Climate change intensifies, geopolitics shift unpredictably, and threats to global security and health continue to rise. We share one village – the Earth – and its complexity deepens. In such times, we need collective intelligence that crosses disciplines, nations, and values.

Dialogue holds unique power: it gathers scattered voices, integrates diverse perspectives, and weaves them into responses to shared challenges. In our interconnected age, that power matters more than ever. Artificial intelligence – especially generative tools like ChatGPT – adds a new voice to the table. With its capacity to process and connect information, AI opens fresh possibilities for dialogue while also introducing risks. We therefore need not only human collective intelligence, but also the co-intelligence that can emerge when humans and machines learn to think together.

In education, collaborative problem-solving has always been both a pathway and a purpose of learning. Communication, collaboration, and ICT literacy are recognised as core 21st-century skills – yet classroom discussions routinely leave many students on the sidelines – a pattern that tends to worsen online. Here, AI arrives not as a replacement but as a catalyst, breathing new energy into educational dialogue aimed at nurturing collective intelligence.

Around the world, industry and academia are exploring how to integrate AI into dialogue. Early experiments like the Stanford Online Deliberation Platform and Frankly AI offer distinct directions but remain exploratory. In this context, we introduce the AI moderator online platform, created by SmartLearning in Denmark, inspired by a vision of AI acting as a moderator. Built on the OpenAI API, it can work with different large language models. Its distinctive feature is multi-party participation: multiple users can join the same shared conversation space in real time from different devices. The platform supports text-based interaction, with optional audio.

The challenge of learning with mobiles (LWM) approaches in the Global South

‘If you want to hide something, put it in a book.’ This quip from Alphaxard Kimani, farmer-philosopher of central Kenya, programme leader for The International Small Group and Tree Planting Program (TIST), and long-time friend and colleague, succinctly captures the core problem of learning with mobiles (LWM) strategies in many rural communities across the Global South: most LWM strategies are text-based and most text-based approaches to learning are ineffective.

Kimani’s point is that traditional approaches to learning don’t always resonate with learners. This is the challenge facing LWM approaches in the rural communities where Kimani lives and works: LWM initiatives tend to be derivatives of formal classroom environments and are misaligned with the pedagogical needs of rural adult learners accustomed to educational models based in non-formal small group dialogue.

Kimani is no stranger to the requirements of learners in rural East African communities. He is an expert small group trainer, who for 15 years has been helping adult learners across East Africa use dialogue to facilitate learning within non-formal small groups. Therefore, he knows that typical LWM approaches fail to employ community-based ground rules like kujengana (‘to build up’ in Swahili) and miss an opportunity to root the pedagogy within the lived experience of the small groups.

This is further compounded by the myriad of infrastructure and economic challenges faced in these communities, where there are few smart phones, pay-as-you go data usage models, and limited internet access. Too often the inherent possibilities of mobile technology focus on the capabilities of the technology and do not adequately conform to the affordances available to rural learners, a crucial oversight when working with rural East African residents who face technological access challenges.

Such challenges raise a simple yet difficult question: can a LWM approach address this misalignment?

A learning with mobiles platform designed for rural communities

TIST farmers in Eastern Tanzania practicing a practice called ‘Conservation Farming’ after watching demonstration videos and engaging content in the TIST Learning Centre.
Photo Credit – Mary Gemela

An ongoing Design Based Research project, started as my doctoral project and implemented in partnership with TIST, has sought to answer this question. Over the last three years, we’ve attempted to align non-formal dialogic learning with mobile learning pedagogy while considering the technological affordances available to mobile phone users in rural East Africa.

Called the TIST Learning Centre, this free-to-use learning platform serves adult subsistence farmers across Kenya, Uganda, Tanzania, and now southern India. TIST Learning Centre breaks away from past forms of knowledge production and content delivery by reinterpreting how we support and facilitate learning in the rural Global South, with a focus on:

• Offline access
• Low bandwidth functionality
• Economic resource constraints
• Diversity of participant educational background
• Face-to-face small group dialogic pedagogy

The platform is developed around a ‘bring your own device’ (BYOD) approach to scalability and requires no additional hardware to operate. Learning content is available in English and Swahili, with translation into indigenous tribal languages and Tamil coming soon, and it is presented in text and aurally through recordings in local dialects. The current suite of educational content, developed by TIST participants, is primarily focused on agricultural education, the carbon cycle and carbon markets, and TIST programmatic. Learners primarily engage with content to improve farming techniques and to explore the TIST programme. Efforts to expand the range of content are ongoing, with an expressed desire from learners for a greater range of topics such as theology, literature, and science.

