WS #270 Understanding digital exclusion in AI era

Summary

This discussion focused on digital exclusion in the era of AI, particularly in relation to marginalized groups and developing countries. Panelists explored challenges and potential solutions for bridging the digital divide between the Global North and South. Key issues identified included lack of infrastructure, limited internet access, language barriers, and inadequate digital literacy, especially in rural areas.

Speakers emphasized the need for human-centered approaches in AI development, involving local communities in design processes. They stressed the importance of creating content in local languages and investing in education and capacity building programs. Success stories were shared, including youth-led initiatives to improve digital literacy in schools and communities.

The discussion highlighted the lack of AI policies and regulations in many countries, calling for international collaboration to establish universal guidelines. Panelists also addressed the challenge of including both youth and older populations in AI adoption. The potential of AI to support sustainable development goals was discussed, though concerns were raised about the risk of data-poor languages being left behind in AI development.

Participants agreed that multi-stakeholder collaboration, including governments, private sector, and civil society, is crucial for addressing digital exclusion. Key factors identified for ensuring digital inclusion in AI included education, public awareness, capacity building, and human-centered approaches. The discussion underscored the urgency of taking action to prevent further widening of the digital divide as AI technologies advance.

Keypoints

Major discussion points:

– Digital exclusion and the AI divide between Global North and South

– Challenges of AI accessibility in rural communities and for marginalized groups

– Need for inclusive, human-centered design of AI tools and policies

– Importance of education, capacity building, and digital literacy

– Role of youth and international collaboration in shaping AI policies

Overall purpose:

The goal of this discussion was to explore digital exclusion in relation to emerging technologies, particularly AI, and discuss ways to make AI more inclusive and accessible, especially for underserved communities and the Global South.

Tone:

The tone was largely constructive and solution-oriented. Speakers acknowledged significant challenges but focused on sharing ideas, success stories, and recommendations for improving AI inclusivity. There was a sense of urgency but also optimism about the potential for positive change if the right steps are taken. The tone became more interactive and collaborative when audience members joined the discussion near the end.

Speakers

– Moderator: Facilitator of the discussion

– Maxwell Beganim, African coordinator for Anglophone region for Open Knowledge Foundation Network, former steering committee member for IGF and Youth IGF in Ghana, former executive member of Internet Society Ghana chapter

– Jaewon Son, Doctoral Researcher at Karlsruhe Institute of Technology

– Speaker 3: Doctor from Chad

– Bendjedid Rachad Sanoussi: Technical expert

– Speaker 4: Maxwell, researcher on AI

Additional speakers:

– Florent: Professor of law at the University of Zurich

– Ram Mohan: From Identity Digital and Critical Infrastructure company

– Mbongi Nimsimangasori: Postdoctoral researcher with the Johannesburg Institute for Advanced Study in South Africa

Digital Exclusion in the Era of AI: Bridging the Divide

This discussion focused on the critical issue of digital exclusion in the era of artificial intelligence (AI), with particular emphasis on its impact on marginalised groups and developing countries. The panel explored the challenges and potential solutions for bridging the digital divide between the Global North and South, highlighting the complex interplay of technological, social, and policy factors that contribute to this growing disparity.

Key Challenges

The speakers identified several key challenges contributing to digital exclusion:

1. Infrastructure and Access: There is a significant lack of affordable and reliable internet infrastructure in many areas, particularly in rural regions of developing countries. This fundamental gap in connectivity forms the basis of digital exclusion.

2. Language Barriers: The dominance of a few major languages in digital content and AI development poses a significant barrier to inclusion. As noted by the doctor from Chad, explaining concepts like artificial intelligence in local languages can be challenging, highlighting the need for localised content.

3. Digital Literacy: Limited digital literacy, especially in rural areas and among older populations, hinders the adoption and effective use of AI technologies.

4. Policy Gaps: Many countries, particularly in the Global South, lack comprehensive AI policies and regulations, creating uncertainty and potential risks in AI development and deployment. Jawan specifically highlighted the lack of AI regulations in universities.

5. Data Inequality: Ram Mohan emphasized the growing divide between “data-rich” and “data-poor” languages, which could lead to the marginalisation of minority languages in AI development.

Proposed Solutions and Approaches

The discussion yielded several potential solutions and approaches to address these challenges:

1. Inclusive AI Development: There was strong agreement among speakers on the need for inclusive, human-centred approaches in AI development. Bendjedid Rachad Sanoussi emphasised the importance of putting humans at the centre of AI design to ensure respect for human rights. This approach involves engaging local communities and end-users in the design process of AI systems.

2. Localisation and Language Development: Speakers stressed the crucial role of creating content in local languages to improve digital literacy and make AI more accessible. This includes developing AI tools and interfaces in indigenous languages. The moderator highlighted the need for multilingualism in AI, citing Tanzania’s linguistic diversity as an example.

3. Education and Capacity Building: Investing in education and digital literacy programmes was seen as essential, particularly in rural areas. Maxwell shared success stories of youth-led initiatives like K-Works for Schools, which have helped bridge digital gaps by providing computer labs, internet access, and digital skills training to students in Zimbabwe.

