AI for engagement and tutor presence

As a learning and instructional designer, generative AI tools are one of the most exciting technological advancements to have caught my attention. These innovative tools can potentially revolutionise how we approach content creation in the field of education. While the concept of AI can seem daunting, I believe that if used correctly, it can offer significant benefits in terms of efficiency and quality.

The potential of generative AI tools lies in their ability to generate content tailored to specific learning needs. With these tools, one can request output suitable for learners of a particular demographic or level. This means that educators can produce more personalized learning experiences for their students, making learning more engaging and effective. In addition, generative AI tools can help to create more diverse and inclusive content that caters to different learning styles and preferences.

You've probably heard of ChatGPT or Mid Journey, but there are a lot of other Generative AI tools with all sorts of uses.

AI tools and models

1. GPT-4: OpenAI's flagship language generation model that has been used for a wide range of applications, including chatbots, content creation, and language translation.
2. DALL-E: An AI model developed by OpenAI that generates images from textual descriptions. It has been used to create everything from animals to furniture.
3. Hugging Face: A platform that provides access to a range of pre-trained AI models for natural language processing and generation.
4. Deep Dream: A visualization tool that uses deep learning algorithms to generate surreal, dream-like images from existing photos.

Of course, as with any new technology, it's important to approach generative AI tools with a critical eye. While the output may be impressive, testing and questioning any generated content is essential to ensure validity. Consultation with subject matter experts (SMEs) ensures the content is accurate and aligned with the learning objectives.

Another aspect to consider is the ethical implications of using AI in education. For instance, if not carefully monitored, AI-generated content could reinforce existing biases or stereotypes. It's important to ensure that the data used to train these tools is diverse and representative and that the content produced is inclusive and accessible to all learners.

A great article by Zoe Larkin – AI Bias - What Is It and How to Avoid It? – provides examples of how AI bias reflects society's biases and what can be done about it.

AI Bias - What Is It and How to Avoid It?

Understand the problems of bias inherent in artificial intelligence algorithms & making sure your machine learning systems don’t inherit human biases

Levity logoZoe Larkin·Content QueenNovember 16, 2022

Despite these challenges, I believe that the potential benefits of using generative AI tools in education far outweigh the risks. These tools can save educators significant time and effort, allowing them to focus on higher-level tasks such as curriculum development and student engagement. Additionally, generating personalised content can improve learning outcomes, helping students retain knowledge and develop critical thinking skills.

Recently, I've been experimenting with the OpenAI API and found some pretty compelling use cases that could make online learning more engaging.

Strong instructor presence has been shown to increase participation, facilitate knowledge acquisition, and foster a healthy learning community. With this in mind, I created a way to get the AI model to ask questions about learning content presented in an HTML page.

I'm not a qualified web developer; I'm more of an enthusiast with a handful of web and coding skills, but I'll do my best to explain how it works.

The learning content I want to use to formulate a question is tagged with an HTML class gpt3-read. At the end of the content, I added an empty div with the id gpt3-results. When the page loads, I use a for loop to add anything tagged with gpt3-read to a variable called content. The rest of the code connects to the API using my API key and constructs a prompt that begins with 'ask a "what do you think" style question based on the following text:' followed by whatever was stored in the content variable. The results are output in the gpt3-results div. I've been impressed by some of the questions it asks.

You're welcome to download the following file and have a play around. You'll need to add your own API key from OpenAI.

I think there's a lot of potential here to use AI to generate questions tailored to individual learners. Imagine, as a learner, logging into your course, being welcomed back and asked a couple of questions related to the last piece of content you interacted with. Or being asked a question related to the last forum post you made. What about using your engagement statistics and the results from your last assessment to restructure the entire course content or develop a personal learning plan?

At this stage, I'm not capable of implementing any of this, but I will have a go at getting the questions working in Moodle.

Using generative AI tools in education can transform how we create and deliver content. However, it's essential to approach these tools critically and ensure that the content produced is accurate, inclusive, and aligned with learning objectives. As learning and instructional designers, we are responsible for staying up-to-date with the latest technologies and using them to benefit our students and support their learning journey.

If you're wondering how I managed to put the example above together with limited web and coding skills… I asked ChatGPT to give me a hand.