AI: Lifting All Boats / DAVOS 2025

Summary

This panel discussion at the World Economic Forum in Davos focused on how artificial intelligence (AI) can benefit global economic growth, particularly in developing countries. The panelists, including leaders from the IMF, Rwanda, Microsoft, and other organizations, explored the potential of AI to boost productivity and GDP growth worldwide. However, they also highlighted significant disparities in AI preparedness and adoption between advanced and developing economies.

Key challenges discussed included the need for better digital infrastructure, energy access, and data scientist skills in many countries. The panelists emphasized the importance of regional collaboration, especially in Africa, to create the necessary scale for AI adoption and development. They suggested that countries could benefit from AI even without being producers, by focusing on effective usage and adaptation to local needs.

The discussion touched on the role of governments in stimulating AI demand through public sector services and the potential for AI to assist in areas like climate disaster prediction. The panelists also addressed concerns about data sovereignty and geopolitical tensions affecting global AI development.

A notable point was the higher trust in AI observed in developing markets compared to more advanced economies. The panel concluded on an optimistic note, emphasizing the importance of individual curiosity and willingness to learn about AI. They suggested that while not everyone can produce AI, its widespread adoption and creative application could lead to significant benefits across the global economy.

Keypoints

Major discussion points:

– The potential for AI to boost global economic growth, but with uneven preparedness and access across countries

– The need for regional collaboration and data sharing to help developing countries benefit from AI

– The importance of building digital infrastructure, skills, and demand to enable AI adoption

– The role of governments in driving AI adoption through public services and regulation

– Geopolitical tensions around AI technology and chip exports

Overall purpose:

The discussion aimed to explore how AI can be leveraged to drive inclusive economic growth and development globally, rather than exacerbating existing digital divides between countries.

Tone:

The tone was largely optimistic and solution-oriented, with speakers acknowledging challenges but focusing on opportunities and potential ways forward. There was a sense of urgency about the need for action, particularly around regional cooperation and skills development. The tone became slightly more cautious when discussing geopolitical tensions near the end, but remained generally constructive.

Speakers

– Vijay Vythianathan Vaitheeswaran: Global Energy and Climate Innovation Editor of The Economist, moderator

– Kristalina Georgieva: Managing Director of the IMF

– Paula Ingabire: Minister of Information, Communication Technology and Innovation of Rwanda

– Bill Thomas: Global Chairman and CEO of KPMG

– Brad Smith: Vice Chair and President of Microsoft

– Hatem Dowidar: Group Chief Executive Officer of EAND

Additional speakers:

– Jennifer Schonberger: Reporter for Yahoo Finance

– David Garza: President of Tec de Monterrey (university in Mexico)

Expanded Summary: AI and Global Economic Growth at the World Economic Forum

This panel discussion at the World Economic Forum in Davos explored the potential of artificial intelligence (AI) to drive global economic growth, with a particular focus on its impact in developing countries. The panel, moderated by Vijay Vaithyanathan Vaitheeswaran of The Economist, featured prominent speakers including Kristalina Georgieva from the IMF, Paula Ingabire representing Rwanda, Brad Smith of Microsoft, Bill Thomas of KPMG, and Hatem Dowidar.

Economic Potential and Disparities

Kristalina Georgieva opened with a striking statistic: AI could boost global growth by 0.8%. However, she highlighted a significant challenge in the form of an AI preparedness index, which reveals wide disparities between advanced and developing economies. Georgieva noted that job exposure to AI varies greatly, with 60% of jobs in advanced economies potentially impacted compared to only 26% in low-income countries.

Paula Ingabire provided a concrete example of AI’s potential impact, stating that Rwanda sees the possibility for a 6% GDP contribution from basic AI use cases. She also shared Rwanda’s digital transformation efforts, with 95% of the population covered by 4G and 70% having access to electricity.

Hatem Dowidar categorized countries into three groups based on AI readiness: AI producers, AI adopters, and those still building basic infrastructure. He mentioned the Edison Alliance’s efforts to connect 1 billion people by 2025.

Infrastructure and Skills Challenges

The panelists identified several key barriers to AI adoption in emerging markets. Vaitheeswaran emphasized that energy access and digital connectivity are foundational requirements for AI implementation. Brad Smith highlighted a critical shortage of data scientists in Africa, comparing the continent’s numbers unfavorably to those in Europe.

Smith also announced Microsoft’s $80 billion investment in AI infrastructure and the creation of a fund starting at $30 billion to support AI development globally.

Regional Collaboration and Data Sharing

A recurring theme was the importance of regional collaboration, particularly for smaller countries and developing economies. Ingabire spoke about efforts on the African continent, including the African Union’s continental AI strategy and initiatives like Smart Africa, to develop frameworks for cross-border data flow.

Smith suggested that countries like Rwanda could benefit from AI infrastructure only if they collaborate with neighboring nations to create sufficient demand and scale. He proposed that the East African community could collectively decide to use shared data centers.

Government Role and Public Sector Adoption

The panelists agreed on the crucial role of governments in driving AI adoption. Smith argued for demand-driven strategies, suggesting that governments should stimulate AI demand by adopting it for public services.

Data Sovereignty and Technological Solutions

The discussion touched on the balance between data sharing for AI development and data sovereignty concerns. Dowidar mentioned technologies that allow for ring-fenced data in other countries while preserving data integrity and sovereignty.

Geopolitical Factors and Global AI Development

Smith raised points about geopolitical factors affecting global AI development, noting that US export controls on AI chips are creating incentives for Chinese development of alternative technologies. He emphasized the importance of ensuring US allies have access to AI technology while acknowledging the need for some level of control.

Trust and Adoption in Developing Markets

Bill Thomas highlighted an intriguing point about the higher level of trust in AI observed in developing markets compared to more advanced economies, suggesting a potential advantage in terms of public acceptance and willingness to adopt AI technologies.

Practical Applications and Opportunities

Dowidar discussed AI’s potential in weather forecasting and disaster prevention, highlighting its practical applications in developing countries. This underscored the diverse ways AI could contribute to economic growth and societal well-being.

Strategies for Inclusive AI Development

The panelists proposed several strategies for more inclusive AI development, including focusing on being beneficiaries and adapting AI to local needs, investing in educational infrastructure, creating regional data zones, leveraging AI for clean energy projects, and developing frugal AI models adapted to emerging market conditions.