4. Infrastructure Development: Speakers emphasised the need for technology transfer and infrastructure development from the Global North to the South. Rashad suggested using community networks and low-cost satellite technologies to improve internet access, as well as leveraging public-private partnerships to expand infrastructure.

5. Multi-stakeholder Collaboration: There was consensus on the importance of collaboration between governments, the private sector, and civil society in addressing digital exclusion. This collaborative approach was seen as crucial for developing effective AI policies and governance frameworks.

6. Youth Involvement: The discussion highlighted the importance of including young people in shaping AI policies and tools, recognising their role as both users and future leaders in the field.

7. Promoting Open-Source AI: Rashad suggested promoting open-source AI platforms to increase accessibility and foster innovation.

8. Environmentally Friendly AI: Rashad emphasized the need for AI tools to use less energy and be more environmentally friendly.

Specific AI Initiatives

Several specific AI initiatives were mentioned during the discussion:

1. Drone Tech Project in Chad: A project using drones for medical deliveries in remote areas.

2. AI Translation Tool in Benin: An initiative to develop AI-powered translation for local languages.

3. K-Works for Schools in Zimbabwe: A youth-led project providing computer labs and digital skills training to students.

Areas of Agreement and Disagreement

While there was broad consensus on the importance of addressing digital exclusion, speakers emphasised different primary factors and approaches:

– Jawan focused on the need for technology transfer and universal guidelines for AI use.

– The doctor from Chad stressed the importance of developing local language content for digital literacy and creating safe online spaces for internet users.

– Bendjedid Rachad Sanoussi highlighted the lack of affordable infrastructure and the need for human-centred design in AI, as well as the importance of affordable and energy-efficient devices.

– Maxwell emphasised the success of youth-led initiatives in bridging digital gaps.

Unresolved Issues and Future Considerations

Several important questions remained unresolved and warrant further discussion:

1. The balance between waiting for government policy on AI and allowing industry to lead development.

2. Strategies for preserving and developing AI for minority languages with small speaker populations.

3. Ensuring older populations are not left behind in AI adoption.

4. Addressing the growing divide between data-rich and data-poor languages in AI development.

5. Developing Afrocentric AI tools, as suggested by an audience member from Zimbabwe.

6. Investing in public aspects of development, as emphasized by the doctor from Chad.

Conclusion

The discussion underscored the urgency of taking action to prevent further widening of the digital divide as AI technologies advance. It highlighted the need for a multi-pronged approach that addresses infrastructure, education, policy, and inclusive design. By fostering collaboration between diverse stakeholders and prioritising human-centred approaches, there is potential to harness AI as a tool for sustainable development and social inclusion. However, significant challenges remain, particularly in bridging linguistic and cultural divides and ensuring equitable access to AI technologies across different regions and populations.

Speaker 1: My name is Maxwell Biganim and I am in Ghana. I currently serve as the African coordinator for Anglophone region for Open Knowledge Foundation Network. I also used to be a steering committee member for IGF, Youth IGF in Ghana, and also an executive member of the Internet Society Ghana chapter as well. I am very happy to be here this morning. Thank you.

Moderator: Thank you very much everyone. So before we move further to our discussions, can you hear me? I hope you can, okay. Yeah, I can’t hear myself. So before we move further into our discussion, so we all, we have always recently have been in these discussions about emerging technologies. We have all this AI, the computational, all these technologies and all that, but then we have been facing an issue that the marginalized group and let’s say people also from the global south have left out when it comes to all these emerging technologies. So there is that this divide that is currently existing, especially when it comes to women, people with disabilities, but also people who are living in rural areas. So in this session today, we are going to explore this now digital exclusion that is currently existing, especially in relating to all these emerging technologies and specifically AI. So the whole discussion is going to revolve around that. So I’m going to be asking questions to my panelists. I mean, they’re going to be asking, I mean, responding to different questions that I’ll be asking them. So just to start with a discussion. So I want to start with my on-site speakers. Start with you, Jawan. In your opinion, what is the most pressing challenge of digital exclusion in AI era? And how can we address it.

Speaker 2: Thank you for your question. I think due to the AI divide between Global North and South, without any cooperation between Global South and North, it will be further that it’s not only that Global South would not be able to proceed further and also bring more economic divide and so on. So I was thinking that like more policies on technology transfer, for example, where the Global North would bring some of the infrastructure technology and helping them to understand how to use the AI skill so that there will be more workforce where they can utilize such a technology and develop further into economic prosperity will be very beneficial in this case. And also I was thinking that one of the main challenges in these days in AI is that there are no much discussion between people and also technology companies that are making those technology where when there is any agreement about how the data will be used and so on, usually it’s between the companies and country and there are not much discussion about how do we bring all the policies to the different stakeholder and the public. So I think there should be more discussion how do we involve all the multi-stakeholder including the public. Yeah.

Moderator: Thank you so much, Xiaowan. So I want to move to you, doctor. So in 2030, we were expecting like to be there when it comes to the sustainable development goals, right? And so how do you think AI can help achieve the SDGs while addressing digital exclusion, especially for marginalized group?

