Hannah Mullane (1:35)

Apple has committed to spending $500 billion.

Ryan Reynolds (1:38)

In the United States over the next four years.

Hannah Mullane (1:41)

Just a snapshot of the huge amount of investment happening in the U.S. in the race to build the infrastructure needed for AI, it may be unsurprising that two countries, the U.S. and China, dominate in this field. Between them, they're responsible for 90% of the world's artificial intelligence infrastructure. But what does that mean for everyone else?

Matthias Neubauer (2:01)

What we do see is a gap between some countries on this planet versus other countries and regions. Europe and other parts of the world definitely have to step up their game. And it's not easy because it requires a lot of capital expenditure.

Hannah Mullane (2:16)

We'll hear about the impact of this growing AI gap and how some businesses are coming up with innovative solutions to try and keep up, including working when everyone else is asleep.

Shiko Gittel (2:26)

You have to find time when not everybody's hogging their resources, like 4am, 5am because then Kenya has not woken up, Europe has not woken up. The US is sleeping. Yeah. So the only person you're competing with is China and India.

Hannah Mullane (2:40)

That's the AI divide, the haves and the have nots all coming up on this episode of Business Daily.

Nicholas Volovic (2:51)

This is an intermediate place where you just clean your shoes. Here, in order to keep the dirt.

Matthias Neubauer (2:59)

Off, you close the doors.

Nicholas Volovic (3:03)

And this is the first part.

Hannah Mullane (3:06)

Nicholas Volovic is a computer science professor at the National University of Cordoba in Argentina. He's running what counts as one of his country's most advanced AI computing hubs from a converted room at the university. A stark difference from the infrastructure you might find 5,000km away in the US.

Matthias Neubauer (3:26)

Let's enter.

Nicholas Volovic (3:27)

But the noise will be unbearable.

Hannah Mullane (3:36)

Nicholas has been repurposing a lot of old computer equipment to run the site.

Nicholas Volovic (3:40)

We are always trying to push the budget. We used very old servers from Supermicro that were from 2012, and we repurposed them adding GPUs. So we added stuff like that to get a new life to the computer. It was the only way to get something related to AI working.

Hannah Mullane (4:05)

To develop AI technology, you need a lot of investment money that Nicholas says the Argentine government is unwilling to give. And he doesn't have much luck getting funding from private companies either.

Nicholas Volovic (4:17)

We are trying to be imaginative and try to convince the more the state governments or even foreign finance institutions or multilateral banks to get loans.

Hannah Mullane (4:31)

How do you try and convince these institutions to give you the money? And how likely are they to say yes?

Nicholas Volovic (4:38)

The strategy that we use is basically human resources. If you don't have people that understand and can handle that technology, that can adapt that technology, you will always be part of a company or other governments that give you the infrastructure and the knowledge. So it's a kind of sovereignty argument that we are trying to use. That's quite important in a sense, because you cannot control anything if you're using a foreign country or a foreign company. We also try to convince the multilateral banks to lend us money using the same sense we need local development.

Hannah Mullane (5:19)

As Argentina lags behind, the country's workforce looking for jobs in AI end up going elsewhere.

Nicholas Volovic (5:25)

The brightest minds in chemistry, astrophysics, computer science, biology, they go to work for data science and AI companies abroad. And you cannot do anything to get them. What can you offer them? A good hardware? No, you can offer good hardware. Can you offer a good salary? No, you cannot offer a good salary.

Hannah Mullane (5:49)

It's a frustrating situation for Nicholas as he tries to convince the people that hold the purse strings that AI research is worth investing in. The president of Argentina Javier Milei announced plans last year to make Argentina a global hub for AI. He said the country had an abundance of cheap electricity and a highly skilled workforce. He even passed a new law offering tax breaks and other incentives to companies willing to invest in AI infrastructure there. But with an unstable economy and as Nicholas pointed out, a flood of young minds moving abroad to work in the sector, attracting big tech corporations is proving difficult. This strategy is happening all over the world as countries pledge billions of dollars of investment into AI and bring in initiatives to encourage private companies to build their data centers in their countries or competing for that corporate investment.

Podcast Advertiser (6:43)

Chinese tech giant Alibaba launching an advanced reasoning AI model overnight.

Shiko Gittel (6:48)

We aim to mobilize a total of 200 billion euros for AI investment in Europe. The Indian government is preparing to launch.

Ryan Reynolds (6:59)

India AI Compute Portal, a platform that will enable key stakeholders to request computing capacity.

Hannah Mullane (7:06)

So why is it so important?

Ryan Reynolds (7:07)

If you host AI hardware within your borders, it does give a government territorial jurisdiction to, say, impose regulations that can govern and shape the development of these very influential technologies.

Hannah Mullane (7:19)

Zoe Hawkins is a research associate at the Oxford Internet Institute and has been investigating which countries dominate in terms of AI hardware and what that could mean for the future.

Ryan Reynolds (7:29)

For many, it's an issue of national security. They might make the argument that with increasing geopolitical tensions, you know, there's a greater appetite for supply chain resilience and a fear about international disruptions. It's not entirely cut and dry. There's a lot of reasons that are compelling, but it comes with trade offs. Like any policy decision, the reason that.

