How AI Is Rewriting Business in Emerging Markets
Lessons from the front lines
-
-
IDEA IN BRIEF | ||
THE QUESTION Silicon Valley reads about AI in emerging markets one headline at a time. What does the whole picture look like, and what does it mean for the Valley? | THE INSIGHT Across China, India and the Gulf, AI is being rewritten for the front line: voice-first, frugal, physical and carried on state-built rails, at a speed and scale the Bay Area cannot match alone. | THE IMPLICATION Silicon Valley wins by keeping innovation at its core and letting the front line shape the edge, in a win-win: learn in China, earn and hire in India, set up in the Gulf, and innovate in Silicon Valley. |
Scenes from the front line
The monsoon stalled in June 2025, and about 38 million Indian farmers knew it was coming. A text message on their phones, sent through a program led by UC Berkeley’s William Boos and the University of Chicago, carried an AI forecast that the rains would pause. They paused for about 20 days. The conventional weather models had missed it. The farmers who acted on the message could plan their sowing.1 The farmers were not alone. Across the emerging world, AI is quietly changing daily life.
In Wuhan and two dozen other Chinese cities, a passenger steps into a taxi with no one in the driver’s seat. At its peak in March 2026, Baidu’s Apollo Go was giving more than 350,000 of these fully driverless rides a week.2
In Shanghai, a commuter orders a coffee in the Luckin app for as little as 9.9 yuan, about a third of what Starbucks charges. She never speaks to a cashier. Every cup she buys becomes a data point for Luckin’s next decision on price, product and store location.3
In a small Indian town, a seller on Meesho calls for help with an order. A voice agent picks up. It is one of about 300,000 seller calls Meesho’s AI agents take every day.4
And in Abu Dhabi, a resident’s license renews itself. TAMM AutoGov, the emirate’s AI public servant, handles routine renewals before people have to ask. In 2025, the government reports, 95 percent of requests on its TAMM super app were resolved through AI.5
None of these scenes took place in Silicon Valley. Yet each of them is AI being rewritten: redesigned for a customer who pays in small amounts, speaks before she types, lives on her phone and expects the state to provide the rails. This is what I mean by the front line. The novelist William Gibson once observed that “the future is already here. It’s just not very evenly distributed.”6 In AI, more of that future is arriving on the front line than many in Silicon Valley realize.
So I put one question to a room of Silicon Valley’s AI builders: Where will your company’s next billion users come from, and who is building for them today?
AI is no longer just being exported to emerging markets. It is being built, tested and scaled there.
Why I see it this way
On October 20, 2026, I gave a talk to the Silicon Valley chapter of the Berkeley Haas Alumni Network in Menlo Park, a few miles from where much of the world’s frontier AI is being built. The room was full of alumni who work at the frontier of AI. For many of us, it was the first time we had discussed emerging markets from the front line in the context of AI. The questions were sharp: on risk, on the concentration of power, on the rise of cheaper open models from China, and on whether AI will bring prosperity first or the perils that many experts predict. I learned as much from the room as I shared with it. This article is my attempt to answer those questions.
I did not come to that room as an AI technologist. I came as someone who spent nearly a decade leading a sourcing operation in China and 18 years running retail in the Middle East, including as Group CEO–Retail at Landmark Group, and leading retail in India. These are markets where margins are thin, customers are demanding and the rules can change overnight. Today I teach them at Berkeley Haas. What I offer is a street view of AI: how it looks from the store floor, the factory, the farm and the government counter.
That view is useful because there is so much noise, hype and confusion about AI, including whether it is a bubble. With Anthropic reportedly discussing a listing at about $2 trillion,7 and Aswath Damodaran warning that the industry would need “two, three, four trillion in revenues eventually” to justify the spending,8 the question is fair. The valuations may prove to be a bubble. The usage, as the rest of this article shows, is not.