Speaker 3: I think it’s important to keep being, I mean, optimistic, but from now till 2030, only six years, I mean, left is quite a challenge, I’m saying so. And especially when it comes to reflect realities in the global south, it’s quite impossible to be honest. But we have to work very deeply in education, for instance, to give the possibility to local communities education skills in digitalization, for instance, or even in different sectors. For example, in the sector of economy, creating jobs, employments, but we have also to invest on how they got infrastructure because you see how the north is working to equip, I mean, very importantly, academic institutions and provide the needed, materials to younger generations to be educated, I mean, in the context of the current AI and all these innovations, I mean, aspects. So in countries like Chad, for instance, where I come from, we need to invest more into public aspect of development, for instance, contribute to create awareness, contribute also to regulate the use of digital devices. But also to protect, I mean, internet users, because it’s important to create a safe place where internet users can feel free to get to be educated or to work online or to be consulted by any doctor or any institutions over the world. It’s important, I mean, to create the same space when it comes to access. how do you get to be educated. All right, so I wanna pick up from the same point that you just said, you talk about employment. And so now I wanna understand, now we have all these rural communities, what do you think are the key challenges? Let’s say taking example of Chad, what do you think are the key examples? I mean, are key challenges in ensuring that the AI tools reach even the rural communities that in Chad? Like, what do you think are the existing challenges? I have the honor to discuss with Honorable Emma Tsoefel, for instance, the ICT Minister of Namibia, who is here. I am very happy to meet her back again. We discussed about this issue several times and it is a challenge because this one, I mean, concern more our local long years because we have communities that only speak the local languages while internationally we use devices or technology into other languages internationally recognize it. So first of all, we need to create in term of digital literacy content using our local languages for, I mean, to allow or to help communities to understand what the message is. For instance, I cannot explain to someone from my community what artificial intelligence means. It could be very difficult to get a word that I can use it to explain what even technology means. But these people, despite the lack of education, they have today a smartphone, tablet, they use WhatsApp, they use Facebook, for instance, they get to be online without knowing anything. So it’s important to develop our. local languages and via this we can create content that our languages are in the top of the program and create capacity building program for instance even to those who have never been to school, the seniors, I mean people of the third age and younger generation, this one is the most important things because our governments and private sectors have to invest more into education by providing I mean the materials needed, equip them and then promote a very sustainable capacity building program even to lecturers and students who are in research program because even these people are disconnected today from the reality. So how can we imagine it will be possible to teach those people who have never been to school if we have not access to information, education as a high level priority?

Moderator: Wow, thank you so much. I think I can really relate to that. I like the point about multilingualism in AI. I come from Tanzania for the record and in my country we have almost 121 tribes and each tribe have their own languages and our national language is Swahili for example. Among all those 121 we still have people who don’t even know the national language so they don’t speak Swahili. So at some point if we would say someday we come with an AI tools that are in Swahili for example they wouldn’t understand because they don’t even speak Swahili so we have been having challenges of making like reaching the rural communities having access to all this digital knowledge, this AI knowledge and all that. So I think there’s really a big work that we have to do for countries like mine. I want to come to you, Rashad. So you have been having technical experience and all that. So I want you to talk about how can we ensure AI and other technologies can be designed well to ensure that there is inclusivity and accessibility, especially for undeserved communities.

Bendjedid Rachad Sanoussi: Okay. Thank you so much, Miriam, for the question. I also like the point of making it inclusive because in Africa, we face a lot of issues when it comes to technology, and we need to have access to Internet, and many people don’t have access to Internet. Like when we are talking about also access, we have the lack of affordable and reliable Internet infrastructure. So in my country in Benin, many people don’t have access to Internet. Even those who have access to Internet, they just use it for social media. So to address that, maybe we can set some community network or maybe some low-cost satellite technologies that can help everybody to have access to Internet. So we need a lot of collaboration to fix that. So to come to your question now about how we can make AI more inclusive and accessible for everyone, firstly, we need to have access to Internet, so access to the technology. It’s why I was talking about making more infrastructure so that everybody can access to Internet. So to ensure that inclusivity and accessibility really fit in our life today, AI and emerging technologies. we need to design them with a human-centered approach, like we need to put the human at the middle so that this technology can respect human rights. This means that we need to involve local communities and also users to design the system so that they can know how it works, and they can also make contribution to design the application and also the technologies as well. I can take one example. If we want to design an AI tool, we can support like use AI to translate our local language. We need to take in account the diversity of African culture and also the socioeconomic challenge. We are talking about Tanzania. We have a lot of tribe, same in Benin, same in Chad. We need to take those in consideration when we are designing some AI tool as well. Another aspect is more about how we can make AI to use less energy so that it can be more green. When we are talking about AI, sometimes it’s about a lot of algorithm and also a lot of energy use. Tool like AI model need to be optimized for mobile device so that we can respect the environment and those AI tool will be green as well. Also, another thing we can do is to promote open-source AI platform so that everybody can have access to this platform, and know how it works, and also can contribute. By doing that, communities, especially youth, innovators, and startups can co-develop. or develop a cost-effective solution for specific issues we are facing in Africa and other parts of the world as well. So that will be my contribution. Thank you. Thank you.