Conclusion

The discussion concluded optimistically, emphasizing the importance of individual curiosity and willingness to learn about AI. While acknowledging the challenges, particularly for developing economies, the speakers highlighted the significant potential for AI to drive inclusive economic growth globally, provided there is concerted effort in regional collaboration, infrastructure development, skills training, and adaptive policymaking.

The session included audience participation and questions, reflecting the high level of engagement on this topic. Attendees were encouraged to continue the conversation on social media using the WEF25 hashtag.

Vijay Vythianathan Vaitheeswaran: Welcome, ladies and gentlemen, to our session on AI, lifting all boats. I’m Vijay Vythianathan, I’m the Global Energy and Climate Innovation Editor of The Economist, and it is my great honor to be your immoderate moderator today for what I promise you will be a lively and, I hope, provocative and educational session as well in thinking about AI. I know there’s no shortage of AI sessions here at Davos or, indeed, any other conference we all have been to in the last year or two, but I think this topic may be the most important of all of them for us to think about because it deals with a much bigger frame in thinking about AI. The topic is, of course, how can AI not fall into the trap of the digital divide that we’ve seen over the years with the internet and energy access, which 100 years on continues to be a challenge in parts of the world for basic electricity, but rather maybe serve as that ability to lift many boats, indeed, all boats, rather than being as Coleridge once observed, that there might be water everywhere, but not a drop to drink. We don’t want that outcome. I think we can all agree on that. So to help us dig into this meaty topic, and one I think that really is one that sometimes tends to get ignored in some of the geopolitical discussions or in some of the challenges about the actual technology itself and guardrails, I’m delighted to have a wonderful panel. To my immediate left, we have Dr. Kristalina Georgieva, Managing Director of the IMF. To her left is the Honorable Paula Ingabire, Minister of Information, Communication Technology and Innovation of Rwanda. We have Hotem Dawidar, who is the Group Chief Executive Officer of EAND. We have Bill Thomas, Global Chairman and CEO of KPMG. And to my far left, Brad Smith, Vice Chair and President of Microsoft. Please give them a nice round of applause to warm them up, get them going. Great. Maybe, Dr. Djurdjeva, I’ll start with you. The IMF has thought hard about this problem. You’ve got a stream of research that’s already been looking at AI and its effects, its diffusion, some of the challenges that there might be. There was some early hope that AI would be different, that somehow we saw some early productivity gains among, for example, frontline employees being able to catch up or maybe a chance for emerging economies to leapfrog. There was the hope. You have research coming out in a few months, which you can give us a preview. What are your economists finding?

Kristalina Georgieva: The good news is that our research shows that the potential of AI to lift global growth is significant. We are at the point when there is one universal common theme across all continents, with the exception of the United States, that growth is simply insufficient to give people and businesses the perspective they aspire for. So, we have growth at around 3%. It used to be 3.8%. And here is the good news. AI can lift growth up by 0.8%. The second part of our research, however, shows that there is very different degree in preparedness, in exposure, in access. We have created a very interesting AI preparedness index that ranks countries on their digital technology, in other words, physical connectivity, in the markets, how well they are prepared for it, regulation, is there a good definition of how AI can benefit society. And what it shows is that the accordion between advanced economies and emerging market developing countries is wide open. It is also a problem when it comes down to exposure to AI. 60% of jobs in advanced economies are exposed to AI positively, they can see lifting up in productivity, or negatively, they may be gone. That goes down to 40% for emerging markets, it goes down to 26% in low-income countries. And the third one that is the most troubling is when we look into the opportunity to translate this whole wonderful thing called AI into growth, and that is the newest research we will be delivering, there is a very significant, about 50% gap between advanced economies and low-income countries. So you said we want to close this gap. We can talk about ways to close it, but I want to make one point in the very beginning, because we talk about boats and water, and yes, there could be water, but if you don’t have a boat, all this conversation doesn’t matter. And my worry is that in developing countries, we still have a very significant part of the economy and people to whom this conversation about lifting all boats is a bit theoretical.

Vijay Vythianathan Vaitheeswaran: Okay, so you’ve sketched out the challenges. Now, some of these, it must be noted, or at least it seems to me, are not unique to AI. It isn’t because this is one of those revolutionary general-purpose technologies that comes along once in many decades or centuries, like the steam engine or electricity itself, grid electrification. The things you’re talking about are perhaps characteristics of… emerging markets with imperfect infrastructure. Almost a billion people don’t have access to modern energy. So if you’re spending, usually girls and women, walking miles a day to fetch crop residue or cow dung, it’s a bit crazy to say you’re gonna become a big winner from the internet economy, or at least in making AI, manufacturing the AI process. But could you be a beneficiary if we’re able to develop products and services that give advance warning of whether a disaster, for example, on a more rapid basis that’s delivered to a smartphone in the village so that everyone has a warning? Are there ways we can find around this? Because if you say you must electrify everything before AI can help you, we’ve been waiting 100 years, and I know there are initiatives there, but I would hope that’s not the only pathway.

Kristalina Georgieva: You’re so right, and that is the attractiveness, that there is so much that consuming AI can deliver to people just on the basis of having a cell phone. You can have diagnostics of your health problem. You can have better access to educational materials. We talked about early warning systems, when to get your crop to market, all these things, you can have a phone. What I’m talking about is translating AI into an integral part of economic growth. And there, it is not enough to just have a phone because producing data, running the systems, that is gonna be a much more complicated than energy-consuming proposition.

Vijay Vythianathan Vaitheeswaran: That’s a fair point, of course. And so, going beyond being beneficiaries of these technologies, how to be a part of the AI economy, a part of the creator economy, and benefit and increase national productivity. I think, Minister Ingabire, you’ve thought a lot about this. You’ve already taken some actions in your country, which has got a good record on digital. Can you tell us? your perspective on how you can continue that success in thinking about extending it to AI? What’s different about AI, or is it really a continuation of some things that you’ve been able to do successfully thus far?

Paula Ingabire: I would like to pick up from what Kristalina was talking about, growth. We have done a study to look at what would be the impact of AI in Rwanda, and what we are seeing is, and I think you talked about 0.8%, what we are seeing is even with basic use cases that we can deploy, we are seeing a potential to create at least 6% GDP contribution. And so we’ve been able to map out the different use cases. We talk about early warning systems for farmers. Farmers make up the largest percentage of our population, and you see that trend similar across the continent. We’re talking about public administration use cases, whether it’s from taxes or the social security claims, and really being more efficient with how we’re using some of these tools. And to your point, Vijay, this is going, definitely it’s going to be a continuation over the already existing foundations that we’ve put in place in our digital transformation efforts. It also means that we’re going to have to think differently on how we continue to close the gap on some of the digital aspects, so it’s making sure that everyone is connected. Today we have about 93% population coverage, but we still have a gap on usage, and the usage gap is driven largely by digital skills, by access to affordable devices, and affordable internet connectivity. But at the same time, we’re also going to be combining those efforts with energy access. Rwanda today has 81% household access to electricity, but when we start to look at what these AI use cases that we’re going to put in place, we start to anticipate what the demand is going to look like and that has called for different ways on how we leverage that. For Rwanda, we’re building on these foundations that are already in place, but at the same time, asking ourselves the question, how do we not lag far too behind on this AI journey.

