George SIemens

The AI future for lesson plans is already here

Another of our intermittent blogs during the #AARE2022 conference
If you want to cover a session at the conference or want to respond to something you’ve heard, please email to check in. Thanks!

George Siemens gave the 2022 AARE conference keynote speech yesterday morning and the takeaway was this: artificial intelligence (AI) is already changing education in Australia whether we’re ready for it or not. He went on to suggest that unless we’re mindful we, as researchers and teacher educators, may not have a seat at the table in deciding what the future of education will look like. This blog post is about the future of AI-generated lesson plans and how we might think about this development today and into the future. Think of it as a concrete case study of exactly what George was talking about this morning where the conclusions might apply to other areas of education

What do today’s AI-generated lesson plans look like?

AI-generated lesson plans are already better than many people realise. Here’s an example generated through the GPT-3 deep learning language model:

Prompt: A lesson plan for English for teaching 16 year old students about the idea of “juxtaposition” within Romeo and Juliet.

Output from GPT-3 (verbatim, with formatting provided by the algorithm):


Juxtaposition is the act of placing two things side by side, often for contrast. In literature, authors often use juxtaposition to create a certain mood or atmosphere, or to emphasize certain qualities of a character or situation.

 In Romeo and Juliet, Shakespeare uses juxtaposition extensively to contrast the two lovers and their families, the Montagues and the Capulets. He also uses juxtaposition to contrast the tragic events of the play with the light-hearted moments.


 By the end of the lesson, students will be able to:

 – Understand the concept of juxtaposition

– Identify examples of juxtaposition in Romeo and Juliet

– Explain how juxtaposition is used to create mood and atmosphere in the play


 – A copy of Romeo and Juliet

– A whiteboard and markers

– A teacher-created handout with examples of juxtaposition from the play (optional)


 1. Introduce the concept of juxtaposition to the class. Explain that authors often use this literary device to create contrast and emphasize certain qualities.

 2. Ask the students to think of an example of two things that are often juxtaposed. Write some of their responses on the board.

 3. Tell the students that they are going to be looking for examples of juxtaposition in Romeo and Juliet. Explain that they should pay attention to how the contrast created by the juxtaposition contributes to the mood and atmosphere of the play.

 4. Give the students time to read through the play, or assigned sections of the play if time is limited. As they are reading, they should be looking for examples of juxtaposition.

 5. Once the students have finished reading, ask them to share the examples of juxtaposition that they found. Write these examples on the board.

 6. Finally, ask the students to explain how the juxtaposition affected the mood and atmosphere of the play. What did it emphasize about the characters or the situation?

 Extension Activity

 Have the students find additional examples of juxtaposition in other works of literature, or in everyday life. They can share their examples with the class and explain how the contrast created by the juxtaposition contributes to the overall effect.

Imagine that you’d set a task for a class of second year Bachelor of Education students to create a lesson plan with that same prompt. How do you think the AI response would compare to some of your students?

Anybody can generate their own lesson plans for a year level, subject area, and topic that interests them with a similar prompt:

Some concerns

It is trivially easy to generate useful lesson plans using this technique. One obvious upshot is that setting assessment tasks for any students in initial teacher education that involve them creating lesson plans isn’t a great idea any more—it’s too simple for them to generate one. Yet there are new opportunities that arise:

  • Why not get students to generate a few lesson plans, look at the patterns, and write something about the essential structure of this thing that we call a ‘lesson plan’?
  • Why not get them to take a generated lesson plan and improve it, annotating the reasons why their changes have made it better?

Another legitimate concern that arises is that inservice teachers might start to use the next generation of AI-generated lesson plans (which will undoubtedly be an order of magnitude more powerful) without critique—or worse, that some jurisdictions might actually request that teachers use such an approach in future.

A word that we need to look to is “design”

The issues raised by AI regarding lesson plans and in many places in education too can be addressed by consideration of design. When design in education is done well (whether that’s learning by design, design thinking, co-design, or within the subject area named “design”) it always places an emphasis on two things:

  1. Authentic problems: such that the learner must always construct an interpretation of the problem before they can address it
  2. Process and rationale such that the output that the student produces is impressive only if their process and rationale support what they’ve done.