Hannah Mullane (7:48)

People are seeing this as an opportunity is there is a lot of money in AI, isn't there? They're seeing the future growth of this digital industry and thinking, I need a piece of that pie.

Ryan Reynolds (8:01)

That's absolutely right. From both a corporate perspective, which is exactly what you'd expect. You're hearing people describe it as an AI race. It's certainly motivating people to make big investments. It's also pushing governments to think about it more holistically and have a bit of a strategy around how do we make our jurisdiction really attractive and build out all of these different capacities?

Hannah Mullane (8:22)

Zoe Hawkins, you're listening to Business Daily on the BBC World Service.

Podcast Advertiser (8:30)

In business, they say you can have better, cheaper or faster, but you only get to pick two. What if you could have all three at the same time? That's exactly what Cohere, Thomson Reuters and Specialized Bikes have since they upgraded to the next generation of the cloud, Oracle Cloud infrastructure. OCI is the blazing, fast platform for your infrastructure, database, application development and AI needs where you can run any workload in a high availability, consistently high performance environment and spend less than you would with other clouds. How is it faster? OCI's block storage gives you more operations per second cheaper. OCI costs up to 50% less for computing, 70% less for storage and 80% less for networking better. In test after test, OCI customers report lower latency and higher bandwidth versus other clouds. This is the cloud built for AI and all your biggest workloads right now with zero commitment. Try OCI for free. Head to oracle.com strategic that's oracle.com strategic.

Ryan Reynolds (9:36)

Hey, it's Ryan Reynolds here from Mint Mobile. Now I was looking for fun ways to tell you that Mint's offer of unlimited Premium Wireless for $15 a month is back. So I thought it would be fun if we made $15 bills, but it turns out that's very illegal. So there goes my big idea for the commercial. Give it a try@mintmobile.com switch upfront payment of $45 for three month plan equivalent.

Shiko Gittel (10:00)

To $15 per month required new customer offer for first three months only speed.

Ryan Reynolds (10:03)

Slow after 35 gigabytes of networks, busy taxes and mintmobile.com.

Hannah Mullane (10:08)

I'm Hannah Mullane and today we're looking at the AI divide. The race for AI dominance between China and the US means these two countries are way ahead of the rest of the world when it comes to AI infrastructure and investment. And while China should close the gap on AI models this year, the US still has a big advantage, with most of the big players in tech choosing to base themselves there. Nvidia, for example, based in the US makes over 90% of the computer chips used by the majority of AI companies around the world. And that's just the start. Earlier this year, President Donald Trump announced the creation of Stargate, a private venture between OpenAI, Oracle and SoftBank, promising $500 billion of investment into AI infrastructure in the US. What happens to other parts of the world that aren't seeing this same rate of it's 3:30am in Nairobi, Kenya's capital, and whilst most of the country is sleeping, software engineers are heading into Carla's offices, an AI startup developing artificial intelligence software for businesses around the world. Kenya doesn't have local computing power for Carla to develop its AI models, so instead they use data centers in other parts of the world. And that often means an early wake up call.

Shiko Gittel (11:31)

We have to optimize for when there's low resource constraints. You have to find time when not everybody is hogging their resources. So we have to optimize that, which is like 4am, 5am, because then Kenya has not woken up, Europe has not woken up, the US is sleeping. So the only person you're competing with is China and India.

Hannah Mullane (11:51)

Shiko Gittel is the CEO of Carla. She's been talking to my colleague Michael Kaloki.

Shiko Gittel (11:57)

When you're training a model, you're sending large files of data to computer system somewhere and then it's processing it and sending you feedback. Now you can imagine if you are 100 thousands of you trying to access that same resource, what will happen. It's like your phone. Yeah. It will say these 15 apps are eating up on your memory. It's the same way. And what I love about the engineers is they found where to host their workloads such that it processes faster because, you know, the further it is, the slower it is. Sometimes they leave it overnight to run so that when Americans have gone to sleep, they're able to run their work. Before Europe wakes up, you're able to run their work. We have a small car, sleeping port, but most of the time people go home and come back in the evening because they find it better to work at that time.

Matthias Neubauer (12:53)

In regards to data storage, is that a challenge? A concern that we have here on the continent, whereby sometimes we have to sort of rely maybe on storing data elsewhere, such as America or other countries, and working with it from there.

Shiko Gittel (13:10)

Most of our data does not reside in Africa. In fact, when you say it, I mean, it was so embarrassing. The claim is we don't have enough capacity to host all our data here, But Africa has 11,000 data centers. How good are they? We don't know. Huawei is encouraging every government to host a data center. Do they have their storage capacity? We don't know because right now we are relying on the big tech companies to host our data.

Matthias Neubauer (13:38)

What do you need to be able to grow as you would like?

Shiko Gittel (13:44)

I know that usually the question for what do you need to grow? Is usually money. Money is a second level priority, first level priorities. Believe. We need people to actually believe Africa has the ability to do.

