Just came across the info about the Harvard webinar happening TODAY (May 2, 2023) 11:00 a.m. – 12:00 p.m. (Eastern Time). You can register here. If you can’t make it, the organizers will provide access to webinar recording for those who register. More info from the event webpage:
Artificial intelligence (AI)—including generative AI tools like ChatGPT—can be a force multiplier for instructors. Join Wharton Associate Professor Ethan Mollick and Wharton Interactive’s Director of Pedagogy Dr. Lilach Mollick as they showcase how AI can revolutionize the way you teach.
In this webinar, you will:
Witness a live demonstration of working with Large Language Models (LLMs)
Learn how to use AI to generate useful teaching materials with ease
Discover the advantages and disadvantages of various AI techniques
Understand how AI can help identify knowledge gaps, uncover misconceptions, and support student learning
Don’t miss this important opportunity to learn from leading experts in the field, enhance your teaching strategies, and get your generative AI questions answered.
I did my own webinar a month ago. Here’s a handout with some tips for instructors and links to slides & a recording:
I’m hoping by the end of the summer to test out the best open source alternatives to ChatGPT to recommend to instructors. Right now there are some options like Open Assistant https://open-assistant.io/ , Hugging Chat (based on Open Assistant), StableLM, Stable Vicuna, etc.
Thanks for all of this Doug, especially the resources in the handout. I’m keen to see what you discover for open source alternatives --I tapped a lot into Hugging Face when the image generators were starting to appear a year ago. Their NLP course looked intriguing though it requires a “good knowledge of python”.
The front face of https://stability.ai/ looks promising, and I just got distracted playing with Dream Studio, but poking around you find a lot of the tools attract with free and credits, or ones like the Photoshop plugin require API keys that are going to require $ too.
With the image tools, e.g. Adobe Firefly, they are providing a lot of options so it’s not just enter prompt and click (I saw Dreamstudio had a negative prompt, what to exclude).
It’s a dizzying array of options, and again, appreciate anything you share from your explorations.
This was a great webinar and I have gone away with some amazing ideas about to use AI. At present I am preparing some draft learning content for short 10-credit / 5-week modules, not the definitive article, but the starting point. I am now able to design a whole module with aim, learning outcomes, indicative content, assessment brief, authentic assessment methodology, marking grid, and reading list in approximately 30-45 minutes. I can produce about 5 a day and then send these to academic faculty for them to proof-read and put their mark on these.
Next step is to share process and see what reaction is, at present nobody knows that the draft is AI generated and those that have read the modules, think they are brilliant.
Question I do have is, are there any AI principles in academia that regulate Generative AI use in creating learning content?