Vijay Vythianathan Vaitheeswaran: And there has been some analysis that looks at the effects of individual countries that may have one shortfall or another. It might be skills, it might be infrastructure, it might be data collection. And the WEF actually put out an interesting report on this, on AI competitiveness through regional collaboration, which I recommend to all of you. What do you see as the prospects or potential for regional collaboration? I know there are a lot of sensitivities at the national level. This is the age of national sovereignty. We see this among the big boys, as it were, and is there an opportunity for emerging economies maybe to collaborate in order to step up that ladder?

Paula Ingabire: Absolutely, and we’re already seeing that happen. Last year we worked with Singapore to put in place a playbook for small states on how we can really, you know, create a proper roadmap on how we build the foundations for the AI economy. We are seeing this even within the Commonwealth member states, you know, creating that kind of coalition which is very important. Now I want to give you an example. When you talk about regional collaboration, at the end of the day, collaboration is very important for developing countries because it helps you to address the resource constraints that you already have. So whether it’s from an energy perspective, whether it’s from any other kind of resource compute, and even data. So we are seeing, at least on the African continent through Smart Africa and different institutions, where we are now starting to work on frameworks that allow for cross-border data flow. And that’s going to be very important, open data access in terms of making sure you have lots of data that you can train these models on. And I’ll end on this one, where regional collaboration comes into play in a very powerful way is when we look at what our energy demand is going to be in terms of AI workloads, we cannot close, you know, we cannot satisfy the demand in the short term. So we’re already working with the East African community on an East African, you know, energy pool. So if there’s excess capacity in any of the countries and one country has a sudden demand for, you know, for energy because of the AI workloads, then we’re suddenly able to do that.

Vijay Vythianathan Vaitheeswaran: see these advances because, of course, historically Africa has found it challenging to have trade across borders, right, or even have visa-free travel. This has been a real challenge. So to see some of these bits of progress, and hopefully on the digital front we can start off on a better footing. I do believe we have a summit coming up in April to try to take on some of these issues, so people who are interested could follow up on that. I want to come to Hatem. Hatem, you can give us a private sector view. Very interestingly, of course, your company has its base in the Middle East, but you’ve got extensive operations in Africa, and I just discovered today in Eastern Europe as well. So you’ve got the cross-country perspective and the bottom-up perspective. What do you see as the principal challenges? We’ve seen the macro sort of challenges laid out by Dr. Georgieva What do you see from the outside view, as it were?

Hatem Dowidar: Okay, so I think there are, you know, across the markets we work, there are several buckets of countries. So some of our markets, like our home markets in the UAE and in Saudi, we have the capital, we have the connectivity, super well-connected networks, as well as having the energy at an affordable price. So some of these markets will be ideal to build data centres that you’re able to not only use, but also teach the models and improve the models and so on. And you also have very strong national policy support in those markets. Absolutely. And we have cooperation with many companies across the world. In fact, Microsoft has a big partnership in the UAE that’s exactly working on this. And then you have other countries that maybe have the energy, but don’t have the access to the know-how and the connectivity, and then some that are actually lacking on both. And there is definitely still a possibility to do this, because there is the less energy that is needed to actually access AI, and a lot of energy that is needed. needed to teach the AI. So there is a possibility where you can have these central areas where we can serve the countries that don’t have the massive energy needed to teach the models, but then we need to relax the AI data sovereignty issues that we discuss. And there are technologies available now that allows you to have ring-fenced data in other countries where you can still preserve the integrity and the sovereignty of this data. And I think this mindset is very important for regulators to be able to actually close the divide rather than widen it. And we have been working in the World Economic Forum as part of the Edison Alliance in connecting people. We actually celebrated that we connected a billion new users, which was something that a lot of the members of the Alliance worked for. Now we still have 2.6 billion that are unconnected, and we need to really work on connecting them. Because if they are not connected, they cannot use AI, they cannot use digital overall. So we need to connect them.

Vijay Vythianathan Vaitheeswaran: Are there, you know, this point you make about having the tools now to be able to ring-fence data in a way that might satisfy national regulators. That sounds very important to me, because what I know about regulators is they don’t want to give up their national sovereignty, whether it’s because of privacy rules or now increasingly involving AI and data sets and other – in an era of geopolitical competition. Have you been able to persuade any particular lead markets that your tools or the tools that you’re advocating can work? Who’s on board with your plan?

Hatem Dowidar: Look, I mean, there are markets that work closely, let’s say, have strong geopolitical ties. And I just don’t want to say names, but we do have a couple of cases now where agreements have been done that allows for data to be handled securely in other markets. There are a few, let’s say, lighthouse cases that allows this to happen. And actually, some of the hyperscalers – Microsoft is one of them. AWS is another that are working on creating these ring-fenced, sovereign clouds that can serve countries from another country while really preserving that integrity and sovereignty there. So there are cases of that already.

Vijay Vythianathan Vaitheeswaran: Some lighthouses in the WEF terminology. Brad, I saw you nodding vigorously and you’ve been called out by company names, so I’m going to give you the next chance to weigh in here. Obviously, you’re a global company, but when you think about moving into markets outside of the United States, what are the considerations that help you make this decision? We heard the UAE example. Obviously, the government is very welcoming. But in emerging markets where they don’t have all of the factors that have been ticked off so far, how do you weigh that decision? What can you do to help them? And when do you encourage sort of regional pooling of talents?

