely, you end up filling this role. An unattainable goal this is for an instructor to individually tutor each one of their students while also teaching the class. The next logical place to look for this assistance was beyond our classrooms to peer tutor centers and sometimes privately paid tutors. Up until recently, a physical tutor was the only option available along with a minefield of limited tutor numbers available, limited expertise, restricted time slots and (sometimes) high cost. How are we as educators to provide every student the same opportunities as Benjamin Bloom identified back in 1984 when he documented that tutored students achieved two grade point averages higher than their un-tutored peers? Enter AI tutoring.

AI can assist students in the role of a tutor at any time of the day, anywhere geographically for zero cost. I agree, it is not the same as a human personal tutor. But honestly, isn’t this step better than no tutor at all?

Below are 4 different options to get you and your students moving with AI as a tutoring study tool:

#1 Create your own tutoring prompt. The AI must understand that it is an expert in the topic and explicitly describe tutoring. One of the exhausting parts of making the AI do this task is to frame the expectation of the tutoring experience as full as possible. The longer the prompt, the better. Try using some of these:

  • Ask me questions
  • Do not immediately provide an answer
  • If I respond incorrectly ask me why I think that is the answer
  • Do not provide the entire process in one response. Break it down and only provide one part at a time after I have had a chance to respond
  • Explain tricky concepts in two different ways
  • Use analogies or metaphors in the explanations
  • Gauge my baseline understanding by asking 5 questions prior to starting the tutoring

#2 Pre-crafted Tutoring prompt. Why develop your own if someone has already done it for you? For instance, the most recent AI tutor prompt from Ethan Mollick is an astonishing 791 words in length! Yeow. Ethan’s website More Useful Things: AI Resources contains a handy prompt library with the aforementioned behemoth of a prompt and many, many more equal in detail. This is well worth the time to simply explore.

#3 Real-time audio/visual AI tutoring. AI tutoring through text may be an excellent format for topics that can be easily expressed in written form, but what about physics, mathematics, or diagrams? Perhaps the student has spent 2hrs on a process or calculation and cannot understand where they are going wrong. What would a tutor typically do in this situation?  Perhaps they would start by reviewing the student’s work.

The newest release of Gemini can access the microphone & webcam of a computer and analyze what it sees and hears in real-time. If the application is used on a smartphone the functionality is reminiscent of sharing one’s surroundings with a friend on Facetime.  Stream Realtime | Google AI Studio

In addition, Gemini now has built-in tutors for specific topics such as mathematics all ready to go.

#4 Free AI Tutor Platform. Here is yet another option, use AI Tutor Pro from Ontario-based Contact Nord. No need to provide detailed prompts, it has already been designed for tutoring. Students can choose to type in a subject or upload content about a concept or topic they are learning. Free.

AI Tutor Options for $$$:

Chat GPT has released an education-specific version of its GPT-4o platform that can be purchased by post-secondary education providers. Perhaps this is the direction higher education will go. See the site here: OpenAI for Education | OpenAI.

Khan Academy has released a tutoring-specific application for anyone. Some school districts in the USA are piloting its use in assisting student. This currently requires a fee to access in Canada. See the site here: Khanmigo.

How do you see these resources being used in your course? How can this resource help empower your students to learn more deeply? Try it out for yourself…

  1. Choose one method outlined above to interact with an AI tutor.
  2. Choose a topic
  3. Experience the AI tutor process

Resources:

• From Khan Academy: Meet Khanmigo: Khan Academy’s AI-powered teaching assistant & tutor
• Blog Post: How AI Could Save (Not Destroy) Education | Sal Khan | TED
• TeachAnywhere: Prompts, Tutors and Assistants

Overview

In August 2024, Marc Watkins published a sub stack called “first drafts in the AI era.” Watkins emphasized the importance of students using original thought to draft an outline when writing rather than relying on AI to help.  He highlighted the value  of wrestling with initial thoughts and organizing concepts as crucial to the writing experience. Watkins was not opposed to AI; rather he argued for writers to generate their own ideas first and then use AI to enhance their work. 