Matthias Neubauer (14:01)

Now, Africa as a continent does not have the same infrastructure that other parts of the world have. Do you feel like the continent will be left behind?

Shiko Gittel (14:10)

Yes. There's no power. So there's a team, part of our team, that is looking at something we're calling resource constraint AI, which looks at how do we build AI in our resource constraint environment. But we cannot wait for everything to be perfect. To start playing. And that's where the idea of resource constrained AI comes in. Edge computing, which does not mean too much power. You can run a model from your phone. People are becoming creative. That's why he's saying we need to showcase some of these ideas to the world. But we have to showcase by building that ecosystem, building the researchers, building the startups, and it has to start with building that demand. I don't think Africa will be left behind because of our constraint. We actually innovate for constraint.

Hannah Mullane (14:57)

Carla, like many other companies, has to rely on the US and China for its compute power. But as other parts of the world race to compete, will that trend change? Exoscale is a European cloud company. Businesses used to have in house equipment to store things like computer programs, emails and apps. They now use companies like Exoscale who can hold all of that stuff on the cloud. And if your business grows, you just increase your cloud space, which makes it easier to scale up and down. It's seeing more businesses looking for a local company to look after their data. Matthias Neubauer is the CEO.

Matthias Neubauer (15:35)

While generally speaking, people and businesses are a little bit worried and concerned if they're not in full control of something, and political developments in the last couple of months certainly made people more concerned and more worried about having full control of the infrastructure. So they certainly want to avoid that someone pulls the plug. They certainly want to avoid that someone has access to data that they should not have access to. And this is where more and more businesses are turning to sovereign cloud providers.

Hannah Mullane (16:08)

From Europe, we know that when it comes to AI and sort of the digital future, China and the US are the big players. So is Europe going to have to catch up in terms of providing that infrastructure and having the data centers and the companies here that are capable of doing what businesses need them to do.

Matthias Neubauer (16:30)

What we do see is a gap between some countries on this planet versus other countries and regions. And what we also see, the AI is becoming more and more critical. Europe and other parts of the world definitely have to step up their game. And it's not easy because it requires a lot of capital expenditure. And then there's also things like trade restrictions coming into play. So certain powerful chips cannot be sold to certain countries due to trade restrictions. From the U.S. for example, the world today is the slowest it will ever be. It will continue to only move faster, and AI will most certainly accelerate that process.

Hannah Mullane (17:12)

We might be seeing more companies looking for more locally sourced digital infrastructure, but that simply isn't possible in all parts of the world and governments might be able to bring in policies to try and persuade private companies to bring their AI businesses to their country. But ultimately it's down to those massive corporations as to where they spend their money. Back to Zoe Hawkins from the Oxford Internet Institute.

Ryan Reynolds (17:35)

So these decisions about investments are being made on commercial terms and not being directed on the basis of equity and or even need. It's a much bigger question as to how do we direct foreign private sector companies to invest in places that isn't their highest return. International organizations, whether that's the UN and other development organizations, the World bank and similar, are very much viewing this as a development issue and agitating within this space and particularly with some of these larger companies to try and encourage them to factor some of those elements in to their investment decisions. But ultimately maybe it's a case of what are the policies and investments that can be made locally. And this is where the pressure comes in for governments to compete with each other almost for these companies investment. And that's why we see such a cascade of different strategies and incentives and policy settings trying to make their own jurisdictions the place that these private sector entities want to put their money. It gets even more interesting when you look at the ownership of this unevenly distributed AI infrastructure. When we look at who owns these companies that run the data centers, we found that of the 33 countries that have AI compute, 18 of them relied on providers from either just China or the U.S. so it sort of aligned with one global powers tech stack. So we had 13 countries relying only on U.S. hyperscaler infrastructure, whereas five were relying only on Chinese hyperscalers. There was 12 countries that are sort of, you could describe as hedging, having a mix of providers from the us, China and Europe. That trend for me is also one that will be interesting to watch. Are more countries sticking with one tech stack or is there more of a hedging as uncertainty in terms of the global order of things comes into question?

Hannah Mullane (19:29)

All the data points towards the US and China continuing to dominate in the AI market. But as more innovation happens around the world and more demand for localized digital infrastructure continues to grow, we might start to see some changes. The reality is AI technology is moving so fast that it can be difficult to predict exactly what the future holds. What we do know is the industry is going to continue to grow, which means its economic potential will grow too. That's all from this episode of Business Daily, produced and presented by me, Hannah Mullaines. Tomorrow Samir Hashmi will be in the United Arab Emirates, which is setting its sights on becoming the world's next AI powerhouse. If you'd like to hear more on our AI series, just search for Business Daily. Wherever you get your podcasts.

Podcast Advertiser (20:21)

Rose is your name, but when you poop, it sure don't smell like roses. You're number two Burns the hair in my nose, but we fault hand in paw. Now your poop don't stink at all. Now that the air is clear, I'll treasure your front and rear, my sweet Ro. Instantly neutralize poop stink with Poop Fighter from World's Best Cat Litter. For the world's best cat.