Silicon Valley reads about emerging markets one headline at a time. DeepSeek’s low-cost model wiped $589 billion off Nvidia’s value in a single day.9 The Gulf signs another gigawatt-scale data center deal. Each story is filed under a different heading. Put them together, and a pattern appears that I have not seen anyone map for Silicon Valley: the most consequential experiments in applied AI are now running in China, India and the Gulf, at a scale and speed the Bay Area cannot match alone.
The world has tilted
Start with the economics. Goldman Sachs projects that by 2050 China’s economy will reach about $41.9 trillion, the United States about $37.2 trillion and India about $22.2 trillion, in 2021 dollars.10 The timing is uncertain. The direction is not.
The users are already there. Emerging and developing economies hold about 85 percent of the world’s people and, by my calculation from IMF figures, about 72 percent of global growth in 2026.11 Microsoft finds that generative AI use among working-age people is 28.8 percent in the Global North and 16.2 percent in the Global South.12 By my calculation, the Global South is already home to about two-thirds of the world’s generative AI users, and its use is far from saturated.
Yet AI capacity sits far from that growth. High-income countries hold 17 percent of the world’s people but 87 percent of notable AI models and 91 percent of AI venture funding.13 Leading Western firms earned just 17 percent of their revenue in emerging markets, at a time when those markets held 36 percent of world GDP.14
Growth lives in emerging markets. AI capacity does not, yet. Whoever bridges the two will create the value. Ram Charan put it in four words: “The world has tilted.”15
Where the next billion will come from
China and India together have about two billion internet users, roughly six times the 324 million in the United States, and about 1.9 billion smartphone users, again about six times America’s 311 million.16 At India’s internet penetration of about 63 percent, more than 500 million Indians have yet to come online. They are already showing up in AI. By web traffic, India is the second-largest market for ChatGPT, and about three in four of its visits come from outside the United States.17 Anthropic reports that India is second only to the United States in Claude.ai use.18
These markets skipped steps. For hundreds of millions of people, the first computer was a phone, not a PC. They went from cash to QR codes without ever carrying a credit card: India’s UPI handled 24.1 billion payments in September 2026, about 802 million a day.19
I have watched this leap first-hand in India. In 2009 and 2010 I led The Mobile Store, then the largest retail chain in India, with more than 2,000 stores. We were selling the phones that carried the leap. The spread of the mobile phone brought a flow of information to people who had never had it. It made quick commerce possible, kept children learning during Covid and after, and put weather advice in farmers’ hands. WhatsApp, Instagram and YouTube turned millions of ordinary Indians into content creators, and retailers used digital tools to transform the experience inside their stores. The next leap is AI: voice-first, in local languages, carried on the same phones, payments and habits. That is where the next billion AI users will come from, and I expect them to adopt it faster than the last billion adopted the internet.
Nobody can know the future, and I hold my forecasts loosely. But one signal is worth watching. I expect that by 2030 India will have more people using AI than the United States, and that emerging markets will be home to most of the world’s AI users. Microsoft’s AI Diffusion Report, which today puts India at 18.5 percent of its working-age population, will show whether I am right.20
What the front line teaches
Andrew Ng has called AI “the new electricity.”21 Like electricity a century ago, it is being wired in differently in each region. Across China, India and the Gulf, five lessons stand out, summarized in Exhibit 1. None of them is a lesson Silicon Valley learns at home. Retail, where I learned to manage, is where AI proves its worth fastest, because the decisions are frequent and the margins are thin.