Fast forward to today. AI tools are increasingly used as collaborative partners in the writing process, and yes, even generating ideas for that first draft. AI can brainstorm ideas, provide inspiration and generate multiple perspectives for almost anything, but the human touch remains crucial for authenticity and originality.

In this exploration, we are exploring  Lane 3: AI as an idea generator for instructors in post-secondary education. It can be used to spark creativity and provide varying perspectives to enhance the teaching and learning experience.

Across disciplines such as healthcare, math/science, and trades programs, AI can generate ideas in meaningful ways.

Healthcare

Case Studies Creation: Generate complex, realistic patient case studies for simulation exercises, tailoring cases to specific learning objectives (e.g., ethical dilemmas, unusual symptoms)

Examples: Springer publishing offers 7 innovative ways to engage a nursing class with AI. Many of these ideas could be adapted across other disciplines.

Math/Science

Experiment Ideas: Develop unique lab experiments for physics, chemistry, or biology, such as ecological simulations or chemical reaction predictions.

Example: An engineering prof at U of Calgary uses AI in the classroom to support students to generate their own ideas.

Trades

Training Simulations: Design virtual or theoretical project ideas, such as creating blueprints for construction or generating simulation ideas.

Example: A trades instructor could use AI to simulate a hazardous scene and collaborate on how best to manage the situation.

In all these cases, the key takeaway is that AI is a tool to support human creativity and critical thinking, not a replacement for it. Whether in healthcare, the sciences, trades, or other fields, instructors can use AI to enhance their teaching, generate new ideas, and bring their work to the next level to support quality student learning.

How?

Here are three AI tools that can be used to help generate ideas beyond Chat GPT and Copilot:

Gamma AI: https://gamma.app/

Free/paid subscription. Can be used to generate ideas for presentations or images. You can import files or URL’s and ask Gamma to enhance or generate.

This one is a game changer when it comes to presentations.

When nursing instructor, Rachel Goodliffe, asked Gamma to generate an idea for a presentation focused on innovative ideas in education she received this:  Empowering-the-Future-Educational-Innovations.pdf

Mymap.AI: https://www.mymap.ai/idea-generator

Free/paid subscription. This is a fun tool that can help to generate flow charts, timelines, and mind maps. You can type in text, or a link/file and the tool will generate content which can then be refined but it gives you a start. Nursing instructor, Rachel Goodliffe, asked Mymap.AI to generate the steps to administering an intramuscular medication into the deltoid muscle ( she has a nursing background if you did not guess already). Once you input information it tells you “I’m thinking….”

While Rachel knows this is too small, she hopes you can see this impressive artefact with information embedded. She will say that information is not as comprehensive as she would provide to her students and is missing safety pieces such as cleaning the skin before you inject, but as a place to begin – not bad.

https://ideanote.io/idea-generator

Ideanote claims to be the “best online tool to quickly generate ideas” and is free. It does ask you to input a company name, so we just added initials. You can only generate 3 ideas in 24 hours, so you must be mindful of that part. However, when we asked ideanote to help us with ideas for a project we’re working on, it came up with some strong suggestions. Give it a try!

By now we’ re starting to think that any AI tool has capacity to generate ideas and if that is your mandate then you are spoiled for choice. Chat GPT or Copilot can all generate ideas to get you started on projects or tasks, it just depends on what you need.

We decided to ask Copilot about this:

Prompt:

Are all Gen AI resources idea generators? What do you think?

Sent by Copilot:

Not all generative AI resources are strictly idea generators, though many do excel in that area. Generative AI can serve a variety of purposes, including:

  1. Content Creation: Writing articles, stories, and even poetry.
  2. Image Generation: Creating artwork, designs, and other visual content.
  3. Data Analysis: Generating insights from large datasets.
  4. Language Translation: Translating text between different languages.
  5. Code Generation: Assisting with programming by generating code snippets.

