- 3 minutes ago
Category
📺
TVTranscript
00:18Hello and welcome to Global Live from the BBC World Service. In the next half hour we'll bring
00:23you the best journalism from our teams across the world. I'm Joe Tidy the BBC cyber correspondent
00:27and this week we're in Silicon Valley, California, the beating heart of the ongoing AI revolution.
00:34This strip of land has shaped modern life from personal computers to social media and now it's
00:39at the forefront of the next tech wave. There's a handful of companies here who are fighting for
00:44supremacy in the race to build super intelligent AI and the stakes? Power, money and maybe even
00:51the future of humanity. Later in the programme I'll be exploring one development that was,
00:56until recently, a sci-fi dream, the birth of domestic helper robots. And it's a few smaller
01:02startups sprinkled around San Francisco that could soon be overtaking the tech giants to make AI
01:08servants in your home a reality. And in the wake of America's capture of the Venezuelan president,
01:15why is Donald Trump so keen to control Venezuela's oil when America produces such huge quantities of
01:21oil itself? We'll explore the global struggle for the planet's precious mineral resources.
01:35Artificial intelligence has been an invisible part of all our lives for years. Everything from our
01:40social media feeds to our sat navs has used some form of AI without us really knowing or caring. But
01:47the
01:47rise of generative AI, programmes that can create text or pictures or videos in an instant, has showed
01:53us just how powerful and pervasive this technology is.
02:01OpenAI caught the world and the rest of the tech industry off guard when ChatGPT went viral in 2022. The
02:09free
02:09website became the fastest growing consumer application in history after being used by one
02:14million people in just five days. Today, 800 million people use it weekly, according to the company.
02:21Tech giants like Google and Microsoft are deep in the race too. And new entrants like Anthropic and Elon
02:28Musk's XAI are all fighting for a slice of the trillion dollar dream, as innovative tools to improve
02:34productivity and creativity are unveiled almost weekly. But this isn't just a US story. It's a race too,
02:44with China's Silicon Valley equivalent, Shenzhen. That was evident last year when Chinese AI app
02:50DeepSeq topped the Apple download charts and rocked the markets as US firms realised they no longer had a
02:57monopoly on AI. But the AI revolution is being viewed with alarm by some environmentalists. They fear it's
03:05driving up water consumption that could lead to shortages, although the exact extent of this is highly contested.
03:13Critics also argue that the energy hungry data centres needed to power this tech revolution are
03:18setting back the global drive to bring down carbon emissions. Google and others have already admitted
03:24that their own carbon reduction targets are under strain. But President Trump has shifted Washington's
03:30priorities away from CO2 reduction as the US doubles down on trying to win the AI race. And as ever,
03:36what
03:36happens here has ripple effects around the world. The darker side of AI has been on display in recent weeks,
03:45as users of Elon Musk's AI assistant Grok have repeatedly prompted its image generation tool to
03:51digitally undress women and even children to make nearly nude images of them. Elon Musk has warned that
03:57anyone using Grok to make illegal content will suffer the same consequences as if they uploaded it
04:03themselves. Campaigners are urging governments across the world to tighten regulation on deepfake images,
04:09or at least enforce laws already passed but untested.
04:16The techno-utopian view is that it'll all be worth it. Technologists promise us that the downsides
04:21of AI will soon be overcome and eventually there will be an age of abundance. Sam Altman of OpenAI and
04:28his rival Elon Musk at Tesla, they talk about a time when we won't even have to work and that
04:34inequality,
04:35all too visible here on the streets of San Francisco, will be eradicated.
04:39We think it can be a printing press moment. We are working to build tools that one day
04:44could help us make new discoveries and address some of humanity's biggest challenges, like climate change
04:48and curing cancer. There are other concerns too. Some of the brightest minds in AI are warning that the
04:55birth of an intelligence greater than humanity's might lead to the downfall of humanity. The shock
05:01of ChatGPT led to a global drive around regulations for AI safety. Slow down and make sure that these
05:08models have humanity's best interests at heart. That was the cry, but no more. Just as climate concerns
05:14have fallen off the priority list, so too have discussions about AI safety as governments prioritise
05:20growth over caution. And many experts also caution that a bubble could be building, with investors
05:26down for big losses if the economic promise of AI doesn't live up to expectations.
05:34Whether we like it or not, the race is on and we're all along for the ride. AI is changing
05:39how we live
05:40and work and the next frontier could be in our homes, courtesy of domestic helper robots. Interestingly,
05:46it's not big tech that's leading the charge here. It's nimble start-ups dotted around San Francisco.