When assessments follow these two ingredients then educators can give students free rein to use whatever tools they have at their disposal. The adoption of AI stops being a concern. When students are being assessed through their process rather than their output, students can use whatever tools are available. The challenge is integrating use of such tools into solving problems through collaboration, critical thinking, cultural understanding, and creativity.

Design as a response to “what should be taught”

George Siemens concluded his presentation by suggesting a list (controversially) of what should be taught in the context of an AI future. A summary/interpretation of his key points of what we should be teaching is:

  • Beingness: what it means to be human in the world, the interconnectedness of all things
  • Systems thinking: how systems change and what complexity is about
  • Technology and how to use it: machine learning and data literacy, computational thinking, collaborating with non-human intelligences

Increasingly, design has become a part of education: design for learning, learning by design, thinking, and so on. The epistemic fluency to design using computational tools in a way that enriches material life and human culture is at the root of all three of these areas. 

For any subject area, teaching using a design approach shifts the focus from knowing content to knowing process. It becomes less about how to get from A to B in a straight line and more about knowing how to frame problems, use tools, and communicate outcomes. More design in education provides one way of responding to this increase presence of AI in education, whether we’re ready for it or not.

It might even provide a response to George’s provocation about McKinsey, Deloitte, or Microsoft trying to get in on a slice of the education sector. Education conceived as design—process rather than output—prioritises the humans involved in the enterprise and makes it harder to sideline educators.

Dr Nick Kelly is a Senior Lecturer in Interaction Design at the Queensland University of Technology, in the School of Design. He is a genuinely cross-disciplinary researcher spanning the fields of Design and Education. He conducts research into design cognition (how designers think), metacognition in learning (how teachers and learners develop their metacognitive abilities), and places where these two things come together (design pedagogy, design for learning, learning by design, design of learning technologies). His specialisation is in the design, facilitation, and analysis of online communities.

Dr Kelli McGraw is a Lecturer in the Faculty of Creative Industries, Education Social Justice at QUT. Currently teaching units in Secondary English curriculum, pedagogy and assessment, her prior experience includes teaching high school English and debating in Southwest Sydney, NSW. Kelli researches the fields of English curriculum studies, secondary school assessment, teacher identity, digital literacy, popular culture and new media texts.

Siemens: the biggest challenges facing education now and ways to meet them

The AARE 2022 conference opens this year with a keynote from George Siemens. Here are some of his thoughts.

We have been hearing about fundamental change in education, often driven by technology, for several decades. Previous theorists, like Illich and Freire, similarly advocated for systemic education change, but their concerns were driven by economics, inclusion, and impact. When “education must change” is now advanced as a narrative, it’s often driven by a motivation to drive use of technology or the outsourcing of some core service of universities or schools. In response to the steady drum beat of calls for change, educators have become somewhat immune and even sceptical. Where is this new reality? Why has covid produced a longing for in-person learning, rather than a great drive for online learning? In our professional lives, the appeal of space and place interactions, while increasingly augmented with online engagement, remains strong.

In this talk, I present three dynamics to consider regarding our future education systems. First, I address the education landscape and the many additional stakeholders now prominently providing some core function. Secondly, I’ll address the conflicting space between data-centric research and complexity-science orientations. Thirdly, I’ll discuss the system of education itself. I believe we are facing a systemic challenge and when looking a decade into the future, it’s apparent that a fundamental change in role and responsibility will unfold for education. 

Before I begin, I want to set context for perhaps the most substantive challenge facing education. Trends can be seen as primary or secondary in terms of impact. Secondary impacts include state government mandates and even national level testing and assessment. A secondary trend may change parts of how teaching happens, the content taught, or how students are assessed. Often, the trend has a short timeline and is connected to the interests and motivations of the party in power. Primary trends, in contrast, are those that fundamentally and structurally change the systems of learning and education. In order to keep the system as it currently is, external pressure must be exerted to keep a primary trend from taking over. Unlike the rollout of national testing, which requires mandates to make things happen, a primary trend requires policy and intervention for it to NOT take over. 

We’ve seen numerous primary trends over the last decade, including the rise of social media and mobile technology. The primary trend confronting education, however, has a long history, dating back to the 1950’s, and is now beginning a rapid and alarming ascent to prominence in all areas of our lives: artificial intelligence. AI presents humanity with a unique challenge that we have not faced before: an agent with intelligence that rivals our own in a growing range of domains. 