So, while idea generation is a significant aspect, generative AI’s capabilities extend far beyond that.

REflect:
  1. What have you used in the classroom or for your personal teaching  as an AI idea generator?
  2. How did it work out?
Resources:

This exploration focuses on Lane 4: AI as editor. AI tools can help students improve their writing by checking grammar, suggesting better word choices, and making their ideas clearer. Editing tools also help students catch mistakes and fix them before turning in their work.

Throughout post-secondary education experiences, many of us have benefitted from colleagues, peers, and professors who provided feedback on our work. Liesel Knaack still attributes her first breakthrough as a writer during her undergraduate when her cultural geography professor provided editing suggestions and numerous opportunities to redo her work. Similarly, when she journeyed through post-grad programs, she had advisors and committees who took the time to provide feedback on her writing projects.

Her experiences are now 10 + years ago and times have changed. For one, we lived through a pandemic that significantly impacted how we study and learn together. Changes such as a shift to remote learning and using digital tools accelerated the arrival of AI editing tools. AI as an editor dovetails with AI as a tutor as both enhance students’ capabilities to refine written content and facilitate learning. AI tutors and AI editing tools have made learning and content creation more accessible and convenient for those with access to and familiarity with the technology.

So how do we use AI as an editing tool? 

Simple. 

AI editing tools can help refine and perfect written content through grammar and spell check. Tools can check for sentence structure and clarity, style, and tone. Some AI editing tools offer content suggestions, such as recommending alternative phrasing, sentences, and even entire paragraphs. Bottom line, AI editing tools assist by flagging potential issues, providing suggestions, and streamlining the editing process. 

How can instructors use AI with students’ assignments?

AI editing tools provide an opportunity for instructors to guide students not only in refining their writing but also in developing critical thinking about writing quality. Here are some ways instructors can incorporate AI editing tools into assignments:

  • Preliminary Edits: Assign students to run their drafts through AI tools like Microsoft Copilot, Grammarly, or ChatGPT before submission. Ask them to submit both the original and the AI-edited version, along with a reflection on how the edits improved their work or aligned with the assignment’s goals.
  • Revision Workshops: Encourage peer review using AI. Students can share their drafts and use AI-generated feedback as a supplement to their peers’ suggestions, promoting collaborative and technologically enhanced learning.
  • Tone and Style Exercises: Assign students a piece of their writing to analyze using AI. For instance, ask them to use Notion AI to shift the tone from formal to conversational, or refine clarity and conciseness, then discuss how those changes impact their message.
  • Grammar and Structure Analysis: Create assignments where students compare AI editing suggestions with traditional grammar guides. This can help them critically evaluate the reliability and appropriateness of AI feedback.
  • AI Limitations Discussions: Include assignments where students analyze the limitations of AI tools. For example, they can identify areas where the tool misunderstood context or offered unsuitable suggestions.

By framing assignments around these tools, instructors can teach students not only how to use AI effectively but also to approach these tools critically, ensuring they remain active participants in their learning process.

Below are 4 different options to get you and your students to use AI as an editing tool.

Microsoft Copilot is likely one of the easiest editing tools for students to use, especially if they are already familiar with Microsoft Office applications like Word. Familiarity means students don’t need to learn a new tool, they can access Copilot directly within the software they use for assignments.

Here are step-by-step instructions for using Microsoft Copilot as an editing tool in Word or other Office applications:

Step 1: Open the Document

  • Open the document you want to edit, select the text and copy into copilot.

Step 2: Select the Text to Edit

  • Highlight sections you want help with or place your cursor where you need editing assistance.
  • For document-wide edits, you don’t need to highlight anything, Copilot can analyze the whole text.

Step 3: Choose an Editing Option

  • Select from the available features, such as:
  • Grammar and Spelling Check
  • Clarity and Tone Suggestions
  • Rewriting Suggestions:
  • Apply these suggestions individually or use bulk approval options.