EXHIBIT 1
Five lessons from the front line
What China, India and the Gulf teach about how AI is built and scaled
Lesson | What it means | Where to see it |
Voice beats text | The next billion speak to AI before they type to it | Meesho, Sarvam, Falcon Arabic |
Frugal and mass market first | AI wins on cost and reach, not in the premium tier | DeepSeek, Doubao, Unitree |
AI gets physical | Factories, streets and farms, not just screens | Xiaomi, Apollo Go, WeRide |
The state builds the rails | Public platforms and state capital carry AI nationwide | UPI, Aadhaar, Stargate UAE, HUMAIN |
Retail is the testing ground | Frequent, thin-margin decisions are where AI proves its worth | Meesho, Luckin, Blinkit |
The usual question about AI is what it can do. The better question is who creates the value, where and how. On the front line, value is moving from incumbents to AI-native challengers such as Meesho; from the core to the edge, to a mid-range phone in a local language; and from tools to outcomes. Front-line firms sell the job done, not the software seat. They use cheap models first, because an AI assistant that costs five cents to help someone buy a $1.50 basket loses money on every sale.
Jensen Huang calls applications the layer of the AI stack “where economic value is created.”22 Not everyone agrees; Franklin Templeton argues that much of the value may stay with scarce chips, memory and power.23 But if applications do capture the value, emerging markets are where most of the users, and most of the testing, will be.
The next billion will speak to AI before they type to it.
India: from talent to AI engine room
India has never lacked people who can master the machine. By my own count, about one in four of the AI leaders seated with the President at the 2026 White House AI leaders luncheon were India-born, the largest group after Americans. Yet leadership in AI rises elsewhere. Invention needs a whole chain: labs, capital, customers and institutions willing to bear failure. Silicon Valley has the whole chain. India has many of the links, but not always the chain.
What India does have is public rails at a scale no one else has built. AI agents can now pay on UPI, in pilots inside ChatGPT and Claude,24 and India’s national ID system ran 283 million AI face authentications in November 2025, more than double a year earlier.25
Blinkit: forecasting at street level. Indians shop daily, in small baskets, close to home. Blinkit delivers in about ten minutes from 1,544 small dark stores, and in the quarter to June 2026 it earned about $1.8 billion in revenue and reported positive adjusted EBITDA.26 The promise works only if each store predicts what its few streets will order in the next hour. India scaled a format that struggled in the West.
Silicon Valley is now helping India build the rest of the chain. Amazon has committed $48 billion to India through 2030,27 Microsoft $17.5 billion for AI and cloud infrastructure28 and Google $15 billion for an AI hub in Visakhapatnam.29 India is Amazon’s largest product-development center outside Seattle,30 and Microsoft’s India Development Center is one of its largest research hubs outside Redmond.31 Google DeepMind’s Gemma 3n model, built to run on phones, was developed in India,32 and Meta is hiring engineers in Bengaluru for AI infrastructure and custom chips.33 Nvidia has more than 8,100 employees in South Asia, 18 percent of its global workforce.34 As Jensen Huang told an audience in Mumbai: “India produced and exported software. In the future, India will export AI.”35
India’s 2,117 global capability centers employ 2.36 million people, and more than 1,200 of them already build AI.36 As the H-1B fee rose to $100,000, six Big Tech firms added about 33,000 workers in India in 2025.37 Retailers are following. In 2011, Best Buy closed its branded stores in China. On October 1, 2026, it announced that its India capability center would move to TCS and become an AI-native capability center.38 Lowe’s center in Bengaluru built the company’s “AI foundry” and cut supply chain simulations from weeks to about an hour.39 Even India’s IT giants are moving from selling hours to selling outcomes: Infosys CEO Salil Parekh says AI services could be “almost half of our work” in two or three years.40
Whether these centers are owned outright or run by a partner, the work is no longer back office. The shift is from labor arbitrage to capability ownership. The links are becoming a chain.
China: the world’s fastest lab
From 1999 to 2008, I was managing director of Alghanim Industries’ China sourcing operation. With our Chinese colleagues, we built Wansa into one of the largest private labels in consumer electronics in the Gulf, and we sourced everything from automobiles and industrial lighting to furniture and steel. I evaluated the feasibility of a Kirby steel building plant in Suzhou, and I visited more than 40 cities and over 1,000 factories. In those years I watched the economy build world-class infrastructure and turn itself into the factory of the world, the strategic sourcing base for companies everywhere.
