Imagine a humanoid robot with not two, but six arms, as if it had dropped straight out of Indian mythology. This is the MIRO U robot, called a ‘Super Humanoid’ by its creator, the Chinese company Midea. Announced just a month ago, this, to me, is the perfect metaphor for 2026—a year when tech will conjure up weird new things, even as the novelties of 2025 get mainstreamed, and most of the new tech will be driven not by Silicon Valley, but China.
, I said you could buy a robot by the end of 2025. I was correct cause you can take home China’s Unitree R1 for $5,900 (a little over Rs 5.3 lakh), while the equally useless US’s 1X Neo can be preordered for $20,000 (almost Rs 18 lakh).
In 2026, tech will go beyond making just promises to actually keeping them. And the promise won’t just come from Silicon Valley, but like the sun will rise from the East—from Shanghai, Shenzhen, and Hangzhou, even as San Francisco VCs debate endlessly about the AI bubble.
The “Bubble” that’s coming, or not
At the onset, let us tackle the question that has driven much of the AI debate in 2025 and will continue in the next year: Is the AI bubble about to burst? As I wrote earlier this year, there is indeed a bubble, one that will need correction, and would cause losses to some, but is unlikely to burst like the dotcom bubble a quarter century ago.
That’s because this time around, with behemoths in the fray, the correction isn’t about AI crashing but about who gets to build it. And the money is not going to just a few thousand lines of HTML code, as in the dotcom bubble, but to building AI infrastructure, which was the original goal of computing’s founding fathers.
But most importantly, the correction isn’t about the technology failing but the playing field levelling out, tilting in ways that should terrify Western policymakers, cause it’s going East, to China.
Look at the numbers: despite everyone from the IMF to your local newscaster screaming “dot-com bubble-like crash”, venture capital for AI didn’t abandon ship. Instead, they doubled down with the first half of 2025 seeing AI startups attract a mind-boggling 53% of all global VC funding, 64% in the US.
Does it smell of fear or ferocious conviction when OpenAI raises $40 billion at a $300 billion valuation, xAI, a novice in the AI field, raises $10 billion, FieldAI gets $405 million to build “universal robot brains’, and Cohere $500 million? This isn’t the dying flutters of an extinct butterfly, but the first pulses of a gambit that’ll be trillions of dollars worth soon.
The pattern is clear: capital isn’t fleeing AI; it’s getting mature. In 2026, this maturation will tilt away from hyped-up startups and slick demos with no path to profitability, and toward the unsexy, essential engines of the real revolution: the agentic AI stacks.
The real AI story of 2026 will be found in the boring, the mundane. Deloitte projects an exponential rise in enterprise use of generative AI via autonomous AI agents. The big story of 2026 is unlikely to be a headline-grabbing new model, but swarms of autonomous agents that read documents, work in customer service, identify supplier risk, and get invoices approved. These aren’t sexy. Nobody will make TikToks about them. But they’ll generate tangible, massive value for everyone involved—the provider and the user of the tech.
2026 will see the emergence of agentic AI stacks, in which companies build the infrastructure, tools, and workflows that enterprises need to deploy autonomous agents at scale. No flashy announcements, only visible profits from AI applications and infrastructure that can be executed fast and cheaply.
The year of Chinese robotics domination
Many new robots were announced in China in 2025—exponentially more than those anywhere else. Most are still expensive proof-of-concepts. But much like Midea’s six-armed, wheeled MIRO U, they are proving to be a deep philosophical divergence that is departing from a Western fixation on anthropomorphising.
Why emulate the human shape, with all its constraints, when you can create for superhuman abilities? China’s attitude toward robotics drives this paradigm shift, and it is charging ahead full steam.