Enhancing small group dialogue through mobile learning

The face-to-face interactions take place between peers within a small group, where knowledge is often shared through demonstration, such as a local farmer showing fellow small group members how to create green manure. These demonstrations, traditionally done in situ at a small group member’s local farm, are where participants create content for the platform.

Building from long-standing oral traditions of learning, Kimani and his colleagues at TIST have honed a set of ground rules for supporting dialogue amongst these small groups:

• Rotating leadership at every group meeting
• Focus on facilitation and servanthood rather than monologic domination
• A unique pedagogical construct known as kujengana, which is the act of small group members verbally recognising the positive contributions of the current leader.

The act of rotating leadership allows each member to experience this positive feedback in the form of kujengana on a regular basis, encouraging confidence and fostering a sense of purpose and community.

The TIST Learning Centre supports and enhances this peer-to-peer exchange by allow participating farmers to capture demonstrations via mobile video recording, which is shared via WhatsApp. These videos are then incorporated into the modules within the TIST Learning Centre, expanding their reach and scope. The Learning Centre provides an opportunity for farmers to share knowledge beyond the geography of their respective rural community and has become a valuable tool to capture and disseminate local knowledge.

Small groups in Eastern Tanzania using side-loading and the Hub model to share and discuss digital content. Photo credit – Mary Gemela

Using small group dialogue to overcome the digital divide

There are pernicious digital infrastructure challenges throughout the Global South. We have found that 60 percent of TIST Learning Centre users access the content on behalf of a small group, simply because connectivity is too expensive and coverage too sparse for each individual to access content on their own devices. When one person downloads the content, it allows individuals to share costs while minimising the need to travel long distances to obtain connectivity.

For example, Joseph, a TIST member in rural Northeast Uganda, lacks reliable network access in his home community. Therefore, he and his colleagues take turns busing each other’s smartphones to a central town once a week to connect to the network and send and receive email. Such usage habits have profound ramifications for the design of mobile education platforms

In fact, we’ve characterised this new approach to mobile learning as the hub model. In the hub model, a single user accesses learning content on the LWM platform and then shares the content a small group (the hub), either directly through dialogue or via an affordance colloquially called ‘side-loading.’ Side-loading involves downloading content to a mobile device and sharing material with other handsets through either a Bluetooth connection or a micro SD card. The hub model allows digital education to function within the limits of digital access in the Global South.

A new LWM pedagogy supporting learners in rural East Africa

However, the hub model also offers a unique benefit: when the side-loaded content is shared amongst small group members, the material acts as a catalyst for dialogic interactions within the small group. As a pedagogy, it leverages indigenous oral traditions along with simple technological affordances to overcome. This is, to again use the words of Kimani, how we make education ‘sticky’ in rural East Africa.

A TIST farmer and trainer in Kenya explains the importance of this approach…

Further research for LWM in the Global South

The LWM platform now has 800 active users and is steadily growing. Analysis of the efficacy of this project has yielded a set of Design Principles for use in LWM approaches in rural communities in the Global South. Ongoing research will test and refine these Design Principles through a deeper assessment of the quality of dialogue produced by this model, investigating the extent to which the model enhances conversations, supports creative problem solving, and achieves long-term learning objectives. Future developments underway include a WhatsApp-based ‘edubot,’ developed in partnership with Meta and InfoBip, which will investigate the ways in which AI can assist in supporting dialogic education in this context.

Findings from this research suggest that adoption of indigenous approaches to dialogic learning can be effective for LWMs in rural East Africa while providing a mechanism to mitigate the digital divide. These findings offer both practical and theoretical insights to researchers and practitioners exploring the use of mobile phones as a tool for learning within rural communities of the Global South, and the concept of kujengana introduces a compelling pedagogical construct for supporting dialogue across all contexts.

Relevance to the Digital Education Futures Initiative

This model is indicative of the types of projects supported by DEFI: digital education should not be about making text-based material more efficient to consume, it should be about leveraging technology to create learning experiences that are immersive, experiential, dialogue-driven and transformative. As DEFI looks to the future we seek to guide the creation of a more inclusive global education system that transcends wealth, class, geography and time. As DEFI grows it will continue to support approaches to digital education which use the affordances of technology to expand interaction and facilitate human flourishing.

A digital human knowledge and action network of health workers

Challenging established notions of learning in global health

When Prof Rupert Wegerif introduced DEFI in his blog post, he argued that recent technologies will transform the notions and practice of education. The Geneva Learning Foundation (TGLF) is demonstrating this concept in the field of global health, specifically immunization, through the ongoing engagement of thousands of health workers in digital peer learning.