Today I see China making a similar leap, this time with AI. If India shows how far AI can reach, China shows how fast it can move. China installed about 354,000 industrial robots in 2025, 59 percent of the world total.41 Xiaomi’s smart factory in Beijing runs around the clock with almost no workers on the line and can make up to 10 million flagship phones a year.42 Chinese firms shipped about 25,000 humanoid robots in the first half of 2026, more than 95 percent of the world’s total,43 and Unitree sells a full-size humanoid for about $5,900.44 Baidu’s Apollo Go has given more than 22 million driverless rides in 27 cities, from Wuhan to Dubai.2
The same speed shows up in software and in cars. Chinese open models such as Alibaba’s Qwen now lead US models in downloads,45 and China counts 602 million generative AI users.46 In one week in July 2026, Chinese models carried 57 percent of the tokens US firms used on OpenRouter, and Coinbase halved its AI spending by moving staff to them.47 BYD sold 2.25 million battery-electric cars in 2025, ahead of Tesla’s 1.64 million.48
Luckin: speed and data beat heritage. Luckin was founded in 2017 to take on Starbucks. It survived an accounting scandal and a Nasdaq delisting in 2020, rebuilt around an app-only model, and in 2023 its revenue passed Starbucks China. By the end of 2025 it had 31,048 stores.3 The app-only model in the opening scene is the point: it turns every sale into a lesson.
The loop is short: research sits next to the factory, regulation allows fast testing, and a product can be proved in China and then scaled worldwide. That is why Harvard Business Review urges executives to treat China “as a platform to accelerate innovation,” as Pfizer, Michelin, BMW, Bosch and Nvidia already do.49 Many Silicon Valley companies are already learning from China, whether they know it or not.
The Gulf: capital, power and an AI-first state
The Gulf plays a different role. It brings sovereign capital, abundant power and governments that want to be AI’s first customers. The UAE named the world’s first minister for AI in 2017, and today 73 percent of its working-age people use AI, the highest share of any country.20
The build is enormous. Stargate UAE, a partnership of G42, OpenAI, Oracle, Nvidia, Cisco and SoftBank, is building a 1-gigawatt AI cluster within a planned 5-gigawatt campus.50 Microsoft is investing $15.2 billion in the UAE from 2023 to 2029, with US licenses to ship Nvidia’s latest chips.51 Abu Dhabi’s MGX closed a $49 billion AI fund in 2026 and backs both OpenAI and Anthropic,52 and Saudi Arabia’s HUMAIN has secured 18,000 of Nvidia’s top chips.53 Nuclear power from Barakah already supplies a quarter of the UAE’s electricity.54
Abu Dhabi: the government as AI customer. TAMM, the super app behind the self-renewing license in the opening scene, offers more than 1,150 government services in over 90 languages, and in 2025 it reported 36 million office visits avoided.5 The UAE now plans to move half of its federal services to agentic AI within two years.55 Abu Dhabi also became Uber’s first fully driverless city outside the United States, with robotaxis from China’s WeRide.56
In the Gulf, the most demanding AI customer is often the state. For Silicon Valley, it is where compute, capital and early customers meet.
The win-win: learn, earn, set up, innovate
Silicon Valley does not win in emerging markets by exporting what it built for Palo Alto. It wins by combining what it does best with what the front line does best. The Valley’s core is innovation: frontier models, platforms, deep capital and a culture that forgives failure. The front line’s strength is the edge: speed, scale, frugality and customers who adopt faster than anyone expected. The model may be the core; language, interface, price point and data residency are the edge.
There is a deeper reason the exchange must run both ways: AI cannot scale on Silicon Valley alone. Most of the stack is built elsewhere. China refines 91 percent of the world’s rare earths and makes 94 percent of its magnets.61 Taiwan’s TSMC makes 72.5 percent of the world’s foundry chips,62 and South Korea’s chipmakers supply 83 percent of the memory that AI runs on.63 The Gulf is building gigawatts of power for data centers.50 India has 21.9 million software developers and is on track to pass the United States by 2030.64 And, as we have seen, most of the world’s people and AI users live in emerging markets. The minerals, the chips, the energy, the talent and the users all sit outside Silicon Valley. The Valley’s biggest role is the application layer built on top of them.
