Throughout Module 5 we learned about generative AI and the pros and cons of using it as well as what large language models are. I also used the SECTIONS Model in order to assess a media and multimedia learning page, that was used in this course.

I decided to use Chat GPT as I always hear people talking about how it can benefit them in many ways. I used the ‘Art of Small Talk’ training game. This game prompt guides users through levels of increasingly challenging small talk scenarios and provides feedback throughout to ultimately improve conversational skills. I know I am not the best with small talk so thought I would have a go! Right away, Chat GPT was correctly me on my small talk. I also found it extremely hard to not dive into a topic. For example, on level 1 I was talking with a senior and we were talking about the weather/ sunshine. I found myself typing out the benefits of Vitamin D and had to correct as that is clearly not small talk. I am unsure how I feel about AI correcting me on how to talk, but the game was overall interesting and I found it difficult to continue to keep up with the small talk. When continuing to experiment with Chat GPT, I was surprised at how the interactions felt very human like. As in talking with the AI, it felt as though I was conversing with a human. I am looking forward to using Chat GPT in the future to aid with certain brainstorming ideas for work.

Screenshot of chat GPT “Art of Small Talk” game Level 1.

Using Chat GPT further, I was attempting to have it produce images and it seemed to not be able to do so. I am unsure why as my understanding with this module was that it could display images. After several attempts to have it produce an image of happiness it continued to state it could not do so. The conversation continued with Chat GPT but I was still unable to produce an image from the generative AI large language model.

A screenshot of attempting for Chat GPT to produce an image on happiness without success.

Further exploration with DALL-E2 – a generative AI tool used to create art and images from text prompts – it would not allow new users to use it and thus I moved on from attempting to create this image. Potentially with Chat GPT and copy right problems occurring with images etc., the generative AI tool may be changing its policies due to these conflicts. It seemed to also be tied to having access if you pay for Chat GPT 4.5.

I moved on to use Stable Diffusion to attempt to produce what happiness, ultimate happiness, and ultimate happiness even happier than happy looked like for an AI generator. It was linked to my google account, and I assume that some of these images are definitely biased as Stable Diffusion was only producing a woman as ultimate happiness. I appreciated that a cupcake was the first image that came up when ‘happiness’ was inputted into the image generator. See images below of what images were produced with different styles (cinematic, animation, fantasy art and cyberpunk) describing happiness with Stable Diffusion.

Images generated from Stable Diffusion when ultimate happiness inputted into AI generator

I am curious as to how Chat GPT will change over the years, and if it will continue to stay free for some time. With Chat GPT and other generative AI sites having biases, and privacy concerns, I can understand how their policies are constantly changing. It surprises me that they rely on historical data, and I wonder if there is a way to have Chat GPT 3.5 to update constantly with more current information in order for it to stay relevant for the times.

Further exploring module 5 with the generative AI tools, I find them helpful with brainstorming of ideas and writing. It also makes the ease of writing out tedious code much easier, something I have always struggled with.

At this point in time, with how many Generative AI tools are available, there is no stopping it now. I think education should embrace this and use these tools to aid students and learning. I am happy this course made me more aware of the value and options with generative AI and what large language models are to help my understanding with their limitations and how they function. I do believe teachers do not need to fear AI replacing their jobs due to the lack of ability for them to think critically. I agree with the use of them, and that as long as students are citing their work with generative AI, this will ultimately aid in transparency with the teacher student relationship and grading of work.

From Module 4 of EDCI 337 I used the Interactive Videos on a WordPress Blog with H5P with the SECTIONS acronym in order to aid in selecting valuable media for learning.

Students: Students of the EDCI 337 class are able to access this media once they are logged into their UVic account. Students would need access to basic wifi or hotspotting from a cell phone with data. There are videos, thus having somewhat of a good connection would be important to access this media. Overall students should be able to access this media with no problems.

Ease of use: If one is able to follow the directions by reading and listening to the video, then it is easy for both instructors and students to use. As we went through how to add an interactive video on a WordPress Blog with H5P in module 4, students are able to achieve this with following the layout and instructions of the media.

Costs: With the course, the realization of making many online videos and media content for free is possible. The H5P website gives a 30 day free trail, thus not completely free for the teacher long term in order to update the content I assume. Thus a cost is associated with this media to be produced by the technology.

Teaching functions: Being able to follow the written instructions and follow the instructions of the video allows students to complete the task effectively and efficiently. Little skills that students have today would be needed in order to complete this task. Allowing students to experiment with H5P would be the main skill needed, and to know how to use the basics of a WordPress site is needed as well. Throughout this course we have had the opportunity to do so, thus the media is relevant and accurate for this technology.

Interaction: The interaction of the student and content is good for this media. It somewhat lacks interaction between leaners and instructors, yet it does do a great job of having links to click in order to have examples and further information that may be needed and wanted by the student. Overall I think this media allows effective learning interaction for the student.

Organisational issues: As a UVic student who has access to the EDCI 337 resources, this media is only able to be accessed through the UVic website (Brightspace from EDCI 337) Yet I understand why this must occur for UVic courses due to privacy requirements, and the fact that students are paying for the course. It seems as though there are free similar resources available on this topic which is great.

Networking: As this video is able to be shared throughout different platforms, the learning groups are able to access it with the right technology.

Security and privacy: This media is secure and protects both the student and teacher. This media does not necessarily need to be private and allowing access could help others learn how to gain this skill of adding interactive videos to a WordPress site with H5P.

This SECTIONS model assessment allows one to select media and multimedia learning in a more logical reasoning. While going through this assessment, it is clear that the media allows for a great example of what can be used for multimedia content. It has text instructions to easily follow, video to aid in the learning and code to copy and paste for ease of use. It also ties into privacy for both instructors and students.

Thank you for reading my blog post. Feel free to comment below!

References

“’Art of Small Talk’ training game…” ChatGPT, 13 Feb. version, OpenAI, 29 Mar. 2024, chat.openai.com/chat.

“’Show me an image of ultimate…” ChatGPT, 13 Feb. version, OpenAI, 29 Mar. 2024, chat.openai.com/chat.