As images of ambulance queues across Europe filled TV screens in 2020, another discussion was starting: how would COVID-19 affect countries with weaker health systems but more experience in facing epidemic outbreaks?

In the global immunization community, there were early signs that ongoing efforts to protect children from vaccine preventable diseases – measles, polio, diphtheria – would suffer. On the ground, there were early reports of health workers being afraid to work, being excluded by communities, or having key supplies disrupted. The TGLF quickly realised it had a role to play in ensuring that routine immunization would carry on in the Global South during the pandemic and then to prepare for COVID-19 vaccine introduction.

Peer learning vs hierarchical, transmissive learning models

Since 2016, TGLF had been slowly gaining traction in the world of immunization learning, with its digital peer learning programmes for immunization staff. These programmes reached around 15,000 people in their first four years, before the pandemic, about 70% of whom were from West and Central Africa, and about 50% of whom work at the lowest levels of health systems: health facilities and districts.

The TGLF peer learning programmes were developed as an alternative to hierarchical, transmissive learning models, in which knowledge is developed centrally, translated into guidance by global experts, which is then disseminated through cascade training.

In the hierarchical model, health workers are merely consumers at the periphery of the process. COVID-19 brought the inadequacies of this approach into sharper focus, as health workers dealt with challenges that had not been foreseen or processed through existing guidance.

No technical guidance could address every scenario health workers faced, such as reaching the most marginalised communities or engaging terrified parents at a time when science had few reassuring answers. They needed to be creative and empowered to find their own solutions. Health professionals learned to rely on each other as peers, learning from each other how to negotiate many unknowns, without waiting for the answers provided by formal science.

The TGLF approach quickly demonstrated its usefulness in connecting peers during the pandemic. In 2020, the number of platform users doubled to 30,000 in just six months (compared to four years to gain the first 15,000 users) and has now trebled to 45,000.

Addressing Covid-19 impacts through challenge-based learning

The foundation of the TGLF approach was the COVID-19 Peer Hub, an 8-month project based on challenge-based learning, which challenged individuals to give and receive feedback as they collaborated to:

• Identify a real challenge that they were expected to address in their everyday work
• Carry out situation analysis, and
• Develop action plans that are peer-reviewed and improved.

The Peer Hub was inspired by the works of several of academics who helped create the Foundation: Bill Cope and Mary Kalantzis, and their technological implementation of “New Learning;” George Siemens’ learning theory of connectivism; and Karen E. Watkins and Victoria Marsick’s insights into the significance of incidental and informal learning.

The Peer Hub demonstrated the creation of a “human knowledge and action network” formed through both formal and informal peer learning combined with ongoing informal social learning between participants. The network was built on the principle that participants were themselves experts in their own contexts, and creators, rather than consumers, of knowledge. Front-line health workers suddenly had the legitimacy and ability to share experiences with their peers and experts from around the globe.

In the first ten days, COVID-19 Peer Hub participants shared 1224 ideas and practices through the Ideas Engine, an online innovation management tool.

Results of peer-led, challenge-based learning interventions

More than 6,000 health workers joined the TGLF COVID-19 Peer Hub, where they:

• Documented and shared 1,224 practices and ideas to maintain routine immunization through the Ideas Engine;
• Developed 700 peer-reviewed action plans, informed by ideas and practices shared through the Ideas Engine;
• Learned to support each other in implementing these plans during a four-week “Impact Accelerator Launchpad;”
• Responded to concerns about vaccine hesitancy in the face of COVID-19 vaccine introduction, by developing a peer-reviewed case study documenting a situation in which they had helped an individual or group overcome their initial reluctance, hesitancy, or fear about vaccination. The resulting qualitative analysis – unique in accessing so many firsthand narratives from health workers – was produced by 734 participants.

Tracking movement of practices and ideas shared through the Ideas Engine between countries

Because while health workers responded positively to opportunities to connect, learn and lead with one another, TGLF is very much a new entrant in a well-established institutional learning environment for global health. Here are some questions we’ve developed as TGLF challenges established norms and ways of working:

• How would you feel as a global expert if you were asked to give up your role as ‘sage on the stage’ to be a ‘guide on the side’ to thousands of health workers?
• Can self-reported data from thousands of health workers evaluated by peers be trusted more or less than a peer-reviewed study?
• What does ubiquitous digital access mean for training programmes that have previously incentivised learner participation in face-to-face events through payment?