Brad Smith: Well, we are spending, as we’ve said, $80 billion worldwide this year to build out AI infrastructure. And the reality is there’s not a shortage of capital to build out infrastructure. That’s great news. And it’s not just ours. You see Microsoft, MGX, which is also in the UAE, BlackRock. We’ve created, separate from that, $80 billion, a fund that is going to start probably at around $30 billion, and we hope to grow to $100 billion. And there will be many other things. So we will go where there is demand. It’s really that straightforward. It’s a very expensive, capital-intensive investment. If you can recoup the investment, you spend the money. Now, when you put this in the context of, call it international economic development, because that’s what it is, I think there’s always two fundamental approaches to spur development. It’s either increased supply or increased demand. A lot of what development banks deal with, I think, is when there’s a shortage of capital. That’s what they exist for. What we need here is a demand-driven strategy. So how do we do that? Well, Rwanda is a great great example. Let’s just look at East Africa. Earlier this year, Microsoft and G42 announced together that thanks to G42’s investment, we together would create a $1 billion data center in Kenya. That is a market with 50 million people. It is hard to spend a billion dollars to support 50 million people in Kenya alone. But we’re doing it. But the real question is, can we grow that? And can we reach Rwanda? We can, but only under one circumstance, that you get Rwanda and Tanzania and Uganda and Kenya and Ethiopia. You get the East African community to decide together that they will all use that data center. And they will put in place among them an agreement to, in effect, think of it as a data zone, just like we have free trade zones. That will get us halfway there. There’s a second thing that I think is needed as well. In these countries, the only way to really kickstart demand is to have the government do it for itself, for its public sector services. These governments have data centers. They have certain typically on-prem services. Move to the cloud. Consolidate demand. Stimulate demand. And then you will have not only one data center, but you’ll probably have an Amazon or a Google or an Alibaba or a Huawei.

Vijay Vythianathan Vaitheeswaran: Right. You’re not advocating a doge for East Africa, are you?

Brad Smith: No. But what is needed is demand. And the thing that will absolutely hold back, let’s just say Africa, is if every country says, I won’t use a data center unless it’s in my own border. Because then, now you’re back on the electricity timeline, which has been going on, now we’re in decade number 15.

Vijay Vythianathan Vaitheeswaran: Right.

Brad Smith: For trying to bring electricity around the world.

Vijay Vythianathan Vaitheeswaran: You can make that argument in ASEAN as well, for example, and other regions where there’s a great divergence between some leaders and a lot of laggards that could benefit from cooperation. A fair point. Bill, let’s come to you.

Bill Thomas: Sure.

Vijay Vythianathan Vaitheeswaran: You’ve heard the arguments aired. I know your team has worked with the WEF to put out some thought, leadership on this. Give us the sharpest point from your analysis and which direction are you, is the boat sinking or is the boat being lifted?

Bill Thomas: Well, I will say-

Vijay Vythianathan Vaitheeswaran: Or do we even not have a boat, as has been posited?

Bill Thomas: Well, the nice thing about going last is you get to tie a bunch of these things together. And I will say this, so we were really privileged to work with the WEF, but it’s really working with the 200 members of the AI Governance Alliance that commissioned the piece that said, what does an intelligent economy look like? And essentially, to Kristalina’s point, what’s the boat we’re even talking about? So Brad mentions infrastructure, and I won’t do it justice in the time I have, but if you look at that blueprint that we put out, you get to a foundational layer that I would just call the tech, of which one element is the infrastructure. It goes to what’s the financing for it, what’s the data model, what’s the compute model. There’s a whole layer there of which one thing is infrastructure. You go to the next layer and it says, what’s the aspiration of the country, or in certain cases, the region? And it says, how are you going to really create the value from AI itself? So making sure you understand what sectors of your economy you should go first to get the best benefit from, and really create that value proposition that says, this is why we want to go there. And then the last one, this is really a trust and growth story, because that’s the growth layer. The last one is trust. All about trust. Our people and the governance that goes into making sure it’s safe and the kinds of things we talk about. I will tell you, that’s the boat. The real question that we’re talking about today, so I would encourage everybody to read the report. That gives you the boat. The real question is, one, not every country, by no means every country, can build their own boat. And already we’ve talked about the fact that that is impossible. Absolutely impossible. Nor should that be the objective. So the real question is, so what? Everybody has to have access to a boat, but not everybody is going to be able to build a boat. So let’s start talking about how do we make sure everybody has access to a boat.

Vijay Vythianathan Vaitheeswaran: Well framed. And if I may, just in the interest of provocation, since we’ve had our initial comments, and before I come to the crowd for your thoughtful questions, maybe suggest that the greatest beneficiaries from AI may not necessarily be those that create it. My last book on the history and future of innovation, called Need, Speed, and Greed, available at critical scores for you, made the point from research that the greatest beneficiaries of inventions throughout history, inventions that became value-creating innovations, are often not the inventor or even the company or the country that came up with something new, but where it was adopted and how it was diffused. And not only that, adapted, modified, improved, attracting local attention and meeting local needs. And in the case of Africa, we can take an African example, everyone benefited from mobile phones, but it was in East Africa that M-Pesa developed mobile banking, pioneering the world in this area, right? And in some ways still leading many countries. And so we can find leapfrogging innovations as well in the adoption and use of innovation. And so if I were to frame it in that way, and equally, you know, most countries, or very few countries invented email, but we all can benefit from it. The list can go on and on. this, if we look at the consumption and usage of AI for development, as it were, or any other purpose, national job creation, there might still be things lacking. If you don’t have access to electricity, it’s very hard to talk about access to AI. So if I were to frame the question as being beneficiaries of that and harnessing it to leapfrog ahead, what are the things that still need to be done? What’s still missing? And that’s a different question than, Brad, you immediately want to jump in there.

Brad Smith: In my opinion, just my opinion, but first of all, that your book is great. And there’s another book that was published last year that looks historically and basically comes to the same conclusion. It’s called Technology and the Rise of Great Powers by a George Washington professor named Jeff Ding. And it’s the same conclusion. What really drives economic growth for countries is the adoption and diffusion of technology. The second conclusion in his book, when you look at the three industrial revolutions, is that the most important infrastructure you actually then need is the educational infrastructure so that people have the skills to put the technology to use. Now, you take that and you apply it to Africa. To me, the biggest shortage in Africa is actually data scientists. For every 14 data scientists you meet in Europe on a per capita basis in Africa, you’ll meet only one. Yeah, because if you can bring the capital in, stimulate the demand, then you will generate the data and put it to use, including in local languages, but only if you have data scientists. And the good news is it costs a lot less money to train data scientists than it does to build data centers. But I would prioritize that.

Vijay Vythianathan Vaitheeswaran: Yeah, Dr. Georgieva.

