Why capital is moving from bits to atoms

**Shrehith Karkera** (0:00)
Hello, folks, you're tuned in to Finshots Daily. In today's episode, we take a look at how AI and advanced robotics are changing the economics of global manufacturing with dark factories, and what this transition means for India's industrial strategy.
Before we head to the story, here's a quick word from Team Ditto. We're hosting a free two-day insurance masterclass that helps you build real financial security by understanding health and life insurance the right way. Right from understanding how to protect your family, choosing the right cover amount, and knowing what truly matters during a claim, to how hospitals process claims, the mistakes bios usually make, and how to choose a policy that won't disappoint you when you need it the most. We will explain it all in plain language. Head to the link in the description and save your spot. Now to the story.
History has a rather uncomfortable lesson about what happens when a society starts running out of people. Take the Roman Empire, for instance. At its peak, agricultural workers produced food, craftsmen manufactured goods, and merchants transported them across regional borders. Soldiers secured trade routes, and workers paid taxes that funded state infrastructure. Then Rome grappled with repeated shocks, such as invasions, wars, and food supply problems that led to a population decline. Historians say these demographic contractions affected agricultural production, wages, prices, and government revenues. Because a smaller population meant that you had fewer people to cultivate the land, manufactured goods, and fewer taxpayers. This situation highlights the core problem of demographic decline. An economy doesn't just need money, but also needs people to create things, to provide services, and to consume what others produce. So if a working age population continues to shrink over several decades, the negative economic effects compound over time across the entire system. Economists actually have a more formal version of this problem. Stanford economist Charles Jones argues that long-term growth ultimately depends partly on people producing new ideas. Fewer people can therefore mean not just fewer factory workers, but fewer researchers pushing the technological frontier forward. But now, AI introduces a strange possibility into that equation. If machines eventually become capable of performing research, managing production, and operating other machines, economic capacity could increasingly be manufactured like capital, shifting some of the constraint on growth away from population and towards things such as chips and factories and energy. Yeah, that sounds like science fiction. Wait until you look at what's happening in China. A couple of days ago, Chinese humanoid robot manufacturer Unitary Robotics debuted on the Shanghai Stock Exchange. The stock surged almost six-fold on its first day of trading, briefly valuing Unitary at more than $50 billion.
For context, the company generated only about 1.7 billion yuan, which is around 250 million USD at today's exchange rate, in revenue in 2025 alone. And the interesting part isn't the robots themselves, but what they could do to manufacturing. Chinese factories are already testing humanoid robots for inspection, sorting, assembly and logistics. Over the last few decades, if a company wanted to manufacture something at scale, it went to a country where workers were cheap. In the 1970s, China became the world's factory, because it had an enormous workforce, low wages, and a huge network of suppliers that could turn raw materials into everything from electronics to cars. But now, automation has started to change that equation. China has spent years building increasingly automated factories. Some have even gone so far that they are called dark factories, where manufacturing facilities can operate with very little human intervention. At BYD's Zhengzhou manufacturing complex, the company says that automated robots handle 98% of the welding, assembly and quality inspection. And that isn't limited to China. Even in India, at Polymatech's semiconductor facility in Kanchipuram, robots assemble chips inside a clean room through the night, while engineers largely just monitor operations from outside. But hyperautomation creates a different kind of vulnerability too. Polymatech's director estimates that a serious power fluctuation could destroy around $1 million worth of production. That's where humanoid robots come into the picture. They are being designed around the physical environments humans already use.
A conventional industrial robot is usually built for one very specific task. If you want it to perform another task, you may have to redesign the production line around it. However, a humanoid robot, at least in theory, could walk into a workstation designed for a human and perform several tasks without requiring the entire factory to be rebuilt. And that could mean if a robot can perform the same repetitive task for longer hours with greater consistency and without needing breaks, companies may eventually stop thinking about where they can find the cheapest workers and focus more on where they can produce the cheapest. For instance, a factory in China may still have higher wages than one in India. Still, if it has better automation, cheaper robots and a better component ecosystem, reliable electricity and engineers who can integrate AI into production, it could still manufacture at a lower cost than India. In that world, cheap labor is no longer enough to guarantee that factories will move to India. This is where China's advantage becomes particularly interesting. China installed roughly 295,000 industrial robots in 2024 alone, accounting for 54% of all new installations globally, while India installed a record 9,100. So essentially, China is building a new ecosystem. It has manufacturers producing the components, engineers developing the software, factories where robots can be tested, and an enormous industrial base capable of adopting them at scale. As more factories automate, demand for robotic components increases. As more robots are deployed, manufacturers collect more data and improve their systems. And finally, as production scales, the cost of the technology falls. All of this makes automation viable for even more factories, and creates a feedback loop that countries with a large human-centric base struggle to replicate. And investors appear to be betting that this feedback loop is becoming commercially important. According to Crunchbase, companies broadly classified under physical AI, including robots, autonomous vehicles, drones, industrial automation, and sensors, all of this attracted $47.4 billion of venture funding across 521 deals in the first half of 2026 That was nearly 80 times higher than the first half of 2025, and almost 4 times the amount raised in the second half. Robotics itself has seen an extraordinary surge. In fact, they have also surpassed fintech companies in new investments as of the first half of 2026 And that could indicate that the AI investment story is beginning to move beyond companies that produce information on screens, and towards companies that can manipulate the physical world. India, however, has a different opportunity. We don't necessarily need to beat China at building humanoid robots. Our advantages could lie further up the stack. Someone still has to design the production system, maintain the machines, train the AI, integrate different pieces of equipment, and troubleshoot the factory when something inevitably goes wrong. Those jobs require a very different skill set from the repetitive factory work that automation is replacing. A country that buys robots without developing the workforce to operate and improve them could end up with highly automated factories, but little domestic capability to build or manage that technology. That means India's real challenge isn't choosing between people and machines. It is making sure that its people move up the value chain quickly enough as machines take over the lower end. Another economic question also lies beneath all of this. If machines eventually allow companies to produce more goods with fewer workers, productivity could rise dramatically. Consumers could benefit from cheaper products. Companies could earn higher profits, and countries could produce more without needing an appropriate increase in their labour force. But the gains would not necessarily be distributed evenly. A factory worker whose job disappears because of automation doesn't automatically become an AI engineer. Without enough new opportunities, the economy could become significantly more productive, while a section of its workforce becomes economically redundant. That is why the transition matters so much for India. Our demographic advantage has always been one of the biggest arguments for why manufacturing should move here. A large working-age population can be an enormous asset when factories need millions of workers. But it can become a liability if the economy creates fewer jobs than the number of young people entering the workforce. The countries that win the next manufacturing race may not simply be the ones with the cheapest workers. They may be the ones that can combine labour, robots, energy, software, and industrial know-how at the lowest total cost. The goal therefore shouldn't be to replace India's workers with robots, or to protect every job from automation. It should be to make the combination of an Indian worker and an intelligent machine more productive than either one could be on its own. Because if China is building factories where robots do most of the work, India's opportunity may be to build an economy where people become exceptionally good at telling those robots what to do. And that could determine whether India merely becomes the world's next low-cost manufacturing hub, or one of the world's most efficient manufacturing hubs.