Working with DEFI and other similar institutions, TGLF looks forward to:

• Exploring and demonstrating the credibility of what we do through critical independent research and commentary
• Demonstrating the potential of our approaches to large institutions and their donors;
• Developing a bigger picture of how other sectors are adapting to the affordances of digital learning technologies;
• Meeting others innovating in digital learning to be inspired and cross fertilise.

­We look forward to fruitful dialogues!

New collaboration with Dr Ingunn Ness and Dr Rupert Wegerif

Dr Ness is also a co-editor and contributor to the new book, Dialogic Pedagogy, Creativity, and Learning –along with DEFI founder Rupert Wegerif.

The main editor of the book is Olga Dysthe, pictured below receiving the King’s Medal of Merit for her work on dialogue.

Preparing for an uncertain future

Dr Ness started her presentation by stating that “one of the main purposes of an education system is to prepare students for an (uncertain) future. This was music to the ears of a DEFI audience, as it matches a key idea from our recently-published DEFI Future Skills report – the perception that the uncertainty will affect life in the future more so than in the past.

According to Dr Ness, the curricula in Norway has been updated to address this challenge, with creativity, collaboration and communication skills emphasised. However, she then pointed out a key question and the driver behind her research: In order to teach for creativity, we first need to know: what is creativity?

What is creativity?

Put simply, creativity is “to create something new,” says Ness, and, “we need disciplinary knowledge to have new ideas which we can be creative about.” She would return to this concept of disciplinary knowledge later in her presentation.

She went on to break down creativity into “Little c” – personal, “what if” breakthroughs – and “Big C,” which includes major, society-level breakthroughs. Dr Ness would like to see young learners encouraged to explore their little C creativity, so that they can generate Big C ideas later on.

As she continued reflecting on creativity, Dr Ness highlighted a tendency to distinguish between artistic creativity and scientific creativity, despite the underlying similarity. While acknowledging that creativity can differ across groups and domains, Dr Ness contends that Einstein and Picasso would have had nearly identical psychological processes!

Case study – Stimulating creativity through polyphony

As Dr Ness took us into the work behind “The Room of Opportunity,” her passion for the topic shone through. In fact, she admitted that she collected more data than was necessary, as she was so excited to observe and capture the work being done by the study groups.

Her study looked at 6 phases of group work, which can allow the group to move from a challenge or need through to a solution. As seen in the graphic below, the first three phases combine to build the knowledge platform, fulfilling the need for disciplinary knowledge which Ness referred to in her thoughts on creativity.

The second half of her model is where the ideas are developed. And yet, Ness identified that it is the three central phases – Polyphony, imagination, idea formulation – which form the “Room of Opportunity.” This metaphorical room is where groups can spend time, moving through the two or three of the phases repeatedly, while the creative process generates ideas that may lead to a solution.

Quite simply, groups wanting work together and build knowledge will need to have openness, curiosity, respect, trust and psychological safety. Ness used Kahn’s (1990) definition of psychological safety as “to feel safe in the team and not afraid to contribute,” then simplified this further to “not afraid of asking silly questions.”

Facilitating creative processes

In her third study, Ness stepped away from the six-phase model to look at how creative knowledge processes are facilitated. Her first observation was that imagination and creativity are stimulated by tension, both between disciplines and group members.

Therefore, she outlined four complementary roles in group dynamics that, when balanced across the activity, can create useful tension and facilitate creative thinking.

A model for student dialogue

The result of her work with students is the STEPRE model. As with the adult groups, there are six phases, but the names have largely changed. Polyphony remains, of course, and Ness explained that the polyphonic tasks allowed students to “practice creative knowledge processes” through dialogue. Polyphonic tasks also allow the teacher to practice letting go of control, which Ness pointed out is losing control of the student group.

Conclusion

In under an hour, Dr Ness managed to take her audience through nearly 70 slides, introducing herself and Bergen, Norway, exploring the basics of creativity and future skills, then looking at her research work with both career and student groups. That she was able to explain both the foundations of her work and results of her research in such a short time is a credit to her own dialogic skills as well as her expertise of the subject matter!

Dr Ness closed by reminding the audience that her phase models are meant as inspiration and a way to structure dialogic teaching – not as a rigid model – something that she demonstrated by presenting both the six-phase model for professional groups and the STEPRE model for use in schools.

She also pointed out that education needs to provide training not just in how to engage in creative dialogue, but in how educators and group leaders can create a framework for dialogue. And finally, she mentioned another book of interest to the DEFI audience – Creative Learning in Digital and Virtual Environments.

We greatly appreciate Dr Ness’ time with our group and, as Genevieve says below, look forward to her next seminar!