I see four moves, one for each region and one for the Valley itself, as Exhibit 2 shows. Each is an exchange: Silicon Valley gains something it cannot get at home, and the region gains something it cannot yet build alone.
Learn in China. You do not need to win the China market to win from China. You need to be close enough to learn its speed: run a learning node there, test fast and scale globally, with clear rules on what knowledge is shared. Learn its speed, but keep sensitive technology and customer data at home. Silicon Valley gains speed. China gains partners, investment and research jobs, as companies such as Pfizer, BMW and Bosch build innovation work there, and a route to global customers.
Earn and hire in India. To earn is to treat India as a market: the way to reach a billion customers, not just a source of talent. To hire is to build AI capability there. WhatsApp has about 500 million users in India and more than 100 million in the United States.65 By users, India is already bigger than the US for several American platforms. By revenue, it is still small. That gap is the opportunity. As C.K. Prahalad wrote, if we start recognizing the poor “as resilient and creative entrepreneurs and value-conscious consumers, a whole new world of opportunity will open up.”66 Silicon Valley gains a billion customers, the talent to serve them and growth. India gains good jobs and the missing links in its chain.
Set up in the Gulf. Build compute and partnerships in the Gulf to serve the region, and design for redundancy from day one. Silicon Valley gains compute, capital and early customers. The Gulf gains know-how and a path beyond oil: Microsoft’s UAE investment, for example, includes a $1.5 billion stake in the Abu Dhabi AI company G42 and US licenses to bring advanced chips into the country.51
Innovate in Silicon Valley. This is the Valley’s biggest role. Its models, platforms, capital and global reach are unmatched. The opportunity is to build the applications with and for these markets, and then scale them worldwide. Exhibit 3 sets out what each side offers, what it gains and the risk that comes with it.
EXHIBIT 3
The win-win, region by region
What each side offers, what it gains and the risk to manage
Region | What it offers | Silicon Valley’s move | What the region gains | Risk to manage |
China | Speed: integrated R&D and manufacturing, robots, EVs, open models | Learn: a learning node with clear knowledge-sharing rules | Partners, research investment and a route to global customers | Geopolitics, export controls, knowledge leakage |
India | Talent, users and public rails at billion-person scale | Earn and hire: AI-native capability centers and products for the next billion | Good jobs, capability ownership and the missing links in its chain | Low revenue per user, data rules |
The Gulf | Power, patient capital and compute | Set up: compute and partnerships to serve the region | Know-how and a path beyond oil | Regional conflict, chip approvals |
Silicon Valley | Models, platforms, capital and global reach | Innovate: build and scale the applications, with and for these markets | Speed, talent, users and growth | Building for the wrong customer |
What Silicon Valley leaders should do
Founders: design for the next billion. Build voice-first, in local languages, for the phone. Price for India, then scale up. Plug into public rails such as UPI and WhatsApp rather than building your own.
Company leaders: turn capability centers into product centers. Make your India center an AI-native product center, not a cost center. Open a learning node in China with clear rules on what is shared. Test one AI product in a high-velocity market before a global launch, and spend time on the ground.
Investors: look past the chip makers. Tomorrow’s winners may be the companies that use AI best, not only those that make the chips. Back AI-native challengers in India and the Gulf, price in the geopolitical risk and value patience.
Prosperity for billions
“AI is one of the most important things humanity is working on,” Google’s Sundar Pichai told a town hall audience in San Francisco. “It is more profound than … electricity or fire.”70 If that is true, AI’s greatest test is not in Palo Alto. It is whether AI lifts the billions on the front line. I see four frontiers where it already does, each one a business or a public service at scale, not charity.