Kristalina Georgieva: I very much like taking this direction to talk about what would make. Africa competitive in this area. And the most important message that came from you, Brad, and actually I’m sure you would agree, is that individual countries, even the big countries in Africa, are gonna lose if they don’t figure out a way to work with others. And I want to make a broader point. We are now in a more fragmented world. We no more have the aspiration of developing countries to succeed because they’re part of integrated global economy and they can rely on exporting goods and services to others for their development. Countries need to think, in this environment, what can they do? And to my mind, working more in regional and sub-regional settings is number one priority for countries to figure out ways to benefit from creating critical mass, but also creating regulation that they adopt, all of them, they operate by the same rules, thinking of skills and capabilities, including the data scientists you’re talking about as common public good. And all of us, and I put this also on us at the IMF, we all have to think, how can we support it? I’ll give you one example. At the IMF, we do, occasionally, regional-based economic assessment, and that is mostly done, most frequently, regularly done for the Eurozone. So we have annual consultations with the countries of the Eurozone because they share common currency. And I’m thinking of expanding that approach on economic opportunities to groupings of countries where there is that desire. to succeed on the basis of creating that common objective and common resources and common public good. You mentioned ASEAN, absolutely. You guys in your neighborhood, you have to do it. You have, otherwise we can talk here until the cows go home and progress won’t be made. And I have a question for the audience. How many in the audience are AI producers? You participate in generating the new economy of AI. Hands up.

Vijay Vythianathan Vaitheeswaran: Hands up. How many are at companies or?

Kristalina Georgieva: Race, race, companies in debt.

Vijay Vythianathan Vaitheeswaran: Okay. Right. Less than 10%.

Kristalina Georgieva: How many of you are AI consumers?

Vijay Vythianathan Vaitheeswaran: Hands up, how many actively try different tools? There we go.

Kristalina Georgieva: I sort of rest my case.

Brad Smith: But there’s another question because I think those are great questions. How many of you have a phone, a smartphone? I assume everybody, right? Okay, if you have a smartphone, you are generating data.

Kristalina Georgieva: Yes.

Brad Smith: That is the fuel for AI. Because we talked a little bit about, is there a data shortage, which there is. But the good news is everybody uses a phone as generating data. Every satellite that passes over Africa every day is generating data. It is a lot easier to generate data today, whether you want to or not, frankly, than it is to find data scientists who can put that data to work.

Hatem Dowidar: I’ll actually add to that, maybe something a bit positive, given that I feel we’re getting a bit gloomy on the capabilities and so on.

Kristalina Georgieva: I don’t think we’re getting gloomy. I think it’s actually very exciting because we are talking about the universe of the possible. We build the boat, we decided that we are gonna bring countries together to get on that boat. So things are going very well.

Vijay Vythianathan Vaitheeswaran: We want to take the boat from the lake to the ocean.

Hatem Dowidar: You sailed the boat safely as well, so actually there is a lot of things that AI is doing today for Africa and for the global south, and frankly seeing some of the things that happened earlier this year, sorry, late last year in Spain, for example, with climate disasters, there are room for AI to do more, and we’re actually working with UNDP on integrating weather forecasts together with AI models, together with data extracted from social media, so because when disasters happen, people start tweeting and whatever about weather in Africa, weather anywhere, people send messages, start doing it, so to combine all this data together to create more accurate forecasts of things that are happening and to be able to warn people earlier on. So we have already started working on this with UNDP since last September, hopefully this year we can have something that is more working, and this is something that AI is delivering today, the thought at that time was the global south and the emerging markets, but I think with what we’ve seen also with the weather patterns, whether the forest fires in California or the floods in Spain, I think this is a global phenomenon that we see these freak weather incidents, and by having early warning, using AI, we’re able to help a lot of people.

Vijay Vythianathan Vaitheeswaran: This can be one of the great public goods. I’m going to give my audience notice, I’m going to come to you in a couple of minutes, so please gather your best questions, we’ll do what time will allow, I’m going to come to the Minister for a follow-on, and then I’ll come down the line to Bill.

Paula Ingabire: Maybe Vijay, I just wanted to weigh in, because we keep referencing Africa on this panel, a couple of things. When you look at some of the efforts that are being put in place, it’s really how do we come together, not just to aggregate demand, but also to have a single unified strategy. Recently, we had the African Union put in place an AUAI continental strategy, and that sort of like lays out the building blocks of what we want to invest in. And I agree with what Brad was saying, one of the main areas is around skills. How do we, and earlier we were having a conversation, Africa has a growing, youthful, digitally savvy population, and to the point that you’re making about being users or early adopters, Rwanda has seen that. We never created drones, but the way we use drones to deliver, you know, medical products to rural areas, and then that has been able to expand to different markets where, and even in doing that, we didn’t only just use that use case, we were able to create regulations that now many countries are benchmarking, and to create an ecosystem. And so the last point I want to make is what we’re seeing, and you did mention the summit that is happening in April 3rd and 4th, again, the summit, it’s the global AI summit on Africa. The idea is to have a collective voice, to think together about what the priorities are in terms of adoption, what building blocks that we need to put in place to co-invest as countries, but also at the same time, creating those AI hotspots on the continent where we can test and experiment with these applications, but also allow for scale. And for that to happen, we’ll need to have streamlined regulations and policies, and I must say, when I look at really what the future holds, there’s really a lot of promise in how we are coming together as a continent to really drive AI adoption in an equitable manner.

Vijay Vythianathan Vaitheeswaran: Well, this is very promising, I must say. Something quick, though, because I want to get to the audience. You don’t want an angry mob turning on you.

Bill Thomas: I definitely don’t. I’ll have them turn on Brad, though. So we did a piece of research that comes to a point that was made earlier, and it was all about trust in AI, trust in AI. And the interesting thing is, when you think about a cell phone and the human spirit, trust in AI is so much higher in the markets that we’re talking about, those that have the potential to be left behind. They are in the 80s and the 90s. People who feel that, given… given the opportunity, this will make it better for them. Way up there. Skepticism is highest in the markets that have the highest access. So there is a real opportunity here to create momentum, the combination of a phone and the human spirit.

Vijay Vythianathan Vaitheeswaran: That is a wonderful note for me to wind up this part of the session. I’m going to come to questions. We have microphone runners. One thing I want to say, this is a live stream session. I know some of you are on social media. The hashtag to use is WEF25. So please do so. Let’s get a microphone here to the front row, to the lady. And then I’ll move around the room to another part of the room. Short question, short answer, please.

Audience: I’d love to ask Brad a question. So we all know that AI requires a lot of energy. And we’d love that energy to be reliable and clean. And would you tell us a bit more about what Microsoft has been doing to leverage AI to accelerate the permitting and licensing process to deploy more clean energy?

