The Economist. Hello and welcome to The Intelligence from The Economist. I'm Rosie Bloor. And I'm Jason Palmer. Every weekday we provide a fresh perspective on the events shaping your world.
Some communities in Canada are so remote they aren't connected to the power grid, to roads or anything much. Often these groups are also feeling the brunt of climate change. But as our correspondent found, efforts to install green energy there are still hitting obstacles.
And everyone knows you should get plenty of cardiovascular exercise. Getting the heart revving just has so many positive effects now and later. But you know what else does? Strength training. Good things come to those who lift weights. But first...
Until now, a small number of American companies has dominated the field of artificial intelligence.
firms like Chipmaker Nvidia and Cloud computing's Oracle have attracted vast investment to become some of the highest valued companies in the world. But suddenly that dominance looks less certain. Talk futures are down sharply this morning amid fears of an AI stock bubble burst. Over the weekend, a chatbot released by an obscure Chinese AI startup called DeepSeek became the most downloaded app on iPhones.
Deepseak claims that developing its new model cost a fraction of what American competitors have spent. And yesterday, in response, American tech stocks plummeted. Donald Trump called Deepseak a wake-up call for U.S. companies. The release of Deepseak AI from a Chinese company should be a wake-up call for our industries that we need to be laser-focused on competing to win, because we have the greatest scientists in the world.
So can a cheap Chinese upstart really change the field of AI? Over the course of a day, the euphoria over AI transformed into panic. Geisgruven is our global business writer.
So you saw the NASDAQ 100 stock exchange index drop by 3% yesterday for recovering some of its losses. You saw a lot of the really big tech companies suffer losses as well. So that's companies like Alphabet, Amazon and Microsoft. Though again, some of those companies recovered afterwards. But the really big news was that Nvidia, which is this superstar AI chip maker,
It's share price dropped by over 10% yesterday, and that wiped hundreds of billions of dollars off its market valuation.
DeepSeek is a Chinese AI startup, and over the past month or so, it's released two very impressive AI models. The most recent one is called R1, and was released only a few days ago, and that was the real trigger for all the market chaos.
The thing you have to know about the kind of AI models that DeepSeek has released is that they both perform really quite well. They've done a lot of innovative things under the hood to really get the best performance out of these models, but they are also incredibly efficient and cheap to make.
Because of American trade restrictions on AI chips, Chinese firms have been forced to be very innovative in the way they make AI models. And so the combination of these high-performing models and the fact that they are very efficient is the root cause of the turmoil we saw yesterday in the markets. So Guy, the big question is, is deep-seak actually any good?
Well, at the moment, the models seem to perform quite competitively to Western rivals, so that the kind of models that OpenAI makes, they also seem to have attracted a lot of users. So over the weekend, DeepSeek's app shot to the top of the app store rankings. So they're obviously very popular with consumers as well. So on the face of it, they seem to be both really quite good and basically quite cheap to make.
And they need less computing power to operate, is that right?
Yes, exactly. So the lower costs come from basically using less computing power. And the reason that's important is because over the last couple of years, since the launch of chat GPT, American tech firms have spent an enormous amount of money and made an enormous amount of money, both buying and selling computing power in order to power AI models.
Just last year, if you look at four of the really big tech firms, they spent about $180 billion just on data centers, and most of that was to power AI. That figure is up more than 50% from the year before, and it's predicted to go up by another 30% or so this year.
And so there's been this enormous spending glut on all sorts of pieces of infrastructure connected with AI. And the implication of these new deep-seat models is that maybe that kind of level of spending and that amount of infrastructure may not actually be as needed as we previously thought.
And as you say, Silicon Valley companies have spent years spending billions on AI investments. And this seems to have caught them somewhat on the fly. What are the implications then for the US tech industry?
The kind of clearest initial implication concerns Nvidia, which is a specialist AI chip maker, and that is that it may not be able to sell as many chips as we previously thought. If you could now train your AI models using fewer chips, then that looks like it could be a big problem for Nvidia, which is why its share price sank so sharply yesterday.