Moderator: Thank you so much, Rashad. I really liked the point when you talked about the human-centred approach when we are developing all these AI systems, because assuming, if you assume the needs of someone without clearly understanding the needs that they really have, and then you develop solutions for them, it may not meet their needs, because you have created something that may not be useful for them. So it’s really important that we engage all these communities. Now we have talked about all these marginalised groups. Have them in the room, understand their needs, like, what do you want? How can we assist to ensure that you guys are not left behind in all these discussions and all that? So we have talked about all these challenges. Now I want to, for you, Maxwell, now I want you to talk about the success stories. Can you share success stories where youth-led initiatives have bridged digital gaps and improved inclusion?

Speaker 4: All right, thank you very much. I think this is, I’ll just share my own success stories on some of the projects that I have led in that. And thank you to the wonderful three speakers that I’ve spoken. And I think I’m also so much educated right now. Yeah, so when it comes to the success stories, I think it’s first important for us to understand the problems that exist, which is the digital divide in many, most of the communities that I have had the opportunity to work with. Firstly, there’s this project that I started with a colleague, and it is called K-Weeks for Schools. And I realized that most of the senior high schools in Ghana, for example, had computers. and their computers were not functional. And those that were also functional were just left abandoned. There were no softwares, there were no applications that students could engage. And so we came up with this project called K-Works for Schools, where first we go into the institutions and then we train teachers to understand the concept of digital literacy. We take them through just about an hour of digital literacy, how they can understand it, how they can also look at all the parameters of digital literacy. And as teachers, don’t forget that they are custodians of knowledge. And so once the teachers appreciate digital literacy, it becomes easy to transfer in their classroom. So then after we then decided to do the installation of K-Works. K-Works is actually an offline educational resource that enables students to have access to contents like Wikipedia, TED, and all of that. So we also took an advantage of also training the students. So because the students were very many, we decided to sample the class reps so that they will also serve as ambassadors or digital ambassadors to their other colleagues. And we train them on how to use the digital literacy, the K-Works, and also had digital citizenship masterclass for them. Now we realized that based on just one that we started, a lot of schools started reaching out to myself and my colleague to expand this project. So we decided to do it in addition for other senior high schools. And it was gaining the traction and I’ll share some materials later. Then we decided that no, the senior high schools cannot always only be the ones that benefit. And we decided to take it back to the basic schools where we train them on digital literacy. Some of the schools that we actually went, they didn’t even have computers, but we also understand. that once people have the understanding, it shifts their mind in thinking along that tangent. So that was also very important and a very good success stories. As I tell you, K-Wigs has moved from Ghana to African region and now we are even training ambassadors in various countries to also do some of this project as well. So that is just one of the projects that we did in bridging the digital gap as well. And also there is also a project called The Life Project that we worked on with Paradigm Initiative. And that was to also target atypical people, people that ordinarily do not have access to digital literacy or digital skills where we’re able to mobilize people who have left school, people who are no more in school to train them to have some of these skills. And I think it was very, very, very successful project because now some of them are into graphic design and some of them are into a lot of areas when it comes to digital device. So because of time, these are the two fundamental or the two key success stories that I want to share with us all in terms of bridging the digital divide.

Moderator: Wow, thank you so much. Congratulations for the great work that you’re doing. Now, I wanna open this few minutes to the floor. If you have, can you please share if you have in your specific countries, if you have policies that are helping regulate AIs in your AI in your countries, is there something like that that anyone on the floor that would like to share? Anyone from the floor sharing from your respective countries, any policies that ensure that there’s equitable access in AI, or maybe we can start, do you even have any policies that you would like to share? Okay, thank you so much for the great work that you’re doing. I’m gonna open it up to the floor. have policies for AI in your countries? No. Okay. So there is a great work we need to do. So maybe I can come back to you, Jawan, now. Now we have seen here, let’s say from the floor, my country, we don’t, they don’t. So how can we ensure like international collaboration and help in ensuring that we have proper regulations of like AI in our respective countries and all that. Yeah.

Speaker 2: Thank you. Yeah. In my university in Germany, even like, while it’s the first university that actually had the email server in Germany, however, they also doesn’t have any regulations about what is acceptable to use AI or not. For example, in research or your studies or tests. So I always thought like, why? Like, we are always talking about like, we should have all the adopter in AI. We should like make them aware of it as so on. But what are the, I don’t know, the universities and like international organizations doing for them to know what is right and wrong? I think in this case, the international cooperation or I would say the international alliance comes in hand where there should be a universal guideline and regulations about how the people are knowing like what’s okay and not to use AI, for example. For example, I think yesterday in the youth session, they were talking about AI and education. So that while we talk about marginalized people who doesn’t know how to use AI, however, are we ready for them to actually use AI in universities or in companies where like just providing the AI is. not the end of everything. However, I think it begins with all the problem about the privacy and everything where we do not have actual like internet governance ready for them to include everyone in the end user having a right to say about how they use the data or not. So I think in that case, international cooperation should really focus on first thinking of how do we integrate all the end user and stakeholders in this governance. And second of all, having the proper regulations have a universal guideline about how we deal with such a problem.