Brad Smith: Thank you. Yeah, I would say first, I think those of us who build data centers have to take on an obligation to invest, to add to the grid the electricity for them. And we need to take on the responsibility to bring on carbon-free energy to do so, or at least offset what we’re going to be taking from the grid that may not be carbon-free. And the interesting thing, if there’s one thing I’ve encountered worldwide, it takes a long time to get a permit from a government agency. Perhaps maybe everywhere outside of China, they may do it better. There are so many places today where it takes more time to get a permit to, say, construct a wind turbine than it does to build and construct the wind turbine. And so we massively need more work to accelerate this. And part of it is non-AI steps that governments can take, put more resources in place, streamline processes. But the fascinating thing about permitting, and we’ve got, I think, most recently 872 permitting processes going on around the world, is that there’s not really a doubt as to whether a permit’s going to issue at the end of it. You might have to make some modifications. But fundamentally, what you’re doing is taking a spec and comparing data to it. And then a human being has to do all that work and identify where there’s a mismatch, and then that focuses the conversation. But when you just think about that, AI is really quite good at that.

Vijay Vythianathan Vaitheeswaran: So it’s ideal for that task, isn’t it?

Brad Smith: Exactly. And so there are now more and more, but it’s just starting. Municipalities, states, and countries starting to use AI to accelerate progress.

Vijay Vythianathan Vaitheeswaran: I’m hearing more and more reports on how AI is being applied to the regulatory process and nuclear power.

Brad Smith: Exactly.

Vijay Vythianathan Vaitheeswaran: Last time you and I spoke, Brad, your company is helping revive the Three Mile Island facility in the US. But this is something that’s being pursued by many companies to clean energy. But the permitting process itself can be accelerated through applications of AI.

Brad Smith: And nonprofits are playing a key role as well.

Vijay Vythianathan Vaitheeswaran: Let’s see another question. I see a hand in the back. And then we’ll swing over to the other side as time permits. A short question, and please identify yourself.

Audience: Thank you so much. Jennifer Schonberger with Yahoo Finance. There’s a technology race now when it comes to AI. The US is putting in place barriers, export controls, particularly when it comes to exporting to China to protect AI advancement in the US. So how does this play into your thesis for AI boosting global growth?

Vijay Vythianathan Vaitheeswaran: Thank you. Thank you. Who wants to take that? Hatem, you had a point about chip barriers as a geopolitical obstacle when we were chatting earlier. Do you want to expand on your views?

Hatem Dowidar: Look, I think that definitely controlling technology doesn’t help in bridging the divide. So the access to these technologies across the world would help more countries in the South build and be able to operate also in this. However, the world is what it is today.

Vijay Vythianathan Vaitheeswaran: And it’s even more so now.

Hatem Dowidar: Yeah. The world is what it is today. And I think, look, I think that’s a good point. that while, of course, we know that especially in AI, U.S. technology is far superior, but there are other parts of the world, there are other technologies coming from Korea, there are other technologies coming from China, that maybe today is not as developed, as sophisticated, but it’s there as well. So there will be possibilities for everyone to find something to work with. And of course, we can solve that by the same idea that Brad talked about, which is having the clusters in places that would be acceptable to the countries surrounding it and acceptable to the U.S. where you can have AI applications running on these big data centers and serving other places as well. So there are ways to go around it. It’s not optimum, but it is possible.

Vijay Vythianathan Vaitheeswaran: Picking up on Hadam’s point, Brad, I want to come to you, put you on the spot here. Are we seeing, in effect, because of geopolitical rivalry, two spheres of influence? Ones that will have access to U.S. chips, U.S. software, U.S. technology, which may well be in the lead, although we’re running, our analysis suggests that the Chinese are catching up in various ways, in part because of the restrictions, driving them towards frugal models, maybe less energy intensive, more creative ways of solving problems that might be more adapted to local conditions and emerging markets. So could we see sort of a geopolitical opportunity for those that are excluded from some kind of American friend zone to go a different way with it? And is that a threat to you or a concern for you?

Brad Smith: I think you raise a really important question. I’d say a couple of things. Almost inevitably and indisputably, when the United States imposes export controls and says that China cannot get U.S. chips, it creates an incentive for China to develop its own, which it had already, but this adds to that incentive. And I think that there’s a lot of similarity, interestingly enough, between, say, the U.S. AI ecosystem and the Chinese AI ecosystem in terms of investment, innovation, growth. They’re both doing quite well. I think the real question is not so much whether the United States government will allow its chips to go to China, because I don’t really sense that view changing, but it’s to the rest of the world. the world. And yeah, that’s where, you know, frankly, in the final week of the Biden administration, they put out this new rule. And I just think it is critically important to the world that all of America’s friends and allies, which is virtually the entire world, have the ability to count on this supply. Now, the US has put in place, I’ll call it qualitative standards, so that there’s a high security, you know, there are some restrictions on how these chips and AI models are used, and that they don’t provide inadvertent leakage to, say, China. The US also put in place quantitative caps for certain countries. And I think there’s room for a healthy discussion about whether you really need, call it a belt and a suspender, a qualitative standard and a quantitative cap, or whether, you know, I don’t know too many people who run around Davos with both a belt and a suspender. We all seem to do pretty well with just one or the other. And I think where we are technologically, we might consider that as an option.

Vijay Vythianathan Vaitheeswaran: So, we have these more… In the application of this perhaps applied flexibly, the way, for example, US sanctions policies have been adapted over time. They don’t always work effectively. They may have a signaling effect more than a practical effect. So, we’ll have to see what this new era brings. I have time for one quick question. I’m going to go all the way to the back since I neglected that side, but it has to be extremely short question, 30-second answer.

Audience: David Garza, president from Tecla, Monterrey, a university in Mexico. My question has to do with education. You mentioned education skills. And the specific question is, who is the actor that must be most involved in this aspect of the re-skilling, re-skilling? Is it the private sector? Is it the educational system? Is it others? And if AI can actually… Are you seeing initiatives of AI actually to be involved into accelerating this? same process.

Kristalina Georgieva: So, you know, the answer, if I were to offer it, would be all of the above. Yes. How is this for short?

Vijay Vythianathan Vaitheeswaran: Short and sweet. Any follow-on? Another sentence or two on who needs to do this? Bottom-up skilling versus big, I mean, the Tec de Monterrey is the finest technical institution in perhaps all of Latin America, so is it the sort of the techs and the MITs of the world or is it bottom-up skilling?