05:53And according to some of these companies, it could be this year that we start to see these robots in
05:57homes.
06:00Domestic helper bots are coming. Billions are being poured into this new AI frontier,
06:08a global gold rush underway.
06:10You just go make us a cup of coffee. I'm on.
06:15But how close actually are we to the sci-fi dream of a robot butler?
06:20Oh, holy. I've never seen this in my life.
06:24Good accuracy. Oh, I spoke too soon.
06:27It's not just investor money at stake. The tech will test safety and privacy in our most intimate of
06:33settings, our homes. And as with so many tech races, it seems like it's China versus the US.
06:39And Silicon Valley is taking center stage.
06:43I've come to Palo Alto to meet the CEO of AI startup Sundae.
06:50Yes. Yeah, Tony. How's it going?
06:53Tony Chow began working on his home robot while at uni, but dropped out to build Memo full time.
07:00While many domestic robots in development are controlled by someone teleoperating them,
07:05Sundae has trained its bot to do many tasks completely on its own.
07:09So can it see me around it?
07:13Right now it doesn't react to you, but it can perceive the world with all the cameras and other
07:19sensors. Can you just go make us a cup of coffee?
07:24It's after 2 p.m. Would you like me to go ahead and make that cup of coffee for you?
07:29Yes, please go ahead.
07:31I'm on it. Making your coffee now.
07:34Tony, can I just check with you? There's no one that we can't see doing this for the robot?
07:39This is all autonomous?
07:40Yes. It's one single neural network controlling the whole body movement of the robot.
07:45Wow. It was all going smoothly. Then Tony paused the demo as he saw something was off.
07:58Maybe we need to restart this and restart this.
08:02I think there's a... Can you stop it?
08:06I think we might need to do it again. I saw there's some peeling off here. Sorry about that.
08:12Let's just try again. Maybe I'm overly cautious.
08:18Oh. Holy. I've never seen this in my whole life.
08:24It's really easy.
08:26After a reset, Memo smoothly cleared away the table.
08:31There's obviously work to be done to get it ready for homes,
08:34but there's no doubt this is one of the most capable bots in the world.
08:43How have you got the robot to be this advanced? How have you done it?
08:47So I think normally the way we train AI is to teleoperate the robot in people's homes or in
08:55other places and gather the data and train the AI. And what we're doing is different,
08:59as in we don't need robot data to train the robot. And instead, we build these gloves. And people just
09:05wear their gloves in their homes and collect data for us. And that gives us really diverse data because
09:10we now see like more than 500 homes and also all the different ways they go about doing the task.
09:16Sunday's army of robot teachers are all being paid to do repetitive,
09:20everyday tasks, a reminder of the human drudgery underpinning how these machines learn.
09:25The staff allow cameras and sensors to record their every move at home.
09:29But Sunday plans to start shipping memos next year. So if the bot needs to be remotely controlled,
09:35it will raise privacy considerations for customers, as well as questions about safety in a domestic setting.
09:41Like there will be kids running around. Kids will be climbing on the robot, right?
09:44So we really want to make sure that we put safety as the most fundamental consideration
09:51when we design these robots.
09:55Some robotics companies are already out in the wild, gathering training data in public.
10:02Here it is. Isaac's been trained by watching hundreds of hours of laundry folding. For the last
10:07couple of months, though, it's been doing it on its own for real, here and in six other sites in
10:12San
10:12Francisco. Deployment is the strategy. Nothing counts if it's sequestered away in a lab.
10:20We'll do a live demonstration. There we are. Okay.
10:24This is a lot slower than a human. So some people might be watching this saying,
10:29oh, that's not very impressive. But can you tell us why it is impressive?
10:32This thing can run all day. You often hear that robots are best suited to tasks that are dull,
10:38dirty or dangerous. We want to make clothing folding optional for people.
10:43When you first put this robot here, how long did it take to do one t-shirt?
10:48I think that we were looking at 2, 2.30. So already, what, in how many months?
10:53That's in a month and a half. In a month and a half.
10:55You've got down to 1.30, just over 1.30. We've planned to start shipping this year.
11:00They say it'll be able to tidy up and fold laundry, but it's not clear how much of that
11:04will be autonomous. Across town, at Physical Intelligence, they're not interested in building
11:10robots themselves. They're focused on creating the AI software that can be used to make any robot
11:15capable of doing chores autonomously. The rest of AI, like in ChatGPT, for example,
11:21we're used to training these models on an internet of data, but we don't have that web of data for
11:27robotics. So we want to be able to actually kind of essentially breathe intelligence into any sort of
11:32physical embodiment, whether that's kind of a standard, typical robot, a humanoid robot,
11:37or even something that looks closer to an appliance, for example. Like all these companies,
11:42Chelsea and her team are developing their AI software using reams of videos of humans doing
11:46chores. Their approach is proving successful. Investors like AI giant OpenAI are on board.