And then there are probably quite big ramifications for the cloud computing giants, Google, Microsoft and Amazon. We don't quite know how this is all going to play out yet. If you speak to different people on Wall Street, they say different things. Some people think that this will be good for them because one of the big inputs into the AI that they sell has become cheaper to make.
Other people think that this is actually quite bad news because those were the only companies that had enough cash to spend large amounts of money on building out this infrastructure, and that created a kind of moat for their business model. And if anyone with a reasonable amount of money now can also make these AI models, then that might start to erode that moat. So at the moment, for the big tech firms, we really don't know how this is all going to shake out, I think.
But presumably for us as consumers, more LLMs and more chatbots means more competition, cheaper, better things. That's a good thing, right?
Yes, ultimately this is probably quite good for us as consumers. I think this may well kind of trigger another AI arms race as lots of startups and big companies try to produce ever cheaper AI models. And so we as consumers, I think will benefit enormously from this. And we'll hopefully we'll see more AI being kind of injected into our phones, into our apps, into our laptops and all kind of much more cheaply.
But that said, there are still plenty of uncertainties around this innovation and kind of the ramifications of it as well. We don't entirely know how the new costs and new economics of AI are going to play out. We don't really know what this means for geopolitics. So if America feels like it's starting to lose the AI race, it has going with China,
That could lead to further trade restrictions, and that could be quite bad for consumers and for markets as well. And another uncertainty hanging over all of this is really the ultimate demand for AI. So as much as AI chatbots are quite fun to play with,
In order for AI to really take off, it has to be quite useful and quite well adopted in big companies. And at the moment, big companies are struggling to take AI from pilots to large scale projects. And how exactly the demand plays out will be another factor which will determine the future path of artificial intelligence. Guy, thank you so much. Thanks so much, Rosie.
And yes, I know sometimes it feels like you're hearing about AI and insert random topic here. But believe me, this is a good one. Tomorrow, our science podcast, Babbage, is looking at how artificial intelligence may change sport. Only subscribers can listen. You know what to do.
It was during the pandemic that I started following the story of one green energy program happening in Canada that was working to replace all of the diesel-generated electricity in the country's remote northern and indigenous communities.
These are communities not connected into the power grid of the south. Many are so remote, there is no road that reaches them, so access is limited really to expensive flights on regional airlines, ships in the summer for those that are coastal. And in the winter, when the lakes and rivers freeze, roads are built across them, across the tundra, through the forests to bring in freight.
Burning diesel in these communities is bad for people's health, and it's bad from an emissions perspective, and it's incredibly expensive and onerous to get fuel to these communities by these methods. So I was really interested in efforts to install green energy in these places. It seemed like one of those climate policy slam dunks the Canadian Liberal government had promised, but over time, following this story, I realized it just wasn't that simple.
And so I was the very first cohort to go through the Indigenous Clean Energy Summer Intensive Training Program in 2016. During one of those online conferences that was just so pervasive during the pandemic, I met a man named Darryl Brown, who was already working in this field.
At the time, he had already successfully installed a solar project in Gulbei First Nation in 2018, which had eliminated a quarter of that community's reliance on diesel electricity.
Daryl is indigenous himself of Cree Heritage, and this was in many ways how academics and policy makers have suggested programs like this should be run in Canada. The government should be empowering indigenous leaders to build bridges with indigenous communities to try to make amends for the past treatment of indigenous people. Okay, so not only are you reducing emissions,
And in meeting Daryl, I was immediately inspired by his drive to bring green energy to First Nations. They also become revenue generators of power and electricity. So now they have ownership in their own destiny. And there's jobs, employment, training. So everything that encompasses Indigenous renewable energy work is really very effective, I think, in lifting all of our First Nations out of poverty.
His first project was in a community that had road access, so logistically it was easier than the next one he tried to set up in Seisidene First Nation. In this community, he wanted to build a wind, solar, and battery-powered microgrid. And this community is about 300 people in the very northern stretches of the province of Manitoba, 900 kilometers from the provincial capital.