Reducing poverty. In Telangana, more than 7,000 chili farmers using AI advice raised yields by 21 percent and earned about $800 more per acre each crop.71 Lenders such as Aye Finance in India and M-Shwari in Kenya use AI credit scoring to reach people with no credit history.72
Education for all. In Karnataka, an AI copilot cut a teacher’s lesson planning from about 40 minutes to 10, in English and Kannada.73 Since the 2025–26 school year, every public school in the UAE teaches AI, from kindergarten to grade 12.74
Affordable healthcare. Qure.ai’s AI chest X-ray reads helped screen 7 million people for tuberculosis in a year.75 India’s free telemedicine service, eSanjeevani, has given 12 million patients diagnoses aided by AI.76
A sustainable climate. The monsoon forecast that opened this article reached 38 million farmers, up to four weeks ahead.1 Google and BCG estimate that AI could help cut 5 to 10 percent of global emissions by 2030, although that estimate is contested.77
This is where emerging markets and Silicon Valley need each other most. The front line has the problems, the users and the rails. Silicon Valley has the models, the platforms and the capital. Together they can make prosperity the story of AI.
So we return to the question the room kept coming back to: prosperity or peril? Some of AI’s most respected voices paint a doomsday scenario. Geoffrey Hinton puts the chance of human extinction from AI within 30 years at 10 to 20 percent,78 and a 2025 book on the subject is titled If Anyone Builds It, Everyone Dies.79 I take these warnings seriously. But the future of AI is not a forecast to wait for. It is a choice, and the choice is ours. We decide which way it goes: in what we build, whom we build it for and how we govern it.
My personal wish is that the warnings are proven wrong, and that AI brings prosperity and benefits all. Let us focus all our energy on making the planet a great place to live for all, by harnessing the power of AI. The answer will not be decided only in Silicon Valley’s labs. It will be decided by whether AI reaches the farmer, the patient and the teacher on the front line, and whether it works for them.
Prosperity or peril: the choice is ours.
The farmers, patients and teachers on the front line are not waiting for us. They are inviting us.
Learn in China. Earn and hire in India. Set up in the Gulf. Innovate in Silicon Valley. The world has tilted, and the next decade of AI will be won by the leaders who keep innovation at their core and stay present where AI is being rewritten.
UC Berkeley Haas has a phrase for it: what starts in the Bay moves the world. In AI, it will move the world only if it is built with the front line in mind.
NOTES AND SOURCES
Figures are as reported in the sources below, read in October 2026. Forecasts are projections, not outcomes.
1. UC Berkeley News (September 16, 2025); Google (September 15, 2025); Center for Effective Global Action (March 16, 2026).
2. Baidu, first-quarter 2026 results (May 2026): peak weekly rides, more than 22 million cumulative rides and 27 cities, including Dubai.
3. Luckin Coffee, fiscal 2025 results; The Edge (February 2024).
4. Meesho, Q4 FY26 shareholder letter (May 7, 2026), via YourStory and Free Press Journal.
5. Abu Dhabi Department of Government Enablement, 2025 highlights (January 8, 2026).
6. William Gibson, as quoted in The Economist (December 4, 2003).
7. Financial Times, as reported by Fortune, “Anthropic reportedly plans a $2 trillion IPO in October” (August 13, 2026). The valuation had not been formally set.
8. Aswath Damodaran, interview reported by Benzinga, via Yahoo Finance (January 27, 2026).
9. Forbes, Nvidia’s record one-day loss after DeepSeek (January 27, 2025).
10. Goldman Sachs, The Path to 2075 (December 2022); real GDP in 2021 US dollars, and emerging-market share of global equity market capitalization.
11. International Monetary Fund, World Economic Outlook (October 2026): world growth 3.1 percent, emerging and developing economies 3.7 percent, advanced economies 2.1 percent; author’s calculation on purchasing-power-parity weights (about 60 percent at market exchange rates).