Moderator: Great. Thank you so much. I’m coming back to you, doctor. You’ve done some research on AI, right? I want you to help me share examples where AI was successfully implemented to solve this, the digital exclusion. Is there any instance maybe you can help us share?

Speaker 3: Well, very largely when it comes to achieve this, for instance, using or benefiting from the use of AI, for instance, in Chad, we have drone tech, which is an initiative that provides support to rural communities when it comes to, I mean, assist them in social, I mean, how do you call, issue, for instance, or when there is a social crisis, for instance, we we intervene to bring, I mean, resources or materials, I mean, to help the community. So for instance, during the crisis between Sudan and, yeah, in Sudan, for instance, we have refugees that were in the border of Chad. and we’re seeking for help. So we use it, I mean, those drones to bring, I mean, support today to the refugees, but also in terms of education, we have initiatives that help, I mean, children to understand better alphabet and vocabulary or even talk to people, for instance, in different languages using these tools. So it’s important to invest in this issue because we have so many challenges to address. And for that, we need a multi-stakeholder collaboration because none of the stakeholder alone can act and, I mean, address these challenges only if we contribute together as one in a strong multi-stakeholder, I mean, body. So I call the, or even urge, I mean, the UN and its partner and our governments, I mean, to take this one into serious, I mean, engagement, especially in countries of the South. It is essential, I mean, to align ourselves with today’s realities when it comes to also tackle the modernization challenges, we are a little bit far and we only solve this with digital inclusion, for instance, and digitalization of the society because I’m seeing the worry in 10, 20 years, perhaps we’ll be having perhaps AI tools that will be able to give 100 or let’s say 95% of consultation, let’s say medical consultation and even provide like a prescription while in countries of the South, developing countries. If you explain this to someone. it could be very strained. So we need to work a little bit hard to align ourselves with the realities in advance.

Moderator: Thank you so much. Thank you so much, Dr. Rashad. Now we have talked about all these challenges. We have seen we don’t have proper regulations when it comes to AI within our specific countries now. But we have also seen the divide that is currently existing between the rural communities and the urban communities. So what steps do you think we can take to ensure that we are bridging the digital divide between the rural and the urban

Bendjedid Rachad Sanoussi: areas? Okay, thank you for the question. I think we have a lot to do. We have many issues when it comes to inclusivity. And from a technical perspective, I think we can maybe first explain the infrastructure because we need to have infrastructure first so that we can have access to the internet and those technologies. So to do that, maybe we can also do community work. You know, in Africa, many countries do not have access to 5G yet. So we can also, by expanding the infrastructure, working on 5G technology or develop some satellite-based internet. You know, Starlink is growing in Africa and it’s sometimes more affordable for some people in a remote area. So we can leverage on public-private partnership to do that as well. After expanding our infrastructure, we can also promote energy access. Like, you know, in some rural communities we don’t have available electricity and when we talk about Internet and connectivity, it’s more about access to electricity. If you don’t have electricity, you cannot have Internet. So it’s really crucial to have access to energy so that we can promote digital inclusion and using also renewable energy like solar and wind to do that. Another way it may be to have affordable device and solution because the issues of money is also crucial. So we need affordable device so that people can buy this device and those device also can, should be energy efficient device so that it will not use a lot of energy. And we can also do a lot of local content like localize the content and service. Like, you know, we have issues about language barrier. So the solution should be more localized and we can develop a lot of content when it comes to our local language on those technology as well. And we can also leverage on our communities to do that. Like, if you want to do the solution, we need to do it with our community so that we can empower the local youth and also encourage the entrepreneur to do a lot. I see in the room we have many entrepreneur here. So it’s really important as an entrepreneur to build the solution that we really need to our communities with them. Thank you.

Moderator: Thanks so much Rashad. And since we are out of time, our time is almost over now, I want to go last to you, Maxwell. So what do you think? What role do you think young people can play in shaping AI policies and tools to ensure that they address the needs of diverse populations?

Speaker 4: Yeah, so I think that this is a very important question. I think first, we need to be very inclusive or in the design approach of all these policies around AI should be very inclusive, where young people do not only participate, but are involved in crafting some of these policies as well. And I think that for young people to be able to also be included when it comes to capacity building and capacity enhancement, how to even understand and leverage on the languages of AI as well, you don’t necessarily need to be a programmer or a deep developer or hardcore developer to be able to understand the parameters within which the artificial intelligence or AI works. And you would agree, or we would all agree that AI has come to stay. And so in order for young people not to be, maybe I’ll use this term as AI immigrants, they should be able to understand the parameters, it should be also included in mainstream education for young people to understand. When we talk about AI, then young people cannot think critically around some of the use of AI as well. So I think the inclusive approach is very important in crafting and shaping this in order to allow young people to be involved. And even at the high level discourse or at the high level, plenaries and conversations around this, we need to also get the input of young people in shaping the policies as well. Yeah, so these are the thoughts that come in mind. And also there should be programs that would be structured in a way to also make sure that the capacity of young people are built to understand AI.

Moderator: Thank you, Dr. Maxwell. So now we have hard panelists. I wanna open the floor again to my participants. Is there anything, any question that you have to the speakers, any contributions, something that you want to add from this ongoing discussion? Yes, please.