11:53Another tech giant, Tesla, is deep in development of its own humanoid robot, Optimus. But it's not
11:59yet clear what it can do. And it's not just in the US where people are getting excited about human
12:05-like
12:05robots. Chinese company Unitree is already dominating the market with its G1 robot, seen here in a demo
12:11given to us by its UK seller, Scan. This is happening right now. People are buying them,
12:16people are coding on them, and people are learning. It's entirely operated. It's not autonomous at all,
12:22as of yet. It's hard to get into China to film, but the humanoid robot industry is so hot there
12:29that the government recently warned a bubble might be building set to burst. What's really interesting,
12:36driving around and speaking to all these robotics companies in Silicon Valley,
12:39is there's a lot of confidence. They think they are building the future, and they're certain it's
12:43going to be a huge moneymaker. But there's also some nervousness as well, because there's lots of
12:48things that we wanted to film and we wanted to show, but we can't. They are scared, because of course,
12:54this is a race. A race to bring to market a domestic robot.
13:00A company that hopes to be the first is One X. The Norwegian-founded company moved here last year,
13:06and millions of dollars are riding on its bot NIO becoming a hit.
13:09Someone is operating NIO right now and watering all the plants. I've just been told this is what
13:14they do all day. So there are people in the office operating NIO with a VR headset, carrying out tasks
13:21and
13:21testing all day long. Good accuracy. Oh, I spoke too soon.
13:28So interestingly, NIO is struggling with this particular handle, but the team here change the
13:34handles all the time, because they're trying to train the robot different ways to open different
13:38handles. They had it there. Founder Bernd Bornic is pushing hard to get NIO out to homes.
13:45So what can NIO do right now for me in my home?
13:50Does a pretty good job of like tidying. So like resetting my living room and my house in general.
13:56Wiping surfaces, like cleaning the counter, like these kind of things.
13:59And that's all someone with a VR headset?
14:01It's a mix. It's a mix. So in my home, we have a lot of data. So a lot of
14:06the stuff in my home can
14:06get automated because we have data there. Periodically, someone kind of steps in and helps if the robot
14:11does not know exactly how to move on. Early adopters will have to be comfortable
14:15then waiving their privacy. And as with all these bots, battery life is going to have to be improved,
14:20as NIO can only work for about four hours at a time. But Bernd promises that NIO will be able
14:25to carry out many tasks and charge itself autonomously by the time they start shipping it.
14:30Your first set of customers then, they have to be quite wealthy,
14:33and they have to be willing for NIO to make mistakes, essentially.
14:36Well, I don't think they have to be quite wealthy. It's thousands of pounds, isn't it?
14:40I don't know how much it is. Well, it depends.
14:42£20,000? Yeah, like are most people who have a car?
14:46So you liken it to a car? Yeah, I mean, it is a pretty affordable car.
14:50It depends on your needs, right? A lot of our early customers are people that actually
14:54will have a lot of value from this. And for a very large subset of that, it will be way
14:59higher
15:00value than having a second car. Getting the right customers is more important in the beginning.
15:06Distribution with the right customers. So we can really use these amazing people
15:09that are our early adopters to help us make this work.
15:14Outside of the tech bubble, some, including the International Federation of Robotics,
15:18think it could take 20 years before domestic bots become truly useful and accepted.
15:24But we have heard that before about other futuristic AI technologies.
15:28So keep your seatbelt fastened, please.
15:33In some ways, brand new revolutionary technology does just sort of creep up on you.
15:38If you think about driverless cars, I've been talking about driverless cars in
15:41my reporting for a decade at least. But now they're here. And they're just sort of normal.
15:47In many cities across the US and in China, driverless cars are an everyday sight.
15:52AI robotics companies are convinced their tech will go the same way. And not only will their
15:58devices be truly useful, we'll all eventually want one in our homes.
16:05Now to South America. Donald Trump has said Venezuela will be
16:09turning over up to 50 million barrels of oil to the US. This comes after the military operation to
16:15remove President Nicolas Maduro from power. But why does America want more oil when it's already
16:20the world's biggest producer? Laura Garcia from the BBC's Global Journalism team explains why.
16:33The United States produces more oil than any other country in the world. And President Trump talks
16:40about it a lot. Take the oil. Keep the oil. Drill, baby, drill.
