George Veletsianos, PhD

Tag: artificial intelligence

GPTs and one student’s custom version of ChatGPT

By

On November 15, 2023

In teacherbots

A few days ago OpenAI released GPTs, which are “custom versions of ChatGPT that you can create for a specific purpose” (announcement here: https://openai.com/blog/introducing-gpts), meaning that one could produce a GPT that is dedicated to a specific set of tasks. I’ve seen a few of these so far, including Mike Sharples’ chatbot that uses his book to offer teaching advice https://chat.openai.com/g/g-RCHNUwnD1-teachsmart, Mairéad Pratschke’s expert in digital education and learning design https://chat.openai.com/g/g-hrPUmXB5X-digital-professor, and this degree builder I came across https://chat.openai.com/g/g-KVB8vSaAJ-degree-designer.

One of the more interesting use cases I saw was one student’s Reddit post titled “I just replaced my chemistry professor with AI:” https://www.reddit.com/r/ChatGPT/comments/17slpti/i_just_replaced_my_chemistry_professor_with_ai/ 

 “I gave it the prompt: you are a college professor teaching inorganic chemistry 2 thermodynamics. The scope of your class is covered in the uploaded documents.

I then uploaded my professors PowerPoint slides and copied and pasted the chapter from the book. All the exercises, extra problems, and a thermodynamics properties table. I also included a summary of the topics covered.

I had to double prompt it to only teach from the documents and upload pdfs seemed to work a lot better than .txt files.”

Lots of areas to reflect on here, including student creativity, privacy, and ethics.

Recommendations on the use of Generative Artificial Intelligence at Royal Roads University

By

On October 5, 2023

In generative artificial intelligence

Today I met with Royal Roads University’s Board of Governors to present the work that we have completed in relation to Generative AI. I appreciated the opportunity not only to meet with the board, but also to hear comments and questions around this work and AI more general.

Because this was a public session, I thought it might be beneficial to share the recommendations we put forward. The full report will likely appear on the university website but for those of you who are tracking or thinking about institutional responses, this short timely summary might be more valuable than a more detailed future report.

As background: In March 2023, Royal Roads University established a generative Artificial Intelligence (AI) taskforce. Chaired by Dr. George Veletsianos, the taskforce consisted of Drs. Niels Agger-Gupta, Jo Axe, Geoff Bird, Elizabeth Childs, Jaigris Hodson, Deb Linehan, Ross Porter, and Rebecca Wilson-Mah. This work was also supported by our colleagues Rosie Croft (University Librarian), and Ken Jeffery and Keith Webster (Associate Directors, Centre for Teaching and Educational Technologies). The taskforce produced a report with 10 recommendations by June 2o23. The report (And its recommendations) should be seen as a working document that ought to be revisited and revised periodically as the technology, ethics, and use of AI are rapidly evolving. The recommendations were:

  1. Establish and publicize the university’s position on Generative AI
  2. Build human capacity
  3. Engage in strategic and targeted hiring
  4. Establish a faculty working group and foster a community of practice.
  5. Investigate, and potentially revise, assessment practices.
  6. Position the Centre for Teaching and Educational Technologies as a “go-to” learning and resource hub for teaching and learning with AI.
  7. Future-proof Royal Roads University [comment: a very contextual recommendation, but to ensure that this isn’t understood to encourage an instrumentalist view of AI or understood to mean that the institutions solely focus on AI, the report urged readers to “consider that the prevalence of AI will have far-reaching institutional impacts, which will add to the social, economic, political, and environmental pressures that the University is facing]
  8. Revise academic integrity policy
  9. Develop and integrate one common research methods course [comment: another very contextual recommendation that likely doesn’t apply to others, but what does apply is the relevance of AI to student research suggesting that research methods courses consider the relationships between AI and research practices.]
  10. Ensure inclusivity and fair representation in AI-related decisions.

I hope this is helpful to others.

So very tired of predictions about AI in education…

By

On July 7, 2023

In futures, generative artificial intelligence, my research, scholarship, short reflections, teacherbots

By people who aren’t AIEd experts, education technology experts, education experts, and the like.

Case in point: “AI likely to spell end of traditional school classroom, leading [computer science] expert says.”

I appreciate cross disciplinary engagement as much as I love guacamole (which is to say, a lot), but I’d also appreciate that we stop wasting our time on these same unfulfilled prophecies year after year, decade after decade.

Will AI impact education? In some ways it will, and in others it won’t. Will education shape the ways AI comes to be used in classrooms? In some ways it will, and in others it won’t.

Truth be told, this negotiated relationship isn’t as appealing as DISRUPTION, AVALANCHE, MIND-READING ROBO-TUTOR IN THE SKY, etc, which are words that readers of the history of edtech will recognize.

Powered by WordPress & Theme by Anders Norén