Step 4: Review and Apply Suggestions

  • Review each suggestion Copilot provides.
  • Accept, reject, or modify suggestions based on your preferences.

Step 5: Finalize and Save

  • Once you’ve made all edits, proofread your work to ensure it meets your expectations.
  • Save the document with a new name if you want to keep the original version as well.

Practice Using Copilot: The more you use it, the more familiar and efficient you’ll become with its capabilities!

ChatGPT can be an effective editing tool for students, offering assistance with grammar, clarity, tone, and organization. Chat GPT is available through a browser or app, making it easy to access from most devices. It can adapt to various writing styles, academic formats, and tone preferences. The tool can offer suggestions for sentence restructuring, word choice, and overall readability providing explanations for edits, helping students improve their writing skills.

How to Use ChatGPT for Editing

Prepare the Text. Upload or paste your content into ChatGPT. and provide clear instructions like:

  • “Check this for grammar and punctuation errors.”
  • “Suggest ways to improve clarity and tone.”
  • “Simplify complex sentences without losing meaning.”
  • ChatGPT will provide edits or rewritten sections. If needed, ask follow-up questions like:
  • “Why did you suggest this change?”
  • “Can you rephrase this paragraph to sound more formal?”
  • Incorporate the suggestions into your work. If a section doesn’t feel right, ask ChatGPT for alternative phrasing.
  • Double-Check Accuracy:
  • Ensure factual correctness and proper citations, as ChatGPT focuses on language rather than content accuracy.

Using ChatGPT as an editing tool is fast and cost-effective. The free version offers robust functionality, with paid options for more advanced features.

Steps to Access ChatGPT

  1. Visit ChatGPT  and create an account.
  2. Choose a plan – Free will work just fine
  3. Start a conversation and paste your text for editing!

Grammarly simplifies grammar, clarity, and tone, serving as a great tool for clean, professional writing. To get started with  Grammarly  create an account and install the application or browser extension on your computer. While you can upload entire documents, you also have the flexibility to input specific sections of text. This versatility ensures you can tailor your editing process.

Here is an example of Grammarly in Action

Let’s edit the following text to demonstrate Grammarly’s capabilities.

Mispellings and grammatical errors can effect your credibility. The same goes for misused commas, and other types of punctuation . Not only will Grammarly underline these issues in red, it will also showed you how to correctly write the sentence.

After inputting this text into Grammarly, here is what happened:

  • Red underlines indicate spelling and grammar issues, such as “Misspellings” (corrected to” Misspellings”) and “effect” (corrected to “affect”).
  • Blue underlines highlight clarity improvements, suggesting revisions to wordy phrases or awkward constructions. For example, “can possibly help you revise a wordy sentence in an effortless manner” becomes “can help you revise wordy sentences easily.”

These features not only correct errors but also empower users to refine their writing for greater clarity and precision.

NotionAI can be used as an editing tool to refine and perfect content. It can detect and correct grammatical errors, punctuation mistakes, and spelling errors. Like Grammarly, it will analyze sentence structure and clarity, style, and tone, and provide suggestions for rephrasing sentences, adding or removing content, and improving overall flow.

Getting started with Notion AI is straightforward. Here is a step-by-step guide to help you begin.

Step 1: Create a Notion Account . You will need an email address, Google account, or Apple ID to create an account.

Step 2: Enable Notion AI. Once you have a notion account, navigate to the Notion settings by clicking on the gear icon in the top right corner. Scroll down to the AI section and toggle the switch to enable Notion AI.

Step 3: Familiarize yourself with Notion AI features. After enabling Notion AI, explore the various features in the following places:

  • Page editor: Notion AI suggestions will appear as you type in the page editor.
  • Block Menu: Click the three dots next to a block to access Notion AI features, such as text generation and content suggestions.
  • Page properties: Some Notion AI features, like page organization and research assistance, can be accessed through the page properties panel.