Educationally, this presents a significant problem. In 2022, Generative AI has grown in influence and prominence. AI can now generate and create in domains that we have previously seen as exclusively our own: art, literature, and scientific discovery. DALL-E 2 and Stable Diffusion have created art that has won state art competitions. Moonbeam can create writing that has surprising coherence. LaMBDA can carry on conversations that are human-like. After being promised for decades that we would give up routine and mundane tasks to AI while retaining creative activities for ourselves, AI is emerging as an active competitor for our most human skills. Research and scientific discovery is now a pairing of human and artificial cognition. The entanglement that happens at the intersection of the two is spilling over into non-technical domains and sociologists, educators, and psychologists are evaluating how this interplay occurs and how it should be managed and supported.

Education, and all of society, moves forward with the looming AI trend in the background as the overarching development of the current era. The education landscape itself is undergoing significant commercialization and reliance on external stakeholders. Schools and universities are no longer primarily self-contained ecosystems. Instead, the fragmentation of function that defines globalisation has arrived. Online program managers support the development, marketing, and recruiting of students. International programs rely on a global recruitment network. Behind the scenes, consulting firms who had previously mainly addressed the needs of big business and large government now provide services to university and school leaders. Policy papers and guidance documents are produced by every major consulting firm in Australia and the prospect of big economic gains through innovation is a salivating prospect. Big technology is increasingly managing core university computing and security and privacy are now off loaded to these firms. Underpinning all of these transitions is the digital revolution and the data it produces as each student movement and interaction and engagement is logged and recorded. 

Digitization produces data and data produces analytics. For researchers, a conflict is unfolding reminiscent of the science wars of the 1990s. Data has won. All research – quantitative, qualitative, mixed – is digital in capture or analysis or publication. To this end, the quantitative side has resolutely and decisively ascended to the throne. The real space of debate now is on how to move data-centric research from focusing on isolated studies to instead begin assessing and evaluating holistic systems. The “science war” emerging is one where the expression of data is the primary concern. Systemic modelling and holistic assessment sits in conflict with NAPLAN and standardised testing. Research conducted is now increasingly focusing on digital spaces or at least spaces that have a digital component: AI predicting how protein folds,  sensors capturing remote environmental data,  psychologists evaluating the mental health of students in digital settings. A complexity science approach to research moves from granular and limited scope research that occurs in sanitised or limited context settings to including multi-faceted and nuanced contextual data.  

When systems change, inefficiencies are created. Organisations and individuals who evaluate and exploit those efficiencies reflect Gould’s punctuated equilibrium (or Kuhn’s paradigm shift): a sudden and significant phase change. This has been experienced in many sectors already, including the move from physical state music and movies to digital, the shift to on demand rather than broadcast media, and the move to networked media rather than centrally controlled. The accrual of inefficiencies – doing the things afforded by previous philosophies and technologies – is confronting education. How should we teach when AI is better at many cognitive tasks than we are? What should we teach when we can find and access the world’s information from our phone?

Looking a decade into the future, international organisations such as OECD see a world where technology is central to learning, where systems of education are dramatically different from what we see today, where AI is a co-learner, where focus on wellness and wellbeing are increasingly important. Educators have long been the end recipients of government initiatives, quasi-scientific pedagogical approaches, and somewhat short-sighted policy changes. The real work of education leadership is the work of systems change. Systems makers – those who create the structures that others work within – needs to be claimed by organisations such as AARE. The future of education is one that will only emerge to serve the broadest range of stakeholders when all participants have the ability to have a voice and to shape the conversation. Finding points of leverage in shaping learning systems through policy, research, funding, and planning landscape is the critical work of today for educational leaders.

Professor George Siemens is the professor and director of the Centre for Change and Complexity in Learning UniSA Education Futures. He researches networks, analytics, and human and artificial cognition in education. He has delivered keynote addresses in more than 40 countries on the influence of technology and media on education, organisations, and society. He has served as PI or Co-PI on grants with funding from NSF, SSHRC (Canada), Intel, Bill & Melinda Gates Foundation, Boeing, and the Soros Foundation. Professor Siemens is a founding President of the Society for Learning Analytics Research. In 2008, he pioneered massive open online courses (sometimes referred to as MOOCs).