So this has meant logistical nightmares from the beginning. I was meant to go to Saisudani. Last time I was in Canada, but when I got to the airport for our charter flight, we were all met with disappointments. When this low ceiling would have prevented us potentially from landing,
So he says you want to roll the dice like you're gonna- This was just one example of the things that have slowed this project down for years. The pandemic slowed progress, and with climate change, those ice roads I told you about. The season is becoming shorter and more unpredictable. So getting freight into the community is becoming a much harder task. Another aspect of this is that the provincial utility companies are not incentivized to make deals on these novel projects.
And then lastly, history and politics are playing a role here too. And that might be the most delicate of the situations to navigate. Part of our struggles in the last little while is trying to recover from our forced relocation back in 1956, based on
A former Saisidane chief, Evan Yasi, spoke to me about the lasting trauma his community faces because in 1956, the Saisidane people were forcefully relocated by the Canadian government.
They stripped us from our traditional homeland and dropped us off on the shores of Hudson Bay with no provisions to get us through. By the 1970s, half of the community had died. They faced persecution, impoverished living conditions, and substance abuse also played a role. Eventually, the leaders of the community just took off and brought their people back to their traditional lands. But quite frankly, by then, the damage was done.
And so the psychological trauma that had impacted our people from that is still lingers today, but we are resilient in that way to try to get back on our feet. And so we've been developing the community since then.
They've been relocated by a Canadian government department, Natural Resources Canada, claiming that they were over-harvesting Caribou, which was a traditional food source. And Natural Resources Canada is the very same department task now with overseeing green energy projects in these communities. Initially, former Chief Evan Yassie completely threw his support behind the project.
be more self-sustaining with the challenges that are coming about in terms of high cost of living and the difficulties and construction of Winter Road to make sure that we get all our essential goods into... So the project had the support of the former Chief in Council. But because of all of the delays, there was also a local election to contend with in the First Nation. And a new Chief came in, and he has been more hesitant.
And you know what? No one is blaming him for his hesitancy. This is what Daryl had to say on that. No, the way they've been dealt with over the years, historically, they don't trust anybody and I don't blame them.
But let's say she done a, with any First Nation project that we touch, community engagement is the foundation of the entire project. In order to get community buy-in, you have to show up, get into their community, and show them you appreciate them.
So the project is still moving ahead, but it's been trimmed down now. They are hoping that they can still install solar panels and batteries that will cut down the community's reliance on diesel fuel by about a quarter. But the year's long slog continues.
The particularly interesting case to look at right now, given Prime Minister Justin Trudeau and his liberal parties slide in popularity. As part of their climate-first mantra, the Liberals had promised to completely eliminate diesel use in these northern communities by 2030. And it's quite clear looking at this one case that it was never going to be achievable.
No one from Natural Resources Canada agreed to be interviewed for this segment. But a spokesperson did point out in an email that $453 million in funding has been committed since 2018 to fund these projects.
But the project in Sei Sei DNA has taken years of work and government backing to get even this far. That is one community, and there are 178 small remote settlements dotted across Canada's Arctic and Sub-Arctic region that would need the same kind of commitment.
Making the financial case for the development of these projects is quite hard. But from a policy perspective, it is the perfect opportunity for the country to marry its efforts on indigenous reconciliation with their climate commitments, particularly because indigenous communities are feeling the brunt of climate change's impact so far.
The business case is that you can't leave these communities alone any longer. They should have had help 20 years ago and to burn that much diesel and let those emissions keep going when we're supposed to be going towards net zero. Somebody's got to help them. So this is a start. It's not the solution, but it's a partial solution. And if this was a promise Justin Trudeau's Liberals wanted to make, it needed more financial and logistical support to actually deliver on what they set out to do.
January, everybody's a gym bunny. New Year's resolutions still have their momentum. Those crowds will soon level out, hide wager. But one trend seems likely to stick around for those who continue to work out, aiming not just to be fitter, but also stronger. Strength training seems to be having a bit of a moment.
Tim Cross is a senior science writer for The Economist.
So there's a lot of competition in the gyms now for the dumbbells, not for the treadmills. Yeah, so Planet Fitness, which is one of America's biggest gym chains, they actually mentioned this on an investor call recently. They said a lot of their locations, they're taking out some of the cardio stuff and replacing it with strength equipment.