16:45But he doesn't just talk about America's oil. We're going to have a presence in Venezuela as it
16:51pertains to oil. So why is he interested in more oil when the US has so much of its own?
16:59To understand this, we need to take a deeper look at oil and how it's shaped the US and the
17:04world.
17:06Oil fuels our transport and powers our homes. But it's also used to make plastic, asphalt and even
17:12some lip balms. And to make all those different things requires different types of oil. Depending
17:18on its density, sulfur content and ability to flow, oil is classed from light sweet crude on one end
17:24to heavy sour on the other. It's easier to refine light crude than heavy, which is thicker.
17:31And once that's done, the light is mainly used for products like gasoline and jet fuel. While heavy
17:37oil can be used as fuel for ships, material for roads and lip balm, among many other things.
17:43That's generally reflected in the price. Light sweet crude, the blue line on the chart,
17:48is worth more than heavy, the red line. And so it's more expensive to buy.
17:54And this is key to understanding American oil.
17:58In 2025, the US sold 13.4 million barrels a day. But at the same time, it bought nearly 2
18:07million
18:07barrels each day from other countries. So why not just keep those extra barrels it produces?
18:13It all comes down to light versus heavy. America's oil is 80% light. But most US oil refineries,
18:22like these along the Gulf Coast, were built to deal with heavy. That's because most oil available to
18:29the US in the 20th century was heavy sour crude oil imported from Latin America and Canada.
18:35But in the early 2000s, there was a seismic shift in US oil production. Advances in technology meant
18:42light crude oil trapped in shale rocks could now be extracted at scale. This means there's a mismatch
18:48between most of the oil America has and the type it can refine. Once a refinery has been built,
18:55it's very difficult to change it and requires millions and millions of dollars worth of investment.
19:01It's not worth doing that because as we saw on this graph, America can sell its light crude for more
19:06and buy heavy for less. It makes good economic sense. So let's look at the top 10 countries with the
19:15biggest oil reserves. Many are known to extract heavy crude. Three of them, Venezuela, Iran and Russia,
19:23are currently under sanctions by the United States. Despite this, small amounts of oil from Venezuela
19:29have trickled into the US. That's because after Venezuela struck oil, it was mostly American companies
19:35that helped set up the industry. And that relationship continued for most of the 20th century,
19:41extracting their heavy crude. Unlike other Latin American countries, Venezuela maintained this
19:46positive relationship with the United States. Even during the 1976 nationalization of oil.
19:56But the turning point was when socialist leader Hugo Chavez came to power in 1999.
20:04He asserted state control over the oil industry in particular. He put tougher conditions on foreign
20:11oil companies and that's what the US government and US oil companies didn't really like.
20:16When Hugo Chavez died in 2013, Nicolás Maduro became president and continued these policies.
20:23In 2019, a World Bank tribunal ordered the Venezuelan government to pay compensation to US oil
20:29companies. But it wasn't paid. And that's part of what US President Donald Trump is referring to
20:36when he says that Venezuela has stolen American oil. They took our oil rights. We had a lot of oil
20:41there.
20:42As you know, they threw our companies down and we want it back. Venezuela denies these claims.
20:50Things escalated between the two countries at the end of 2025 when the US military seized oil tankers
20:56and blockaded Venezuelan ports. President Trump said this was to tackle narco-terrorism.
21:04On January 3rd, American troops seized the leader of Venezuela, Nicolás Maduro and his wife.
21:09They took them to the United States to face drug-related charges.
21:14Here's some of what President Trump said following that intervention.
21:17We have the greatest oil companies in the world going and invest billions and billions of dollars
21:22and take out money, use that money in Venezuela. And the biggest beneficiary are going to be the
21:28people of Venezuela.
21:30US and Western sanctions in Venezuela over the last decades have made space for competitors to come in,
21:36like Russia, Iran and particularly China. China has been buying around 90% of Venezuela's oil and is a key
21:44trading partner. And it's this sort of influence that the Trump administration wants to stop.
21:50Well, we want safety there. We want to be surrounded by countries that aren't housing all of our enemies
21:56all over the world.
21:57When it comes to Venezuela's oil, things aren't going to change quickly.
22:00We want safety. Despite Venezuela having the biggest known reserves, around 303 billion barrels,
22:07it exports fewer than a million barrels a day. That's because of sanctions and decades of
22:13underfunding and mismanagement.
22:15The infrastructure has to be, in some cases, believe it or not, rebuilt from scratch. Looking at a
22:22three to four year window before any Venezuelan crude in meaningful volumes happens.