Step 4: Experiment with Notion AI features to get a feel for how they work. Try generating text, getting content suggestions, or using the research assistance feature.

Step 5: Customize Notion AI Settings (Optional). If you want to customize Notion AI settings, such as adjusting the level of AI involvement or specifying AI models, you can do so in the Notion settings.

Exploring Notion AI as an Editing Tool

1. Familiarize Yourself with Notion AI:

  • If you haven’t already, sign up for Notion and activate the Notion AI feature.
  • Watch the introductory tutorial on how to use Notion AI for editing. (You can find this in the Notion Help Center or through a quick YouTube search).

2. Select a Writing Sample:

  • Choose a short piece of writing (approximately 200–300 words). This could be a draft of an email, a paragraph from an essay, or a piece of creative writing.
  • Paste the text into a Notion page.

3. Experiment with Editing Features:

  • Use Notion AI’s editing tools to refine your text. Specifically:
    • Grammar and Spelling Check: Identify and correct errors.
    • Clarity and Style Improvements: Use suggestions to make your writing more concise and polished.
    • Tone Adjustments: Experiment with shifting the tone (e.g., make it more formal, casual, or professional).
    • Summarization and Highlights: Ask Notion AI to summarize your text or highlight key points.

4. Document the Process:

  • Save two versions of your text: the original and the edited version using Notion AI.
  • Take note of the specific edits Notion AI suggested and which ones you chose to apply or ignore.

5. Reflection Questions:

  • How did Notion AI improve your writing?
  • Were there any suggestions that didn’t align with your intent or style?
  • Do you see yourself using Notion AI regularly for editing? Why or why not?

Image made with Canva

AI is transforming the way we think about education, offering remarkable opportunities for personalized learning and innovative assessment methods. While it is crucial for educators to establish guidelines for the responsible use of these technologies, it is equally important for students to develop their ability to engage with and critically assess the role of AI in their education. By incorporating AI into assessments, educators can create a more engaging, dynamic, and relevant learning environment that prepares students for the challenges of tomorrow.

How can we effectively encourage students to think critically about their use of AI? To begin, it is essential for educators to carefully consider key factors when revising assessments that facilitate the integration of GenAI into student learning. The STRIVE model, developed by the University of Calgary, offers a comprehensive framework to guide this reflection and promote thoughtful engagement with technology:

The STRIVE values – student-centeredness, transparency, responsibility, integrity, validity, and equity – emphasize a holistic approach to the use of GenAI (see the STRIVE resource for a detailed discussion and examples). By integrating these considerations, we can foster future-focused learning and skill development for our students, providing a valuable reference as we move forward with our AI assessment updates.

This exploration will focus on assessments in Lane 5: AI as Output Evaluated from our NIC 6-Lane AI Assessment Scale.

Here are six resources that offer ideas and suggestions to enhance your assessments:

  • AI Pedagogy Project (Great collection from Harvard University – with a search tool and categories to filter by AI theme, subject, skills, etc.)
  • Guide to Assessment in GenAI Era – Assessment Strategies (Extensive list of assessment strategies with suggestions for avoiding inappropriate use vs. responsible practices for GenAI from the University of Waterloo)
  • Assessment Design using GenAI (Discussions on GenAI assessment design with specific UBC course examples found at the bottom of page in the drop down: How to make the most of GenAI in assessments and assignments – UBC examples)
  • Assessment Design – GenAI-integrated Assessments (From the University of Alberta – specific course examples under Category 3 – Explore Examples)
  • GenAI – AI Output Evaluated (BCIT list of ideas from multiple disciplines for assessments that critically evaluate AI outputs)
  • Discipline-specific examples of effective prompt writing (University of Calgary prompt library with many great ideas for student prompts and course assessments with some discipline specific examples near the bottom of the page)

As you review the resources, which ideas stand out to you as particularly relevant for enhancing assessments in your course?