Bill Thomas: It is 100% all of the above, and the biggest risk is the individual who puts their head in the sand and pretends that the world isn’t going to change, because as long as they’re curious and as long as they have a willingness to learn, then it’s everybody’s responsibility to help them get there.

Vijay Vythianathan Vaitheeswaran: You landed on the right place. We’re almost out of time, but I think rather than trying to sum up every thread, this note of optimism but also of individual empowerment that you pointed out, that everyone’s job is to learn and become knowledgeable about AI, to use it, but as that poll of this audience showed, even if only some of your companies are building AI, almost all of you are using it. The opinion poll we heard about that said the attitudes towards AI are extremely positive in countries where actually people crave and desire access to it. I think this points me towards a more positive view that suggests, in fact, people not only are able to see a rising tide, but maybe they’re fixing the boats, maybe they’re bailing out any water that might be there, and as I like to point out, we’re not just caught in a small pool, but this is the great ocean, and we’re going to ride this boat out for the greater success of humanity. We are out of time. Thank you, everyone. Thank you to those who joined on the live stream. Please give a nice round of applause to our wonderful speakers. Enjoy the rest of your World Economic Forum Davos.

Agreement Points

AI has significant potential to boost global economic growth

speakers

– Kristalina Georgieva
– Paula Ingabire

arguments

AI can significantly boost global growth by 0.8%

Rwanda sees potential for 6% GDP contribution from basic AI use cases

summary

Both speakers highlight the potential of AI to significantly contribute to economic growth, with Georgieva citing a global figure and Ingabire providing a specific example for Rwanda.

Regional collaboration is crucial for AI development in emerging markets

speakers

– Kristalina Georgieva
– Paula Ingabire
– Brad Smith

arguments

Countries need to work together regionally to create critical mass

Regional collaboration is key for developing countries to address resource constraints

Create regional data zones and shared infrastructure

summary

Multiple speakers emphasize the importance of regional cooperation to overcome resource limitations and create economies of scale for AI infrastructure and development.

Addressing skills gaps, particularly in data science, is critical for AI adoption

speakers

– Brad Smith
– Bill Thomas

arguments

Data scientists are a critical shortage in Africa

It is 100% all of the above, and the biggest risk is the individual who puts their head in the sand and pretends that the world isn’t going to change, because as long as they’re curious and as long as they have a willingness to learn, then it’s everybody’s responsibility to help them get there.

summary

Speakers agree on the importance of developing skills, particularly in data science, to support AI adoption and development in emerging markets.

Similar Viewpoints

Both speakers suggest solutions to address data sovereignty issues while enabling regional collaboration and shared infrastructure for AI development.

speakers

– Brad Smith
– Hatem Dowidar

arguments

Create regional data zones and shared infrastructure

Ring-fenced data solutions can help address data sovereignty concerns

Both speakers highlight the importance of basic infrastructure and preparedness for AI adoption, particularly in developing economies.

speakers

– Kristalina Georgieva
– Vijay Vythianathan Vaitheeswaran

arguments

There is a wide gap in AI preparedness between advanced and developing economies

Energy access and digital connectivity are foundational requirements

Unexpected Consensus

Trust in AI is higher in developing markets

speakers

– Bill Thomas
– Vijay Vythianathan Vaitheeswaran

arguments

Trust in AI is higher in developing markets, creating opportunity

Focus on being beneficiaries and adapting AI to local needs, not just creators

explanation

Despite the challenges faced by developing markets, there is an unexpected consensus that these markets show higher trust in AI and potential for innovative adoption, which could lead to unique opportunities for AI impact.

Overall Assessment

Summary

The main areas of agreement include the potential of AI to boost economic growth, the importance of regional collaboration for AI development in emerging markets, and the need to address skills gaps, particularly in data science. There is also consensus on the importance of basic infrastructure and the potential for innovative AI adoption in developing markets.

Consensus level

There is a high level of consensus among the speakers on the potential benefits of AI for global economic growth and development. However, there are nuanced differences in approaches to addressing challenges, particularly regarding infrastructure development and geopolitical factors. This consensus suggests a shared understanding of the opportunities and challenges presented by AI, which could facilitate coordinated efforts to promote inclusive AI development globally.

Different Viewpoints

Approach to AI infrastructure development

speakers

– Brad Smith
– Kristalina Georgieva

arguments

Demand-driven strategies are needed to spur AI infrastructure development

Countries need to work together regionally to create critical mass

summary

While Brad Smith emphasizes demand-driven strategies for AI infrastructure development, Kristalina Georgieva focuses on regional cooperation to create critical mass. Smith suggests governments stimulate demand, while Georgieva emphasizes collaborative efforts among countries.

Unexpected Differences

Overall Assessment

summary

The main areas of disagreement revolve around the specific approaches to AI development and adoption in emerging markets. While there is general agreement on the importance of regional collaboration and addressing infrastructure challenges, speakers differ in their emphasis on demand-driven strategies versus resource pooling.

difference_level

The level of disagreement among the speakers is relatively low. Most participants share similar views on the potential of AI to boost economic growth and the need for collaborative efforts. The differences mainly lie in the specific strategies and priorities for implementation. This low level of disagreement suggests a generally unified approach to addressing AI development challenges in emerging markets, which could facilitate more effective policy-making and implementation.

Partial Agreements

Both Smith and Ingabire agree on the importance of regional collaboration for AI development. However, Smith focuses more on creating data zones and shared infrastructure, while Ingabire emphasizes addressing resource constraints through collaboration.

speakers

– Brad Smith
– Paula Ingabire

arguments

Create regional data zones and shared infrastructure

Regional collaboration is key for developing countries to address resource constraints

Similar Viewpoints

Both speakers suggest solutions to address data sovereignty issues while enabling regional collaboration and shared infrastructure for AI development.

speakers

– Brad Smith
– Hatem Dowidar

arguments

Create regional data zones and shared infrastructure

Ring-fenced data solutions can help address data sovereignty concerns

Both speakers highlight the importance of basic infrastructure and preparedness for AI adoption, particularly in developing economies.

speakers

– Kristalina Georgieva
– Vijay Vythianathan Vaitheeswaran

arguments

There is a wide gap in AI preparedness between advanced and developing economies

Energy access and digital connectivity are foundational requirements

Key Takeaways

AI has significant potential to boost global economic growth, but there is a wide gap in AI preparedness between advanced and developing economies.

Regional collaboration and demand-driven strategies are crucial for developing countries to overcome resource constraints and spur AI adoption.

Foundational requirements for AI adoption in emerging markets include energy access, digital connectivity, and skilled data scientists.

