Justin is founder of Germinotion.
He is an insatiable generalist who weaves solutions using strands from learning design, behavioural design, data analytics, AI and more.
He has a particular interest in using AI to supplement and enhance human reasoning. Between projects, you will find him reading P.G. Wodehouse and eating chocolate biscuits.
This fascinating episode covers so much:
If AI models have a personality
How when it comes to personalised bots that will be with you through your life, we’re almost there now
How AI helps us learn complex tasks at a much younger age - and what that might mean to education
How multimodal AI will become a new set of eyes and ears, augmenting and accelerating our learning
How our minds will extend into AI through the notion of transactive memory
The distinction between a recipe, technique and tactic when it comes to AI - and why tactics and techniques are so much more effective than recipes
The two ways to prompt AI - outcomes-based, and process-based
What a good AI workflow looks like for educators
The potential of institutionalising mediocrity by getting AI to do the same as we’ve done before, only faster
How you can get an AI to reason with itself through chain of thought and reflection prompting
It's the most in depth discussion I've had to date, and Justin was a joy to listen to.
For more information on Justin's prompting techniques you can check out his videos: https://www.youtube.com/playlist?list=PLrwOhkDQ6tVGdD_Zuob72ZiabM2ojFy9x
Episode #10: Justin Germishuys on transactive memory, extended mind, and why prompt tactics are better than recipes