While the West is trying to build the perfect humanoid, China is attempting to manufacture the practical, what-can-we-put-to-work model through mind-blowing vertical integration and cost engineering. While Tesla’s Optimus may have $150,000 in component costs, Chinese producers can now make up to 90% of the parts for their robots within their borders, slashing prices rapidly. The Unitree robot, which you can buy today for a fraction of the cost of its Western rivals, is the first salvo in a price war the West is neither ready for, nor can it hope to win.
Subsidies are a push not to protect its robotics industry, but to build the entire ecosystem in-house. It includes factories, supply chains, chip equivalents (such as Huawei’s Ascend), and AI brains (DeepSeek, Qwen). They are weaving them into a single, scalable industrial policy.
The ambition is for a 1-million-unit market by 2030, and China is positioning itself to capture most of it. The recent “Robot Olympics” in China and the viral videos of robots dancing on stage with Chinese pop stars demonstrate China’s ambition.
A world of two tech ecosystems
The US-China tech war moved from cold to superhot under President Trump this year. Tariffs were turned into weapons, rising to 125% on certain goods. Export controls were tightened, and even Nvidia’s deliberately crippled H20 chips for the Chinese market were banned. The aim was to suffocate China. The outcome, instead, was the supercharging of China’s entire tech ecosystem.
Washington’s fatal miscalculation was assuming its sanctions would hammer Chinese innovation. Instead, it only hampered the viability of the West’s own tech.
By the time the harshest US blow fell, Chinese tech giants had already pivoted, stockpiling chips, building hybrid systems that used domestic silicon, and doubling down on homegrown AI models that, as shown this January, were better and cheaper to produce than Western ones. Alibaba, Tencent, and others didn’t just survive the US attack; they improved on dimensions such as cost and speed that few Western companies had bothered to explore.
2026 will accelerate this bifurcation into a permanent division. The world will increasingly work in two parallel tech stacks—a Western one centred on OpenAI, Google and Nvidia, and a Chinese one based on DeepSeek, Alibaba and Huawei.
For Global South nations, the decision is more likely to be made for them based on economic feasibility, availability, and the absence of political preconditions. In other words, most of them are likely to favour China. Hence, the real danger to the West’s technological hegemony is not necessarily espionage, but rather better economics and strategic patience, which the Chinese have proven they are best at.
Sovereignty, Quantum, and the infrastructure of power
The US-China rift points to a more profound shift: the emergence of sovereign technology. The wars in Ukraine and the increasingly fraught situation in Taiwan underscore that data pipelines and cloud infrastructure are geopolitical assets. Thus, 2026 will bring demands for localised, compliant data regimes from both countries and corporations.
Cyber resilience and data sovereignty are about to move from IT issues to essential, board-level priorities for business continuity. The era of blindly trusting the fate of critical infrastructure to a globalised, remote cloud is coming to an end.
Meanwhile, the quantum computing race is shifting from physics to engineering. 2026 is not going to be the year when quantum breaks an applicable encryption scheme or designs a miracle drug, but it’ll be the year that paves the way for us to achieve exactly that one day.
Advances in quantum error correction are taking us beyond the “Noisy Intermediate-Scale Quantum” (NISQ) era and into more stable, reliable systems. The story of quantum computing is emerging out of the shadow of many “ifs” to a definite “when”. And that transition will usher in a fresh wave of government funding and startup energy dedicated to quantum software.
China’s deep, quiet investments in the field are likely to yield more announced milestones in 2026, demonstrating that even Quantum is an actual marathon, not a solo sprint.
When the Personal is AI
In India, we have a tradition of metaphoric marriage to nature, where a woman or man marries a tree. 2025 became the year of the union of carbon and silicon when the first marriage between an actual human and an AI was officially solemnised as a 32-year-old Japanese woman, Yurina Noguchi, married an AI-generated persona named Lune Klaus Verdure, with the vows exchanged between her and a digital figure displayed on a smartphone screen. And it fit right in: if 2024 was the year humans had affairs with AI, the future was bound to see marriages.
