There’s no question that the rapid evolution of Generative AI has thrown a curveball into the education space, but Australia’s teachers have been adaptive as ever and finding ways to incorporate this technology into their lessons.
And despite concerns about AI replacing teachers, studies show that while AI can generate impressive creative outputs, it fundamentally relies on human intervention. In short, AI isn’t a threat – unless it learns creativity.
In 2025, the focus has been on ensuring teachers and students not only use AI safely and responsibly but in a way that complements – not undermines – learning.
Byron Scaf, the CEO and co-founder of Stile Education, has played a key role in steering the conversation around artificial intelligence in education – making it clear that technology should serve to empower, not eclipse, teachers.
Below, The Educator speaks to Scaf about breaking down AI myths in education, the non-financial supports schools urgently need, and how a fresh approach to science teaching can help close the STEM skills gap and better prepare students for the future.
TE. You’ve mentioned that AI in classrooms is often misunderstood. What’s one common myth you’d like to clear up – especially when it comes to how AI can support, rather than replace, teachers?
There’s a growing fear that AI will replace teachers, but that assumption comes from a fundamental misunderstanding about what teaching actually involves. People outside the system assume it’s mainly about content delivery, so this current wave of AI-powered apps we’re seeing are frankly quite hostile towards teachers, aiming to ‘disrupt’ education by removing the teacher from the equation entirely.*
Anyone who has spent real time in a classroom knows that teaching is so much more than information delivery. Yes, students go to school to learn, but much of what they learn isn’t just academic. It’s about how to navigate the trials and tribulations of adolescence. That emotional and social development doesn’t happen overnight, and it certainly can’t be sped up by AI. Even if it could, I like to think that we wouldn’t want it to.
We need to come to terms with the fact that AI will radically shape assessment. A few years ago, we used to talk about setting “non-Googlable questions”. That ship has sailed. There’s almost no school-level question now that AI can’t answer. Sometimes, a good old-fashioned paper test will do. But, we’ll also need to lean into socratic seminars, engineering challenges and other more sophisticated (offline!) performances to judge progress of learning.
The other big opportunity where AI can have real impact is in reducing the crashing admin load on teachers, creating more time for them to interact with students and ensuring they have the time, headspace and rest needed to do their job.
TE. Funding always makes headlines, but as we know, it’s not the whole story. From your perspective, what kinds of non-financial support do Australian secondary schools most urgently need to boost STEM engagement and lift outcomes?
We expect our educators to be “teacher by day, lesson writer by night”. Contrary to popular belief, there’s actually plenty of money in the system, yet it continues to blow me away how much we undervalue teachers by expecting them to create everything from scratch. From lessons to resources and assessments. These are all things that can be offloaded so that teachers can focus on the things only they are equipped to do - connect with students, respond to needs in real time and build engagement.
The right curriculum resource providers should serve as true partners, making life easier for teachers and providing the flexibility they need to tailor lesson plans to their individual classrooms. For example, at Stile we’re seeing more teachers asking for ways to incorporate AI into their resources and assessments, so we build the tools and lessons to help them do exactly that. Others prefer to explicitly leave AI out of their classrooms, and we support that approach too. Schools invest good money into these providers, and have every right to think of them as collaborators, not just a content transaction, and who can lift the quality of STEM education to take real pressure off teachers.
Because let’s face it, getting students to care about STEM subjects can be challenging. If we’re serious about boosting STEM engagement and improving outcomes, we need to start treating teachers like the highly-skilled professionals they are and cut inefficiencies. High-quality, curriculum aligned resources that don’t need to be reinvented in every classroom, every department, or every school across the country. Nationally scalable tools, particularly in STEM subjects, would drastically reduce teacher workload which currently is estimated to take an average of 50 hours spent creating resources per one hour of time spent teaching.
TE. Stile brings together curriculum and pedagogy to support science teachers. How is that approach shifting in 2025 as schools navigate new curriculum demands and rising student disengagement?
There are around 10,000 schools in Australia. I like to quip there are at least 10,000 hand-crafted science programs. In truth, there are many more. My assumption that every department perfectly plans together is actually the exception, not the rule. Most science programs are built by individual teachers after hours, meaning there are likely well over 100,000 individual programs. Thanks to the dedication of teachers, some of these programs are incredible, but quality varies widely.
Now introduce new curriculums and teachers have to do it all again. We’ve made quality in STEM education entirely dependent on how much unpaid overtime a teacher is willing and able to put in.
That’s exactly what we’ve built Stile to address. We’ve created a program, lesson-by-lesson, that covers everything you’d need to teach from the first day of Year 7, to the last day of Year 10. From practical activities to pre-tests, and everything in between. Because no two classrooms or 100,000 classrooms are the same, every part of it is fully customisable by teachers.
Additional context:
It’s the support graduate teachers need, with the flexibility experienced teachers deserve and demand.
We have lots more time to make something of serious quality.
• Be the antithesis of textbooks: timely, not timeless.
• Build relationships with curriculum authorities to deeply understand their direction and build to that. Centralise the hard work of understanding changes to the curriculum standards and building new resources to match.
• Build resources in the context of recent, relevant events, and contexts that matter to kids.
• Match (without going overboard) their content delivery style. For example, short, TikTok style videos for key scientific concepts.
TE. There’s been a lot written about the STEM skills gap in Australia’s schools. What are some practical, actionable steps school leaders can take right now to help students become truly future-ready – not just prepared for the next exam?
If school leaders want to prepare students to be truly future-ready, one of the most powerful steps is to create space for real-world learning, particularly in STEM subjects which are practical and inquiry-based by nature. Exploration, experimentation and applying knowledge to real problems is the basis of a STEM skillset. When students engage with real scientific questions and authentic data, they develop critical skills that exams alone can’t measure.
We recently partnered with Lunaria One, an Australian research initiative, to create a bespoke curriculum that involves schools in a global-first scientific mission to grow plants on the moon. Through the Stile Plants Unit, students will collect and analyse data on plant growth in specific conditions, and share their findings with the Lunaria One team in preparation for their real-life lunar mission in 2026. This direct connection still teaches students fundamental curriculum-aligned science, but additionally gives their learning meaning.