Audience: You mentioned a lot about young people, but I’m wondering as opposed to the young population, what can we do to the elder to make sure that they have equal access to AI or internet? Because a lot of older people nowadays, they have ingrained ideology opposing AI, and they also due to their physical or cognitive issues, they have trouble getting access to the internet. And some of them are also lacking enough money. So how do you think we can address that issue? Who wants to respond?

Speaker 3: Thank you for the question. Well, for those who are professional, I think it would be a little bit easy to support the process, because we can just, I mean, call companies to have like a capacity enhancement program. to these professionals, any time there is a certification or any programs comes out, we can organize such capacity enhancement program to reinforce their capacities in certain subject. But for those who are unprofessional or never work or don’t have a quality education, we can work on digital literacy by creating, as we say, content using local languages to help them understand the use of these tools or, I mean, digital devices or whatever in program that can help them, I mean, in their age, in case. Yeah, so that’s why it’s important to keep working very hard to initiate a program that support younger generation to be very well-educated and then align itself with our today’s realities or the 21st and mid-century realities in general.

Moderator: Anyone who wants to add? Okay, so I think, but I think something that I wanted to add on, I think it’s really important to understand. I’ll say, let’s say from my country per se, we can’t go directly and start introducing AI because there’s still a very large group of people that are not digitally literate. So we have people that can’t even turn on computers. They can’t even do a simple Google search and stuff. So I think at some point to help, let’s say, with the elders in that kind of a situation, we might need to start with the basics. So we can’t go to AI if they don’t even get the basics of computers. They need to understand this, the basic stuff, and then moving forward, that will at least be easier. I guess. Yeah. Any other questions from the floor? Any contribution? Yes.

Audience: So thank you very much for these interesting contributions. My name is Florent. I’m a professor of law at the University of Zurich. So I come from a very privileged country, but I think it might be interesting also for you to know that we also have minority languages. We have a language in Switzerland which is spoken by less than 20,000 people. It’s even split up in different dialects. They have difficulties understanding each other. So developing AI tools for these language groups is a huge challenge, just because there’s a lack of data and lack of users as well. But maybe given that there are some developed countries that face similar problems, there is a chance to cooperate amongst continents on these, I think, very important issue. Great, insightful contribution. Yeah. Thank you. My name is Ram Mohan, and I’m with a company called Identity Digital and Critical Infrastructure. One of the things that I’ve been quite concerned about with AI and digital inclusion, especially in the area of languages, as the prior intervention was, I think we are now in an era of data poor and data rich languages. And I think if you look forward in the next five to 10 years, AI systems which train on language sets are going to train disproportionately on data rich languages. They’re going to lead data poor languages to the side. And I think if we don’t take action now, we’re going to start having a risk that the data-poor languages will disappear from the digital infrastructure and the people who speak and use those languages are also going to go away. So I’m quite concerned about that. I wonder if you have a perspective on it. Alice, do you have any contribution on that? It’s not a contribution, but perhaps I can add a little bit more to this question and ask our professor from Switzerland because when I was in Lyon during my PhD program, I used to go to Geneva where I’m more connected with certain programs. So do you think, professor, that AI use or let’s say digitalization can be an opportunity to develop minorities, I mean languages, I mean the less spoken, I mean let’s say languages, because if we work on that, we have no choice only to create programs that will develop these languages, add more vocabularies, create programs that is very limited this vocabulary in this language. So how do you think we can work on that?

Bendjedid Rachad Sanoussi: Thank you very much. Tough question. I’m not sure I do have any ideas how to work on that, but I fully agree that I think it’s important groups into this technology. Yes, okay. Because even right now there are risks to be left behind and there’s simply a danger of just these languages getting extinct because maybe as opposed to some in some of the African countries speak German or Italian on top, so they are able to communicate in other languages, which raises of the language getting extinct. To use that language into these technologies might be a means to promote it and give people chances to communicate technologies. So I think it’s super important, but no solution so far, I’m sorry.

Moderator: Okay, and contribution, yes, at the back. Can you please pass the mic at the back? Over there. Thank you for an interesting presentation, I must say. I think one of the aspects really is that I resonate with what you’re all saying.

Audience: My name is Mbongi Nimsimangasori. I’m a postdoctoral researcher with the Johannesburg Institute for Advanced Study in South Africa. So one of the issues is I originally come from Zimbabwe and one of the issues especially that we face in our country is that these, I don’t know if I may say policy contradiction, right? So the question is, should we wait for government to actually introduce policy or should we let the industry to lead, right? I must say that especially within the media sector who already have AI robots, right? And particularly from an institute called the CITE, it’s called the Center for Technology Innovation, right? It’s mainly a new site, right? They’re already developing an LLM for one of the main languages in Zimbabwe which is Isindere language, right? So that’s one of the critical aspects should we actually wait for government because government is actually delaying and for them to pass these policies, to pass these laws takes so much time, right? So, we have these institutes actually developing these specific, going ahead with AI policy and all. So, in Zimbabwe as well, if I may say that one of the resolutions in terms of consultations with the Internews and the Media Institute for Southern Africa was to sort of develop an Afrocentric, you know, AI tools, right? So, it’s one of the critical aspects that perhaps we can take a lead from as well, right? But I must say that we’re still far from actually developing policy, you know, considering the slow pace at which everything is actually going. But this site, as I mentioned, has really taken a very critical role in actually developing this LLM within, and including actually minority languages, especially within that. We had an uproar recently as well where that particular robot that could not even pronounce its Independent names or African names, and a lot of people are actually complaining around that, right? So, I think coming up with this LLM would really probably assist in that as well. So, that’s my main contribution and very interesting work that everyone is actually doing here in AI as well.