12. Microsoft On the Issues, “The continued state of global AI diffusion in 2026” (September 21, 2026); author’s calculation of the Global South’s share of users.
13. World Bank, Strengthening AI Foundations (November 2025).
14. McKinsey Global Institute, Winning the $30 Trillion Decathlon: Going for Gold in Emerging Markets.
15. Ram Charan, Global Tilt: Leading Your Business Through the Great Economic Power Shift (Crown Business, 2013).
16. Internet and smartphone users, 2026 metrics: Statista and SG Analytics (China); DataReportal, Digital 2026, and Statista (India); Statista (United States). Counts vary by method; IAMAI-Kantar and CNNIC publish different totals, but every source agrees on the direction.
17. Similarweb, chatgpt.com traffic by country (September 2026).
18. Anthropic Economic Index, India brief (February 16, 2026).
19. National Payments Corporation of India data, via DD News (October 1, 2026).
20. Microsoft, AI Diffusion Report (Q2 2026), share of working-age population using AI.
21. Andrew Ng, “Why AI Is the New Electricity,” Stanford Graduate School of Business (March 2017).
22. Jensen Huang, “AI 5-layer cake,” NVIDIA blog (March 10, 2026).
23. Divya Mathur, “Beyond Silicon Valley: Where AI Value Is Really Created,” Franklin Templeton (August 5, 2026).
24. TechCrunch, NPCI, Razorpay and OpenAI pilot UPI payments in ChatGPT (October 9, 2025); Razorpay, agentic UPI payments on Claude (February 2026).
25. Press Information Bureau, Government of India, on Aadhaar face authentication (December 2, 2025).
26. Eternal, Q1 FY27 shareholder letter (July 23, 2026).
27. IANS, “Amazon CEO meets PM Modi, announces $48 billion investment in India through 2030” (June 25, 2026).
28. Associated Press, “Microsoft investing $17.5 billion in India for AI and cloud infrastructure” (December 9, 2025).
29. UPI, “Google to invest $15B in its first AI data hub in India” (October 14, 2025).
30. Business Today, Amazon’s Hyderabad campus opening, remarks by Amit Agarwal (August 22, 2019).
31. Microsoft India Development Center, as reported by Fonearena (March 2025).
32. Business Standard, interview with Manish Gupta of Google DeepMind (July 27, 2025).
33. CXO Digital Pulse, on Meta’s Bengaluru hiring for AI infrastructure and custom silicon (February 24, 2025).
34. Revelio Labs, Nvidia headcount by region (March 2026 data); Nvidia, Form 10-K for fiscal 2026; Digitimes, on Nvidia’s India engineering centers (September 2023).
35. Bloomberg, via Fortune, Jensen Huang and Mukesh Ambani at Nvidia’s AI Summit in Mumbai (October 24, 2024).
36. Zinnov and Nasscom, India GCC Landscape Report FY2026, as reported by The Tribune and Business Today (2026).
37. Rest of World, “Silicon Valley can’t import talent like before. So it’s exporting jobs” (February 9, 2026).
38. TCS press release, Best Buy India capability center becomes an AI-native capability center (October 1, 2026).
39. YourStory, “How Lowe’s India GCC is driving AI innovation for global impact” (April 17, 2025).
40. Salil Parekh at the Bloomberg New Economy Forum, as reported by PTI and The Economic Times (October 13, 2026).
41. International Federation of Robotics, World Robotics 2026, press release (September 24, 2026).
42. Lei Jun, Xiaomi founder, on the Changping smart factory and its capacity of 10 million flagship phones a year (July 2024).
43. IDC humanoid robot shipments, first half of 2026, via Yahoo Finance (September 30, 2026).
44. Unitree Robotics, launch of the R1 humanoid robot (July 2025).
45. MIT Technology Review, on Chinese open models overtaking US models in downloads (February 12, 2026).
46. China Internet Network Information Center (CNNIC), statistical report on internet development (December 2025 data).
47. Fortune, “China’s Moonshot, Z.ai and DeepSeek are challenging U.S. AI labs, and beating them on cost” (July 26, 2026).
