Imagine a classroom in 2035. A teacher begins the day not by marking piles of homework but by reviewing insights generated by an AI assistant. Instead of being bogged down in administrative tasks, she spends more time with her students -- listening to their questions, nurturing their curiosity, and guiding their learning journeys. AI is not replacing her; it is amplifying what she does best: Human connection.
This image is possible, but not inevitable. As I argued earlier in this series, Bangladesh risks slipping deeper into an “ego-system” of silos and distrust if we are not careful. But if we design intentionally, AI can help us move toward an eco-system -- where collaboration, trust, and alignment shape the future of education.
So how do we get there? I suggest four pathways. Each is not a silver bullet on its own. But together, they can help us move from fragmented projects to a thriving ecosystem.
- Impact networks: Trust over turf
During the pandemic, a small group of educators, policymakers, and innovators in Bangladesh tried something different. We nurtured a network called #NextGenEdu. Instead of building another formal organization, the network operated on four principles: Mission, not organization; humility, not brand; trust, not control; and node, not hub.
The idea was simple: When people are united by purpose, they can share knowledge and act more effectively, even across silos. This is what researchers call an impact network -- a web of relationships built on trust that allows collaboration to flourish where rigid hierarchies often fail.
For AI in education, such networks could be vital. They would bring together teachers, government officials, EdTech entrepreneurs, parents, and learners to test ideas, share stories, and build solutions together. Without trust, collaboration collapses into competition. With trust, even small groups can make big ripples.
- Delivery units: Coordination that cuts across ministries
Good ideas often die in the maze of bureaucracy. Ministries and agencies work in parallel, but coordination is rare. AI in education will require inter-ministerial cooperation -- linking not just education, but ICT, finance, labour, and local government.
One way forward is the delivery unit model, pioneered in the UK under Michael Barber. These are small, focused teams designed to cut across ministries, track progress, and solve bottlenecks. Instead of every agency waiting for someone else to move, delivery units create momentum.
Imagine a delivery unit in Bangladesh focused specifically on AI in education. It could ensure that curriculum reform at the National Curriculum and Textbook Board is aligned with teacher training at NAEM, that EdTech providers have pathways to pilot their tools responsibly, and that policymakers have real-time data to guide decisions. Coordination is not glamorous work, but without it, even the best frameworks remain paper tigers.
- Systemic EdTech testbeds: Safe spaces to experiment
All over the world, AI tools are being marketed to schools. Some may be transformative, others little more than hype. The risk is that Bangladesh adopts technologies in a piecemeal way -- one district tries one thing, another tries something else, with little evaluation or shared learning.
This is where systemic EdTech testbeds come in. The Global EdTech Testbed Network (GETN) describes them as platforms where new technologies can be tested in real-world environments before scaling. Testbeds bring together providers, government, teachers, and parents to see what actually works -- and what doesn’t.
A testbed in Bangladesh could, for example, allow a handful of schools to pilot AI-based tutoring tools, with structured feedback from teachers and students. Policymakers would gain evidence, companies would learn how to adapt their products, and parents would have a voice in shaping what enters classrooms. Testbeds act like safety valves: They reduce the risk of large-scale failure while encouraging innovation.
- Futures thinking: Expanding the imagination
Finally, we need to widen our imagination. Too often, policymakers get stuck between only two options: “Do nothing” or “adopt the latest technology.” Futures thinking offers a way out.
The futurist Jim Dator once said: “Any useful statement about the future should at first appear ridiculous.” By exploring alternative futures, we expand the space of what is possible and desirable. For instance, what if instead of just Artificial Intelligence, we asked about the role of Artificial Stupidity in education? Could deliberately slowing down, making “inefficient” but human choices, sometimes be wiser than optimizing everything?
By engaging teachers, students, and communities in futures workshops, Bangladesh can ensure that decisions about AI are not just made by technocrats or donors. Futures thinking democratizes the conversation, reminding us that technology should serve human aspirations -- not the other way around.
These four pathways are not about technology alone. They are about creating the conditions for collaboration: Networks of trust, coordination across ministries, platforms for experimentation, and imagination beyond the immediate.
The truth is, AI will amplify whatever system it is placed into. If introduced into an ego-system, it will deepen silos and accelerate competition. But if nurtured within an eco-system, AI could strengthen alignment, unlock creativity, and help us work together in ways we haven’t managed before.
A call to action
Bangladesh is at a crossroads. We can treat AI as just another project -- rolled out with fanfare, fragmented across agencies, forgotten after a few years. Or we can use it as an opportunity to do something bolder: to shift from ego-system to eco-system, and to design education futures that are collaborative, holistic, and humane.
If we can dream together, we can work together. And if AI can help us dream together, it can surely help us build the education future our children deserve.
Shakil Ahmed is an educator, futurist and storyteller at Ridiculous Futures, network coordinator at #NextGenEdu and studying his PhD in Futures Studies at Tamkang University, Taiwan.