Countries should focus on being beneficiaries and adapting AI to local needs, not just creators of AI technology.

Geopolitical factors like export controls are shaping the global AI landscape, potentially leading to the emergence of alternative AI ecosystems.

Resolutions and Action Items

Develop regional data zones and shared infrastructure to create critical mass for AI adoption in developing regions

Governments should stimulate AI demand by adopting it for public services

Invest in educational infrastructure and skills development, particularly for data scientists

Leverage AI to accelerate clean energy permitting processes

Create AI hotspots in Africa for testing and experimenting with applications

Unresolved Issues

How to effectively balance national data sovereignty concerns with the need for regional collaboration

Specific strategies for closing the gap in AI preparedness between advanced and developing economies

How to ensure equitable access to AI technologies given geopolitical restrictions and export controls

The optimal approach for re-skilling and educating workforces for AI adoption across different sectors and regions

Suggested Compromises

Use ring-fenced data solutions to address data sovereignty concerns while enabling regional collaboration

Develop frugal AI models that are less energy-intensive and more adapted to emerging market conditions

Apply export controls more flexibly to balance security concerns with the need for global AI access

Focus on being beneficiaries and adapters of AI rather than trying to compete as creators in all markets

AI can lift growth up by 0.8%. […] However, […] there is very different degree in preparedness, in exposure, in access. […] The accordion between advanced economies and emerging market developing countries is wide open.

speaker

Kristalina Georgieva

reason

This comment provides concrete data on AI’s potential economic impact while highlighting the stark disparities between countries, setting the stage for the discussion on inequality.

impact

It framed the central tension of the discussion – AI’s potential vs. the risk of widening global inequalities. Subsequent speakers addressed ways to close this gap.

Consuming AI can deliver to people just on the basis of having a cell phone. […] What I’m talking about is translating AI into an integral part of economic growth.

speaker

Kristalina Georgieva

reason

This distinction between consuming and producing AI introduces nuance into the discussion of how countries can benefit from AI.

impact

It shifted the conversation to focus on how countries can move beyond being mere consumers to active participants in the AI economy.

We are seeing, at least on the African continent through Smart Africa and different institutions, where we are now starting to work on frameworks that allow for cross-border data flow.

speaker

Paula Ingabire

reason

This comment provides a concrete example of regional collaboration to address AI challenges, offering a potential solution to issues raised earlier.

impact

It introduced the idea of regional cooperation as a way for smaller countries to overcome resource constraints and compete in the AI economy.

There are technologies available now that allows you to have ring-fenced data in other countries where you can still preserve the integrity and the sovereignty of this data.

speaker

Hatem Dowidar

reason

This comment introduces a technical solution to a political problem, suggesting a way to balance data sovereignty concerns with the need for cross-border data sharing.

impact

It opened up discussion on how technological solutions could help overcome regulatory barriers to AI adoption.

The real question is, can we grow that? And can we reach Rwanda? We can, but only under one circumstance, that you get Rwanda and Tanzania and Uganda and Kenya and Ethiopia. You get the East African community to decide together that they will all use that data center.

speaker

Brad Smith

reason

This comment provides a concrete strategy for how smaller countries can attract investment in AI infrastructure through regional cooperation.

impact

It shifted the discussion towards practical solutions and the importance of demand-driven strategies for AI development in emerging economies.

Trust in AI is so much higher in the markets that we’re talking about, those that have the potential to be left behind. They are in the 80s and the 90s.

speaker

Bill Thomas

reason

This insight about higher trust in AI among emerging markets introduces a surprising and potentially important factor in AI adoption.

impact

It added a new dimension to the discussion, suggesting that emerging markets might have an advantage in terms of public acceptance of AI technologies.

Overall Assessment

These key comments shaped the discussion by progressively moving from identifying the challenges of AI adoption in emerging economies to exploring potential solutions. The conversation evolved from highlighting disparities to discussing concrete strategies for regional cooperation, technological solutions to regulatory challenges, and leveraging unique advantages of emerging markets. This progression created a more nuanced and solution-oriented dialogue about how to ensure AI benefits can be more equitably distributed globally.

How can regional collaboration in AI be implemented effectively, particularly in Africa?

speaker

Paula Ingabire

explanation

Regional collaboration was highlighted as crucial for addressing resource constraints and creating a unified AI strategy in Africa. Further exploration of implementation strategies is needed.

What specific policies and regulations are needed to enable cross-border data flow and AI collaboration in developing regions?

speaker

Paula Ingabire

explanation

The minister mentioned work on frameworks for cross-border data flow, but more detailed research on specific policies is needed to facilitate regional AI cooperation.

How can data sovereignty concerns be addressed while enabling regional AI infrastructure sharing?

speaker

Hatem Dowidar

explanation

Dowidar mentioned technologies for ring-fenced data in other countries, but further research is needed on how to implement these solutions while satisfying national regulators.

What strategies can be employed to rapidly increase the number of data scientists in Africa?

speaker

Brad Smith

explanation

Smith identified a significant shortage of data scientists in Africa compared to Europe. Research on effective training and education strategies is crucial.

How can AI be leveraged to accelerate permitting and licensing processes for clean energy projects globally?

speaker

Brad Smith

explanation

Smith discussed the potential for AI to speed up permitting processes, but more research is needed on specific implementation strategies across different countries and regulatory environments.

What are the most effective ways to stimulate AI demand in developing countries, particularly through government adoption?

speaker

Brad Smith

explanation

Smith suggested government adoption of AI for public sector services as a way to kickstart demand. Further research on best practices and implementation strategies is needed.

How can the global community address the widening gap in AI preparedness between advanced and developing economies?

speaker

Kristalina Georgieva

explanation

Georgieva presented IMF research showing a significant gap in AI preparedness. Further research on strategies to close this gap is crucial for inclusive AI development.

What role can international financial institutions play in supporting regional AI initiatives in developing countries?

speaker

Kristalina Georgieva

explanation

Georgieva suggested expanding regional-based economic assessments to support AI initiatives. More research is needed on specific support mechanisms.

How can AI be effectively integrated into early warning systems for climate disasters in developing countries?

speaker

Hatem Dowidar

explanation

Dowidar mentioned ongoing work with UNDP on AI-powered early warning systems. Further research on implementation and scaling of such systems is needed.

What are the most effective strategies for reskilling and upskilling workforces for the AI era, particularly in developing countries?

speaker

Bill Thomas

explanation

The discussion touched on the importance of education and skills, but more research is needed on specific strategies for workforce development in the context of AI.