The same thing's happening in the UK. There's a company called Places for People, which runs a lot of the municipal sports centres up and down the UK. And they've said they've seen the same thing and it seems to be being driven by their customers. Their customers are saying we'd like more strength training stuff and maybe a bit less of the cardio, please.
Which kind of runs counter to what we thought we knew about what's good for you, right? The suggestion has been to regular cardio vascular exercise. Well, it doesn't run counter, but it is added onto it. So yeah, that has been the suggestion. But in 2010, the WHO, the World Health Organization, they actually added to that and said, as well as the cardio, you should try and do at least two muscle strengthening sessions per week because that's good for you separately from doing cardio. Well, why though? What does that add to the mix?
So two things I think and one is kind of metabolic or systemic. And I guess the way to think about it is that in the same way that your bones are more than just scaffolding on which the rest of your body is hung. They produce white blood cells and red blood cells and that kind of thing. Muscle is also more than just that organ that moves your limbs around. It actually plays a role in metabolism as well.
And so there's evidence that if you strengthen your muscles and grow them, it improves various markers of metabolic fitness. So it makes you more sensitive to insulin. Your cholesterol levels seem to go down. It seems to reduce your risk of things like heart disease or diabetes or cancer. And this is a sort of slightly newer finding. So the evidence isn't quite as iron-clad as it is for cardio, but I think it's pretty persuasive.
And then the second thing I think a lot of the interest in this from doctors in the WHO comes from the idea that regular strength training can prevent something called sarcopenia, which is the sort of fancy medical name for the decline in strength and muscle mass that everyone experiences as they get older, which in the medical view now is in some way kind of avoidable.
It's probably not entirely avoidable, but you can certainly slow it down dramatically. And there are good reasons to try. So from what we can tell, left to their own devices, your muscles start to shrink, maybe sometime in your 30s. And that disproportionately affects the fast twitch muscles that are responsible for explosive strength. And that process speeds up as you get older and older. And by the time you get into your 60s or 70s or 80s, enough of your muscle has gone that it can lead to what doctors call problems of daily living, which just means
you find it hard to say get dressed independently, or carry shopping, or work in the garden, or even stand up from a chair. And when it gets to that point, and when people's strength and balance is that poor, you start to see an increase in the risk of things like force, which can be quite a big killer of the elderly. It can make it harder for you to recover from injuries or illnesses, and it can even worsen the prognosis of some diseases.
So in a sense, doing a little bit of pumping iron is kind of a little sip of the fountain of youth.
Yeah, that's the idea. And these days more and more people could use a sip at the fountain of youth because most countries are seeing their average age rise. So this sarcopenia, since it's to some extent an inevitable consequence of aging, is only going to become a bigger problem and a bigger burden for health systems. People have done studies in quite elderly populations, even people in their 80s, that have shown that you still retain some capacity to grow muscle even then. And that strength exercise is beneficial and does help.
make you more independent. And then interestingly, there's a secondary consideration, I guess, that people are maybe starting to pay a bit more attention to, which is that strength training may help undo some of the downsides of these increasingly popular GLP1 drugs.
So these are things like wager V that people take for weight loss. They definitely work for weight loss, but the problem with them is, as far as we can tell, some of the weight loss comes from fat, which if you're overweight is good, but some of it seems to come from muscle, which generally isn't. So if they're good or become even half as popular as some of the projections suggest they are, then we might see an awful lot more people prescribed strength training to try and counteract some of the ill effects.
So Tim, I imagine you're convinced now you are about to step out and go pump some iron right now. Yeah, I like rock climbing, which I like to tell myself is sort of strength training. So I think I'll stick with that for now. Tim, thanks very much for your time. Thanks, Jason.
Subscribing to Economist Podcast Plus helps make our journalism possible. Thank you. Search Economist Podcast Plus for a free trial. That's all for this episode of The Intelligence. Let us know what you think of the show. You can get in touch at podcasts.economist.com. We'll see you back here tomorrow.