1. What is the most interesting thing that this GENAI tool is capable of doing? 

2. What is one potential opportunity you see for using this tool for teaching and learning? 

3. What is one potential problem/challenge that this tool presents to educators?

Since that document is so long, NOLEJ truncated it and only used the first 50,000 words. The materials it created were pretty basic: a glossary, quiz questions, concept and flash cards, a summary of the piece. I didn't go through the content very carefully to make sure it was accurate, but assume it was, I guess it's a fine tool for generating some basic learning content.

That said, there was nothing very "higher order" about the content that was created. It's basically a simple study guide for learning terms, phrases, and basic concepts. It might be good for a pre-check assignment to get a sense of how well students understood a reading?

Right now it's a somewhat locked down chatbot. It's not yet embedded with any of our O365 apps, for instance, and in its first week it refused to draw anything, stating that a personal account was required. I just checked again, and it seems to be up for drawing today! But I have no idea what else is getting locked at the admin level.

From the perspective of a teacher of writing, Copilot seems quite rigid stylistically and in terms of its research. For instance, I cannot get it to write a short essay without relying heavily on bullet-points and copious headers. I asked it to rewrite an essay with scholarly sources, and it added a link to scholar.google.com. Not a specific source, just the entire website! And I confounded it pretty quickly by asking for a somewhat niche essay on adaptive sports in New Hampshire that includes scholarly sources. Copilot did find when I merely asked for research, but when I asked for scholarly sources it started to go down a rabbit hole about climate change without any clear connection to adaptive sports.

In general my answers to the questions posed are this, after some interaction with the tool:

I think the most interesting thing it is capable of doing is learning and expanding as it gathers more data and information into its language model.  That means that the answers will continue to get better as more people use it and it gathers more data.  One opportunity that I see in teaching and learning is through its ability to personalize education; similar to adaptive learning.  However, this means students need to be taught how to utilize it properly and how to verify the information it does give them.  This will take a strain off teachers and put individualized help in the hands of learners.  One potential challenge is the accuracy of the output.  Students need to learn how to get the output and then how to verify and apply their own knowledge into those answers to create better work.

The first thing I noticed was that Copilot could not follow my prompt to "improve this prompt" whereas ChatGPT can (and it's pretty good at doing it). So whatever I put in to Copilot seems pretty vague and generic (while also being WAY too flowery with language). I tried a few prompts and found it pretty general.

However, it did two things that ChatGPT doesn't (for good or bad): 1) it fed me ADS for items for sale related to one of my prompts, and 2) it at least cited its sources when providing me withe information, so I could check some of info it was drawing from. I could see this as an activity to help students practice with the importance of source material evaluation.

In one of the courses I teach often, I believe I'm starting to see use of LLMs not just in assignments but forum prompts (way too many uses of "delve" and "firstly"). One challenge I feel like is head of us is: what does it mean to write in the age of AI? What will be valued five years from now and how can we start weaving those skills into our activities so we're not having to talk anymore about how sad we are by the amount of AI-generated work we think we see (or do see) or having to rely on AI to identify AI and help convince students to focus learning and growing vs. clicking an assignment box.

I haven't done that much with it yet for course preparation, but it's interesting as I've noticed features on and off (file upload for example) which proves to me potentially that internally folks are asking questions and maybe decisions are being made actively on the development and use of available tools, which is just interesting. I thought that in particular would be benefical due to its summarizing skills (which I tested) and found really cool! I have so many students not read text due to time, so I may encourage (if this gets turned on) some use and annotating the AI summary with additional findings from the text!

I tested asking some questions about an image of myself I took directly within copilot and to work some AI magic on it (did not work), but I wanted to try.

However, I am finding it valuable to do things like 1. Scour UNH and PSU for end of the semester dates (which helps speed up my decision-making when I have to look at both campuses' information constantly). 2. Generate baseline communications text that helps me get started with writing a short communication. 3. Creating interesting images that I may or may not use.

I want to have students create a work completely generated by AI this semester to see the difference between their human touch and AI and then generate something with more of their influence, so I'm thinking about using Co-Pilot for script generation, image creation, and potentially VO (since it can read text).

Incidentally, the class syllabus specifically mentioned not using AI tools. I.e., I want them to learn how to write not how to get the AI to write for them.

The tool is free up to an evaluatory limit. After that, it's subscription based.

It is expressed through the website that they provide information from the website performance and what it is capable of doing. According to the GPT the website it is used to help students with complex DOCS and PDF to make them less intimidating and more easier to understand; so I decided to try this out by uploading a PDF of my own to test the limits of this GPT.

I asked the program to help me understand what a previous lesson plan meant. Which the results weren’t bad at all; the whole thing was write information all broken down into categories and bolted important things to know the most. I believe this tool can be great to help to understand things a bit more since we all learn differently. I was amazed by the way it actually worked and gave me the information quick and right.

One potential opportunity I see from using this took is definitely to revise and to use just in case if you need that extra help on understanding an assignment. I believe this tool can be used for teaching and learning for both students and teachers.

On the other hand I also believe this can be a problem because if students see this tool they will start to to pay attention in class and just use the tool at home to explain them what they could have learned in present time in class with the teacher. Just like Chat GPT many people including students use this program to do everything for them which is a big problem in the end because they rely to much on it.

One potential problem/challenge that this tool presents to educators is that it is so very accessible, and cheating with ChatGPT has become very normalized. I think, especially with students who are growing up with this program, the lines of support and cheating are being blurred. One potential use of this tool for teaching and learning is that it can support students' learning. It can often explain concepts without revealing the answer if asked. Controlled, it can be a good tool in a classroom.