Moderator: Thank you very much, everyone, for the contribution. This is really an amazing discussion that’s going on here, and we have less than five minutes. So, allow me to please close the session, and I would like to tell my speakers to please tell me if you are to describe in a single word what would be the critical factor that would help in ensuring a digital inclusion in AI? Like in one word, what would be a factor that you think this is very critical for us to achieve digital inclusion in AI? Yes. Okay. Thank you. education? Chair Wong? Public awareness. My online speakers. Capacity. Capacity. Human-based approach. Human-based approach. Human-based approach. Human-based approach. Human-based approach. Human-based approach. Human-based approach. Thank you, everyone. Thank you very much for attending and joining us this session today, and we’re looking forward to seeing you in other sessions. Thank you. Thank you to the panelists, too.

Agreement Points

Need for inclusive AI development

Speaker 2

Speaker 3

Bendjedid Rachad Sanoussi

Speaker 4

Need for technology transfer and infrastructure from Global North to South

Importance of developing local language content for digital literacy

Need for human-centered approach in AI design

Including young people in shaping AI policies and tools

The speakers agree on the importance of inclusive AI development, emphasizing technology transfer, local language content, human-centered design, and youth involvement.

Addressing infrastructure and access challenges

Speaker 2

Bendjedid Rachad Sanoussi

Need for technology transfer and infrastructure from Global North to South

Lack of affordable and reliable internet infrastructure in many areas

Expanding infrastructure and promoting energy access

The speakers agree on the need to address infrastructure and access challenges to bridge the digital divide and promote AI inclusion.

Similar Viewpoints

Both speakers emphasize the importance of creating content and services in local languages to make digital technologies more accessible and relevant to local communities.

Speaker 3

Bendjedid Rachad Sanoussi

Importance of developing local language content for digital literacy

Developing localized content and services

Both speakers advocate for inclusive policy-making processes in AI, emphasizing the need for universal guidelines and youth involvement in shaping AI policies.

Speaker 2

Speaker 4

Need for universal guidelines on acceptable AI use

Including young people in shaping AI policies and tools

Unexpected Consensus

Potential of AI to preserve minority languages

Audience

Speaker 3

Potential for AI to help preserve and develop minority languages

Importance of developing local language content for digital literacy

There was an unexpected consensus between an audience member and Speaker 3 on the potential of AI to help preserve and develop minority languages, despite the challenges posed by data scarcity for these languages.

Overall Assessment

Summary

The main areas of agreement include the need for inclusive AI development, addressing infrastructure and access challenges, developing localized content, and involving diverse stakeholders in AI policy-making.

Consensus level

There is a moderate level of consensus among the speakers on the key challenges and potential solutions for digital inclusion in AI. This consensus suggests a shared understanding of the complexities involved in bridging the digital divide and the need for multi-faceted approaches to address these issues. However, there are still variations in the specific solutions proposed, indicating the need for further dialogue and collaboration to develop comprehensive strategies for AI inclusion.

Different Viewpoints

Approach to addressing digital exclusion

Speaker 2

Speaker 3

Bendjedid Rachad Sanoussi

Need for technology transfer and infrastructure from Global North to South

Importance of developing local language content for digital literacy

Lack of affordable and reliable internet infrastructure in many areas

Speakers emphasized different primary factors for addressing digital exclusion: technology transfer, local language content development, and infrastructure improvement.

Unexpected Differences

Overall Assessment

summary

The main areas of disagreement centered around prioritizing different approaches to address digital exclusion and AI governance.

difference_level

The level of disagreement among speakers was relatively low. Most speakers presented complementary rather than conflicting viewpoints, focusing on different aspects of the same overarching issues. This suggests a multifaceted approach may be necessary to address digital exclusion and ensure inclusive AI development.

Partial Agreements

All speakers agreed on the need for inclusive AI governance, but emphasized different aspects: universal guidelines, multi-stakeholder collaboration, human-centered design, and youth involvement.

Speaker 2

Speaker 3

Bendjedid Rachad Sanoussi

Speaker 4

Need for universal guidelines on acceptable AI use

Importance of multi-stakeholder collaboration on AI governance

Need for human-centered approach in AI design

Including young people in shaping AI policies and tools

Similar Viewpoints

Both speakers emphasize the importance of creating content and services in local languages to make digital technologies more accessible and relevant to local communities.

Speaker 3

Bendjedid Rachad Sanoussi

Importance of developing local language content for digital literacy

Developing localized content and services

Both speakers advocate for inclusive policy-making processes in AI, emphasizing the need for universal guidelines and youth involvement in shaping AI policies.