22:28And while there's a push to use more renewable energy in the face of growing climate change
22:32concerns, oil still plays a crucial role in the global order.
22:41Also this week on the World Service.
22:44As protests have rapidly spread across Iran in recent days,
22:47countries, initially driven by public anger over the country's economy,
22:51BBC Persian, in collaboration with BBC Verify, have been monitoring videos posted online to produce
22:57a dynamic map indicating where the protests have broken out. The map clearly shows just how widespread
23:03the challenge to the Iranian government has become. You can read the full story on the BBC News website.
23:10Now we're going to take you to Central America. Since Donald Trump's return to the White House in
23:15January 2025, the United States has taken a much stricter view on immigration, with sweeping efforts
23:21to remove undocumented migrants from the country. Yet despite the threat of deportation, thousands
23:26from Central and Southern America continue to head north in search of opportunity. Among them are
23:32Guatemalans. Nearly 10% of the country's population has relocated to the US in recent decades, and the
23:39remittances they send back home have become a cornerstone of Guatemala's economy. Atahualpa Amoriz
23:44reports on how remittance homes have become an economic lifeline, and a cultural symbol.
23:55My son went to the United States, and from that time, he started to send remesas to
24:00build his house. My mother went to the United States, and she fulfilled her dream.
24:06She made our house in Guatemala. Guatemala, a country made up of around 18 million, is one of the
24:13poorest countries in Central America. Inequality is extreme, especially in rural or indigenous areas,
24:20where poverty, underemployment, and limited access to public services remain widespread.
24:26For decades, migration has been one of the main ways Guatemalans have tried to escape that reality.
24:31The main destination, the United States. For families who stay behind,
24:36migration has one central economic consequence, remittances.
24:48Antonio is a master builder, and has just completed this mansion for his 22-year-old son,
24:54who migrated to the US three years ago.
24:57My son has a dream of having a beautiful house. He has taken two jobs, construction in the morning
25:03and the night in a restaurant. I haven't traveled, but I believe that the hard thing is to be in
25:09the
25:09United States. Every time he gets the money, he sends it for the material, he sends it for the
25:14payment, and so, as the work has been executed.
25:17Globally, remittances are often discussed as financial flows. In Guatemala, they are visible in
25:23everyday life, and in the built-in environment.
25:36Money sent from migrants in the United States has become a pillar of Guatemala's economy.
25:41In many rural municipalities, remittances exceed all other sources of income combined,
25:47including state investment.
25:48George is a local architect who specializes in American-style homes.
25:55From his studio in the center of San MartÃn, Zacatepeque, he designs houses commissioned by
26:00migrants who want to help their families, or dream of one day returning to live in Guatemala.
26:07The cost of the average house is about $100,000, which is $700,000, which is $700,000,
26:14in advance. Here they like the colors pastel brillant, yellow, green, turquoise,
26:21morado. At the end of the day, it is a shock of who has the best.
26:25Most of my clients come from the United States. I can assure that approximately 75% of them
26:30come from there. It is a small percentage of the people who stay here in Guatemala and can
26:34make the houses that someone in that situation of that country could build.
26:40Maria's mother works in the United States and moved there more than 12 years ago.
26:45The money she sends to Guatemala every month allowed her daughter to move from a small house
26:50made of metal sheets to this three-story mansion.
26:54Maria's mother works in the house. Here I have a room, a room, a room and a kitchen.
27:00And at the third level I have four rooms and a refrigerator. My mother is working in cleaning.
27:08She works there for seven days. She is fighting there. She is fighting, said my mother.
27:13She is fighting. She is fighting. She is fighting. She is fighting. She is fighting.
27:24In recent years, reminences have also become a form of insurance against deportation. Since President
27:30Donald Trump returned to the White House, U.S. immigration policy has again tightened. The
27:36administration has intensified enforcement against undocumented migrants, increasing workplace rates
27:42and deportations. Most Guatemalans now living in the U.S. are believed to be undocumented.
27:48It can have to do with a panic that at any moment they are going to deport them and they
27:52have to
27:52more effort and send them as much as they can. For the new migratory policies of Trump, the
27:57send of remesas back to Guatemala is completely transforming the architecture of peoples and
28:03the peripheries of the Guatemalan cities. The amount of remesas received, for example, in the
28:07year 2024, was 21,000,000,000. It is a barbarity. We are talking about the 20% of the
28:13Guatemalan
28:15Guatemala's Remedance Architecture is ultimately a physical record of global inequality, migration
28:21and power built one house at a time.
28:48Thanks for joining us here in California. Next week, Global Eye will be reporting from Israel.
28:52Goodbye.
29:22When the
Comments