Save Our Sundays is the build-in-public podcast from MyTeacherAide, hosted by co-founders Josh and Jacob. Follow along as they rebuild MyTeacherAide V3, an AI toolbox designed to help teachers save time on planning, resources and admin.
Each episode goes behind the scenes of building an education technology product in real time, unpacking teacher workload, AI in education, product decisions, school feedback, curriculum alignment, privacy, and the messy reality of trying to create tools teachers actually want to use.
It’s for teachers who love teaching, but could do without spending their Sun
AI Is Getting Better. Teacher Judgement Matters More Than Ever.
Last week, we told you AI doesn’t actually know anything.
Which raises a fairly obvious question:
Then why is it getting so good?
In Episode 21 of Save Our Sundays, Josh and Jacob continue their crash course on how AI actually works — and unpack one of the most important concepts for teachers to understand: context.
Modern AI can now search the web, access tools and gather information before generating an answer. And the more relevant context it has to work with, the more you can narrow the range of possible outputs and improve what you get back.
Jacob explains it with a very sophisticated mathematical model involving... guessing which number Josh is hiding behind his back.
Naturally. 😂
But there's another side to this.
AI is getting incredibly good at sounding right.
And that can make it surprisingly easy to switch your brain off, accept an authoritative-sounding resource and put it straight in front of your students.
Jacob shares a real classroom example where an AI-generated resource included questions about a character before that character had even appeared in the film — something the students themselves noticed.
In this episode:
Why modern AI is producing much better answers than it used to
What context actually means and why it matters
How giving AI better source material can improve its output
Why an authoritative-sounding answer can still be wrong
How teachers can use AI for research without outsourcing the research process
Why AI detectors aren't a reliable solution to student AI use
Why teacher judgement becomes more important, not less, as AI improves
Why choosing not to use AI is different from choosing not to understand it
Jacob also shares a practical way he now uses AI for research: instead of asking it to write the research report, he asks it to find and summarise potential sources — then reads and evaluates those sources himself.
You don't have to use AI. But if you're responsible for educating students who are growing up with it, you probably need to understand it.
17 Sept 2026
What Teachers Need to Know About How AI Works | Australian AI Framework Series • Part 7
AI can write a lesson plan, create a worksheet and answer almost any question you throw at it.
There’s just one slightly important detail:
It doesn’t actually know anything.
In Episode 20 of Save Our Sundays, Josh and Jacob continue their look at Transparency in the Australian Framework for Generative AI in Schools by going right back to basics: what actually is AI, and how does it work?
Because if teachers are going to use these tools - and help students navigate them responsibly - understanding what’s happening underneath the chatbot is pretty important.
Jacob breaks down large language models without requiring a computer science degree: why AI is essentially a probability machine, what “next-token prediction” actually means, how models are trained, and why something that sounds incredibly intelligent can still produce an answer that makes absolutely no sense.
Along the way, Josh discovers that ChatGPT does, in fact, know there isn't a twelfth Harry Potter book. So that slightly ruins one of Jacob's examples. 😂
In this episode:
What an AI language model is actually doing when it answers you
Why AI can sound like it “knows” something when it doesn't
A very non-technical explanation of next-token prediction
Why context matters so much when using AI
How AI models are “trained”
Why newer models are getting better at questions older models got hilariously wrong
What all of this means for teachers deciding whether to trust an AI-generated resource
The big takeaway?
AI produces the appearance of thought remarkably well. But understanding how it gets there should change the way we trust and use its answers.
And just as they start getting into the really interesting stuff... Producer Gabby pulls the plug.
Part 2 next week. 😅
10 Sept 2026
You Don’t Have to Like AI. But You Do Need to Understand It. | Australian AI Framework Series • Part 6
You don’t have to like AI. But as a teacher, can you afford not to understand it?
In Episode 19 of Save Our Sundays, Josh and Jacob return to the Australian Framework for Generative AI in Schools and tackle its third principle: Transparency.
And they’re starting with a slightly uncomfortable reality:
You can’t opt out.
Teachers can choose not to personally use AI. Schools can restrict it. Parents can dislike it. But students are already encountering these tools - and if the adults responsible for guiding them don’t understand AI, it risks becoming a repeat of what happened with social media: the kids figure it out first, and the adults play catch-up.
Josh and Jacob discuss why transparency starts with understanding - what AI actually is, what it does well, where it falls short and how it affects learning.
They also make an important distinction: understanding AI doesn’t mean embracing every use of it.
In fact, they argue that students should have very limited access to AI for learning, while teachers - equipped with subject expertise and professional judgement - are much better placed to use it as a tool.
In this episode:
What “transparency” actually means in the Australian AI Framework
Why some teachers understandably want nothing to do with AI
Why opting out completely might leave students worse off
What social media can teach us about getting AI adoption wrong
Why students and teachers shouldn't necessarily use AI in the same way
Three practical ways teachers can start building their AI literacy
Why Brain first, AI second matters more than ever
The good news? You don’t need to become an AI expert.
You just need to understand it well enough to know what it can do, what it can’t do, and how to guide your students through a world in which it already exists.
The adults need to lead this one.
Contact Save Our Sundays
Guest appearances
Does not typically book guests
Based on episode analysis; this does not confirm that the show is currently accepting guests.
Host of Save Our Sundays?
Claim your podcast to manage its listing and keep your show details accurate.
Pod Engine is an independent podcast discovery and analytics service and is not affiliated with or endorsed by this podcast. Artwork and show content belong to their owners. Full legal notice.
Explore this show Podcast research with Pod Engine