Speaker 2

Speaker 4

Need for universal guidelines on acceptable AI use

Including young people in shaping AI policies and tools

Key Takeaways

There is a significant digital divide and AI divide between the Global North and South that needs to be addressed

Lack of infrastructure, affordable internet access, and digital literacy are major barriers to AI inclusion in developing countries

Developing AI tools and content in local languages is crucial for digital inclusion

A human-centered, inclusive approach is needed in AI development to ensure it meets the needs of diverse populations

Youth and local communities should be involved in shaping AI policies and tools

There is a lack of AI regulations and policies in many countries, especially in the Global South

Resolutions and Action Items

Promote technology transfer and infrastructure development from Global North to South

Develop more localized AI content and tools in indigenous languages

Implement digital literacy programs, especially in rural areas

Include youth and marginalized groups in AI policy development

Expand internet infrastructure and promote affordable access

Create universal guidelines for acceptable AI use

Unresolved Issues

How to effectively regulate AI across different countries and contexts

How to preserve and develop AI for minority languages with small speaker populations

How to ensure older populations are not left behind in AI adoption

Whether to wait for government policy on AI or let industry take the lead

How to address the growing divide between data-rich and data-poor languages in AI development

Suggested Compromises

Balancing rapid AI development with careful consideration of inclusivity and ethics

Collaborating across developed and developing countries to address shared challenges like minority language preservation

Starting with basic digital literacy before introducing advanced AI concepts in some communities

We need to create in term of digital literacy content using our local languages for, I mean, to allow or to help communities to understand what the message is. For instance, I cannot explain to someone from my community what artificial intelligence means.

speaker

Speaker 3

reason

This highlights the critical challenge of language barriers in AI adoption, especially in diverse linguistic regions. It emphasizes the need for localized content to make AI accessible.

impact

This comment shifted the discussion towards the importance of multilingualism in AI development and sparked further conversation about language diversity challenges in different countries.

To ensure that inclusivity and accessibility really fit in our life today, AI and emerging technologies, we need to design them with a human-centered approach, like we need to put the human at the middle so that this technology can respect human rights.

speaker

Bendjedid Rachad Sanoussi

reason

This comment emphasizes the critical importance of human-centric design in AI development, ensuring technology serves human needs and respects rights.

impact

It refocused the discussion on ethical considerations in AI development and the need to involve local communities in the design process.

I think we are now in an era of data poor and data rich languages. And I think if you look forward in the next five to 10 years, AI systems which train on language sets are going to train disproportionately on data rich languages. They’re going to lead data poor languages to the side.

speaker

Ram Mohan

reason

This comment introduces a crucial perspective on the long-term implications of AI development on language diversity and preservation.

impact

It deepened the conversation by highlighting a potential future challenge in AI and language, prompting further discussion on how to address this issue.

Should we wait for government to actually introduce policy or should we let the industry to lead, right?

speaker

Mbongi Nimsimangasori

reason

This question raises an important point about the balance between government regulation and industry innovation in AI development.

impact

It introduced a new dimension to the discussion about policy development and implementation, highlighting the tension between waiting for government action and allowing industry to lead innovation.

Overall Assessment

These key comments shaped the discussion by highlighting critical challenges in AI adoption and development, particularly in diverse linguistic and cultural contexts. They broadened the conversation from technical aspects to include ethical considerations, policy challenges, and long-term implications for language and cultural preservation. The discussion evolved from identifying problems to exploring potential solutions and considering the roles of different stakeholders in addressing these challenges.

How can we develop AI tools and content in local languages to improve digital literacy and inclusion?

speaker

Speaker 3 (Dr. from Chad)

explanation

This is crucial for ensuring AI technologies are accessible and useful to diverse populations, especially in rural areas and developing countries.

What policies and regulations are needed to govern AI use and development across different countries?

speaker

Moderator and Speaker 2 (Jawan)

explanation

There is a lack of clear policies in many countries, highlighting the need for international collaboration on AI governance.

How can we ensure multi-stakeholder collaboration, including local communities and end-users, in AI development and governance?

speaker

Speaker 2 (Jawan) and Bendjedid Rachad Sanoussi

explanation

This is important for creating AI systems that truly meet the needs of diverse populations and respect human rights.

What strategies can be employed to bridge the digital divide between rural and urban areas?

speaker

Moderator

explanation

This is critical for ensuring equitable access to AI and digital technologies across different geographic regions.

How can we address the needs of older populations in accessing and using AI and digital technologies?

speaker

Audience member

explanation

This highlights the importance of considering all age groups in digital inclusion efforts, not just youth.

How can we prevent the disappearance of data-poor languages from digital infrastructure as AI systems advance?

speaker

Audience member (Ram Mohan)

explanation

This is crucial for preserving linguistic diversity and ensuring AI doesn’t exacerbate existing language inequalities.

Should we wait for government policies or allow industry to lead in AI development and implementation?

speaker

Audience member (Mbongi Nimsimangasori)

explanation

This highlights the tension between policy development and technological progress, particularly in countries where government action may be slow.