The goal of this project will be to explore and experiment with a variety of open-source Gen AI tools and technologies that align with open education values, including focusing on issues of equity, accessibility, and inclusivity while also offering offline access, reducing environmental impact, and ensuring student privacy. BCcampus will work with partners in the develop of an ethically focused AI powered study tool prototype to compliment the BCcampus open textbook collection.
This is one among many items I will regularly tag in Pinboard as oegconnect, and automatically post tagged as #OEGConnect to Mastodon. Do you know of something else we should share like this? Just reply below and we will check it out.
A hat tip to @clintlalonde for sending me the link to this project. First of all, I see it incredibly valuable and something that seems less than commons, that a team uses a blog to document the progress of a project, that itself becomes an open resource.
I also note that this something enabled by the OpenETC that provides this capability for educators in higher education across British Columbia (and a 2021 OE Award winner).
And for an example, I am proud to see how Harper Friedman, who I got to work with previously on a BCcampus project, so clearly explains the set up of GPT4ALL an open source tool to run GenAI locally on your own machine. See Harper’s post
But for this blog post, I have chosen to feature GPT4ALL from Nomic AI for the following reasons:
Is an open-source software,
Emphasizes privacy,
Can interact with your local documents,
Quick to install (10-15 minutes),
Easy to use and is virtually “plug-and-play”,
Easy to customize the System Message of the model, which tells the model how to behave and interpret the conversation.
This is enough to convince me to give it a try. I know there are several ongoing efforts for this as a solution for locations that have low/no internet connectivity-- I am tagging @Sirbiti and @olimiemma to maybe chime in as they are pursuing this, as well as @danmcguire who has been at this a while and @moodler who has advocated local GenAI approaches-- and I am sure I am missing many others working in the space.
Thanks, Alan, for including me in this discussion. My focus is always on what or how teachers use the results of AI responses with their students for effective learning. Having deep understanding of that makes a difference in all things AI. We’ve been practicing with results from various free AI for translating children’s’ picture books and aligning already available OER content and assessments to local standards. Developing a specific AI than can translate a specific language for use in Moodle is being worked.
Here’s a very promising report from Nicolas Martignoni about using an AI with MoodleBox that I haven’t tested myself:
In MoodleBox 4.10.0 and AI
3 Jul
Here’s a guide to use Ollama LLM integrated with Moodle 5.x in a MoodleBox, a cool way to use AI totally offline!
### Step 1 – Install Ollama
curl -fsSL https://ollama.com/install.sh | sh
On the MoodleBox, we use the `gemma3:1b` model (815 MB).
This model doesn't support the generation of images.
Let's test it:
ollama run gemma3:1b
You can now type a prompt of your choice, e.g. "Explain why generative AI is not reliable?" :-)
It's also possible to use `gemma3:4b` model (3.3GB).
This model support the generation of images, but as of Moodle 5.0.1+, it's not possible yet to use the image generation.
## Step 2 – Configure MoodleBox
1. Open http://moodlebox.home/admin/settings.php?section=httpsecurity
* Remove from "cURL blocked hosts list" field the following lines
127.0.0.0/8
localhost
Today, my friend, Steve Miley, walked me through setting up gemma3:1b on a MoodleBox using the directions above. Then, I got to use it. Without being connected to the internet, the AI on MoodleBox generated the attached two documents. It took about 10 minutes total. Half of which was me typing and thinking.