Explore:

  • Choose one of the assessment ideas from the resources above (or choose another idea you have either created or applied)
  • Reflect on and plan how you might integrate it into your course or a future course.

Additional Resources:

• The STRIVE model (University of Calgary)
• General Assessment Design Strategies (University of Waterloo)
• Teach Anywhere: Prompts, Tutors and Assistants

Through the adoption of a 6-Lane AI Assessment Scale, NIC has strategically moved to promote the integration of AI tools. Thoughtfully used, these tools have the potential to assist learners in ways that we are only beginning to understand.

Since we are exploring the sixth and final lane, Full AI, in the assessment scale, it will be valuable to highlight a few quotes that contextualize where education stands with AI and offer some candid reflections on what is no longer feasible in this quickly changing landscape.

“…by redesigning our assignments to include in-class elements, direct connections to local and current issues, and a focus on the process in addition to the product, we can make GenAI use easier to detect and mitigate when it is not appropriate for learning goals.”

UBC AI in Teaching & Learning

“It is a potentially disruptive but absolutely necessary perspective which understands that permitting any use of AI effectively permits all use of AI, and since it is undetectable and sophisticated across domains, the distinction between previous Levels 2, 3, 4, and even 5 is somewhat arbitrary.”

Leon Furze (blog post)

“…we take the stance that assessment security, in the traditional sense, is only possible at Level 1, but that assessment validity is possible throughout.”

Leon Furze (blog post)

“I think generative AI is unavoidable, not inevitable. The former speaks to the reality of our moment, while the latter addresses the hype used to market the promise of the technology—a sales pitch and little else. Faculty and students have to contend with generative technology in our world as it is now, not as it is promised to be. That should be our focus.

Marc Watkins (Substack)

To conclude, this final paragraph helps put things in perspective that we are not alone in the education field in our wrestling with AI and its unknown impacts. Read the article, engage with the Empathetic Voice Interface, then reflect on our changing future.

“The majority of OpenAI’s 250 million weekly users are students. OpenAI and other generative companies have committed to release versions of their tools for free with few safeguards, in a massive public experiment that defies belief. There is no touchstone moment in educational history that compares to our current AI moment. If you think generative AI is like MOOCs, then I invite you to have a three-minute discussion about that very topic with a multimodal AI tool called EVI—Hume.AI’s Empathetic Voice Interface. You don’t even need an account, simply click the link and pick the synthetic persona of your choice. Get emotional with it and see how quickly it responds to match your mood. Do you still believe this technology won’t profoundly change education, labor, or even society itself?”

(emphasis added to all quotes)

This exploration will focus on Lane 6 of the assessment scale: Full AI. It is important to understand that this lane is reserved for skills that are assessable outside the context of AI. To phrase this another way, the skills that are being evaluated would be unaffected by the students use of AI.

Below are 4 different options to get you and your students moving with Full AI assessment integration:
  1. Have students create a slide deck with AI for a class presentation that they have previously researched and studied
    In this situation, the presentation itself is not evaluated but the content and findings are. Students may need to present the information from previous studies but lack the skills necessary to complete this in a manner that is visual or functional.
  1. Students work with AI to create an original ad campaign
    In this situation students could converse with the AI to establish a quality end product by continual evaluation. Since AI will likely be an available tool to a novice ad creator in the workplace, it would be reasonable for them to practice with it in the classroom environment.
  1. Have students create AI-generated market analysis reports, including predictive modeling of market trends using AI for data analysis and visualization.
    In this situation, the student is not evaluated on the analysis reports or visualizations but rather on their interpretations of the data provided and their proposals for mitigation strategies. Each student would create/generate their own unique data set. Here are examples of such a case study with Chat GPT
  1. Have students use AI for the creation of visual designs, logos, and branding materials, focusing on the refinement and application of design principles to AI-generated concepts.
Reflect:
  • Consider how you could integrate AI fully into a project or activity in your course
  • What are your thoughts? Does this work?
Resources: