The AI (ChatGPT) future: What do we do now? – ASCILITE TELedvisors Network webinar
Tue 2023-Feb-02 TELedvisors, Monash University, AUS
Tim Fawns (Moderator), Colin Simpson (Monash University, AUS), Aneesha Bakharia (University of Queensland, AUS), Trish McCluskey (Deakin University, AUS), Anna Mills (College of Marin, USA), George Siemens (University of North Texas, USA) Webinar Recording: https://www.youtube.com/watch?v=CyvhrWCSGpE
Key Takeaways
- Discussion of the integrity of assessments in the short term (Anna Mills)
- Focus on intrinsic motivation and engagement
- Establishing policies at the class level
- Establish a norm of transparency about what is AI-generated and what is human-generated
- Ideas on how to make it harder for students to use ChatGPT to complete assessments:
- Ask them to analyze multimedia
- Reference class discussions
- Analyze longer texts
- Peer collaboration on text annotations
- Assigning steps in the writing process
- Conferences with students
- Disclose when using AI assistance
- Use outputs as a starting point
- Ask students to submit a list of prompts used
- Awareness raising for students, professors, and staff is important. Students have been using various iterations of AI for years. (Trish McCluskey)
- Four National Science Foundation AI-research projects are running in the U.S. at the moment (see 25:00)
- Students don’t know how to use Boolean language in google searches, let alone in ChatGPT – skilling-up prompt writing
- Advanced prompting – “prompt engineering” is a skill students will need
- Cheating might become, “I’ve built a great prompt and now I’m sharing it”. Prompt mills with targeted prompts for specific courses/assignments.
- Educational designers have been calling for more authentic learning experiences for years. Now these types of assessments are even more important (Colin Simpson).
- Not long from now these tools will be paid applications which will become an equity concern for the community (Aneesha Bakharia).
- Cheating sites may allow for modifying the APIs to make them more effective for assignment cheating in certain courses/contexts.
- Should teachers use AI to grade? Or to teach? How would students respond to knowing their work has been graded by AI? Student work being fed into an opaque tool, is that okay?
- Promising that AI can handle systemically relevant (grunt work) but learning-irrelevant tasks.
EDUCAUSE Member QuickTalk | GPT: The Generative AI Revolution
Mon 2023-Jan-30 EDUCAUSE, USA
Ray Schroeder, Senior Fellow of the University Professional and Continuing Education Association (UPCEA) and Professor Emeritus at the University of Illinois Springfield (UIS)
Webinar Recording: https://events.educause.edu/member-quicktalks/2023/gpt-the-generative-ai-revolution (Free EDUCAUSE membership required)
Key Take-Aways
- Professor Schroeder curated summaries of 60-70 news items and thought pieces related to AI in higher education. He scrolls through these items, discusses various aspects of them, and poses thought-provoking questions about the impacts of these technologies.
- To review his list, click here.
- Highlights:
- Chrome extensions: https://www.fastcompany.com/90846682/best-chatgpt-chrome-extensions
- They built GPT 3.5 with three billion dollars; what can we expect with three times that funding? https://www.marketwatch.com/story/openai-the-company-behind-chatgpt-in-talks-for-tender-offer-that-would-give-it-29-billion-valuation-report-11672953025
- 6 things ChatGPT can’t do (and another 20 it refuses to do) https://www.zdnet.com/article/6-things-chatgpt-cant-do-and-another-20-it-refuses-to-do/
- It’s futile to try to keep generative transformers out of schools.
- Our job at colleges and universities is to provide learning, experience, and training to our students so that they can best utilize the tools they will be expected to use when they join the workforce. This includes generative AI tools.
Digital Learning Trends at Canadian Higher Education Institutions: Findings from the 2022 National Survey of Online and Digital Learning
Tue 2023-Jan-31 Contact North, CAN
Dr. Nicole Johnson, Executive Director at the Canadian Digital Learning Research Association
Webinar Resources and Recording: Click here
Key Take-Aways
- Surveys of faculty and university administrators were conducted in Spring and Fall 2022. (Note that students were not surveyed)
- Spring 2022: 169 individuals representing 91 different Canadian institutions
- Fall 2022: 287 individuals representing 141 different Canadian institutions
- The combined responses from the Spring and Fall rounds of the survey had 456 total individuals representing 163 institutions.
- Key results are summarized in the graphs below: