I just gave a #GenerativeAI workshop for some 300 high-school students in a program of Toronto's Lifelong Leadership Institute. The LLI serves the Candian Jamaican, Carribean and Black community with #education opportunities. We opened with the most meaningful #LandAcknowledgement I have ever heard (by a participant), and I got to have a spirited conversation with a group of wonderful, well informed, smart, and curious students.
Some take-aways:
We polled uptake:
- About 1/3 of students have not used #ChatGPT so far;
- About 1/3 have used it only a few times overall;
- About 1/3 use it regularly. Interesting: of those who use it regularly, the majority are frequent users, i.e. either you have not used it for some reason, or you use it a lot. Not much middle ground.
During the workshop, we focussed on the Sentient Syllabus - Three Principles for AI use in academia...
https://sentientsyllabus.substack.com/p/just-the-basics
... and one question stood out: how can we implement the new thinking around generative AI in classroom practice? Indeed, how. We need to grow practice. I always look for a win-win angle - coercion is the wrong way, but on this topic it is particularly easy: figuring out solutions is absolutely a topic that learners and lecturers must approach shoulder to shoulder. In this case it means: "Better learning support for students, deeper questions for educators".
Of course, the question of #academicMisconduct always comes up: my mantra is "Have the AI think with you, not for you". From that perspective we can avoid making this the next battleground. Liberalize - and teach how the AI's level of performance can no longer be a passing grade. Because that quality is what your future employer gets for free.
#Plagiarism? Open it up. #Alignment? Stop thinking about controls and democratize it instead. #Bing? Yes, that psychotic episode created a foundational piece of writing. Jobs? That's on us: to educate society the value of the human touch. #Sentience? No. But #emergence! It does more than what it was trained for. The future? IDK. I really don't. But stay in control. Learn what it can do, and then determine what it should do for us (always good advice).
:-)