This did not mean that the crisis of suicides being prompted after chats with AI bots, which I talked about in last year’s piece, was reduced. Instead, they multiplied enough in September detailing at least 13 suicides, assaults or murders emanating from chats with AI bots.
Perhaps an AI ban for people of certain ages or mental dispositions, like the teen Social Media ban enforced by Australia and cheered on by parents all over the world, isn’t too far in the offing.
Despite all this, the truth remains that tech-induced loneliness will only rise worldwide in the new year. So much so that in 2026, a marriage between an AI bot and a human, far from making global headlines, will not even elicit raised eyebrows.
Instead, what will become a cause of great concern—indeed it already is—are AI deepfakes. I have been keeping track of AI deepfake porn on the web for some years. This year, I found them missing from PornHub, who seem to have cracked down despite it not making much news.
So, is the net free of celebrity and nonconsensual porn? Quite the opposite. They seem to have migrated to normal social media portals, alarmingly on Instagram, where handles promoting horrendous deepfake semi-porn, both of celebrities and normal people, are popping up faster than weeds after a monsoon.
CRISPR gets crisper
If 2026 is the year AI gets down to business in the silicon domain, when it comes to the biological carbon, this will be the year when CRISPR gene-editing graduates into something much more extraordinary than just treating sickle cell disease and beta-thalassemia, and takes a magnificent leap towards personalised medicine that is affordable.
For the first time, researchers at the Children’s Hospital of Philadelphia developed a custom CRISPR therapy in just six months for a single infant, KJ, born with a rare, fatal metabolic disorder called CPS1 deficiency. They corrected a specific mutation in his liver cell. KJ’s story thus becomes the ultimate proof-of-concept for the future of medicine, which isn’t just about curing diseases but engineering one-off genetic solutions for single-patient use.
2026 will see the expansion of that promise from one to many, as gene editing advances toward treating the root causes of diseases in the liver, heart, muscles, and brain. Advanced clinical trials may soon bear fruit for the hereditary disorder transthyretin amyloidosis (a contributor to heart failure), as well as muscular dystrophy.
Some scientists are pushing CRISPR’s potential to solve even more complex, multi-gene problems, such as tackling high cholesterol and HIV. The change this will bring is profound: Instead of a “one-size-fits-all” pill, you get a “designer fix” platform curated to your needs.
The unglamorous truth about building the future
So, what will 2026 be all about? In AI, I suspect, it is the year when its grand narrative collapses under the weight of its own hype and is replaced by something much more complex: the reality of implementation.
When it stops being mere potential and actually starts living it, and as a species, we humans might finally stop asking if AI will kill us all, and begin looking at it for what it is: the most brilliant tool we have ever created. We’ll stop enquiring about AI consciousness, and shift our preoccupation to API latency, inference costs, and supply chain security for critical chips.
Those who will eventually “win”, by and large, will be those who can successfully navigate the unglamorous: the 10,000 tiny incremental advances that remake manufacturing, logistics, and healthcare from the inside out. It will be a year of Chinese firms prevailing through masochistic execution and scale, and of Western VCs, at long last, having to spot enduring value—not just viral vapourware.
And somewhere out there, in a lab in either Philadelphia or Shenzhen, a scientist will prepare a guide RNA for a patient whose mutation is unique in the world. They will not issue a press release. They will DESIGN, DELIVER, and move on to the next patient. That’s the way the future is being built—not with one drug that reaches millions, but with millions of drugs that are each tailored to individuals.
Thus, no single technology will define 2026. Instead, it’s going to be a multiplex, parallel revolution—in AI pragmatism, robotic utility, geopolitical tech stacks, and now biology—all evolving in concert. Soon, the narrative will cease to be one of hype and become the arduous, brilliant, and thoroughly human toil of integration, scale, and, ultimately, individualised solutions that actually matter.
So 2026 ahoy, and why just six? The world is ready for a sixteen-handed robot. Just make sure they can thread the needle of human expectation.