48. Electrek, BYD and Tesla 2025 sales (January 2, 2026).
49. Jing Qian and Lizzi C. Lee, “Where Does China Fit in Your Company’s Innovation Strategy?” Harvard Business Review (June 2026).
50. G42, Stargate UAE announcement (May 22, 2025); Zawya (October 14, 2025).
51. Microsoft On the Issues, on Microsoft’s investment in the UAE (November 3, 2025).
52. CNBC, MGX closes $49 billion AI fund (July 1, 2026); Bloomberg, on MGX’s OpenAI and Anthropic investments (February 17, 2026).
53. Data Center Dynamics, on HUMAIN’s Nvidia GB300 order.
54. Emirates Nuclear Energy Company, via Khaleej Times (September 8, 2025).
55. Dubai Media Office, UAE Cabinet framework for agentic AI in federal services (May 18, 2026).
56. WeRide and Uber, Business Wire (November 25, 2025).
57. Reliance Retail Ventures, media release on the acquisition of METRO Cash & Carry India (December 22, 2022).
58. Euromonitor data cited by Reuters and Fortune; Starbucks, agreement with Boyu Capital (November 4, 2025).
59. Srikant Gokhale, Karthik Naidu and Vijay Viswanathan, Navigating Emerging Markets, including Chapter 15, “The Blueprint of Strategic Framework for Winning in Emerging Markets.”
60. Srikant Gokhale and Rajiv Lal, Balanced Adaptation, Harvard Business School Working Paper 26-075.
61. International Energy Agency, commentary on rare earth supply chains (October 23, 2025): China’s share of rare earth separation and refining and of sintered permanent magnets.
62. TrendForce, global foundry revenue ranking, second quarter 2026.
63. Counterpoint Research, high-bandwidth memory market share, second quarter 2026 (SK hynix 50 percent, Samsung 33 percent), via Seoul Economic Daily (September 3, 2026).
64. GitHub, Octoverse 2025.
65. Meta, Q2 2024 earnings call, via TechCrunch; Mark Zuckerberg on US WhatsApp users (July 25, 2024).
66. C.K. Prahalad, The Fortune at the Bottom of the Pyramid (Wharton School Publishing, 2005).
67. Encyclopaedia Britannica, 2026 Iran war; AWS and UAE Ministry of Defence statements (March and April 2026), as compiled by Wikipedia; Associated Press, via Fortune (October 11, 2026).
68. TCS, second-quarter FY27 earnings call (October 8, 2026), via Investing.com.
69. World Bank, World Development Report 2026: The Promise of Artificial Intelligence, press release (August 4, 2026).
70. Sundar Pichai, at a Recode and MSNBC town hall with Google and YouTube in San Francisco, as reported by Gizmodo (January 2018). The original reads “more profound than, I dunno, electricity or fire.”
71. World Economic Forum, Saagu Baagu program in Telangana: 21 percent higher yields and more than Rs 66,000 (about $800) extra per acre per crop cycle.
72. World Bank and International Finance Corporation research on AI-based credit scoring in emerging markets, citing Aye Finance (India) and M-Shwari (Kenya).
73. Microsoft Research, Shiksha Copilot pilot in Karnataka (February 2024).
74. Gulf News, on the introduction of AI teaching in all UAE public schools from the 2025–26 school year.
75. Qure.ai, FY24–25 Impact Report.
76. Press Information Bureau, Government of India, on eSanjeevani (February 13, 2026): 282 million free teleconsultations from April 2023 to November 2025, 12 million aided by AI-recommended diagnoses.
77. Google and Boston Consulting Group (2023). A study reported by Euronews (February 2026) found that many such climate claims lack evidence.
78. Geoffrey Hinton, interview on BBC Radio 4’s Today programme (December 2024), as reported by The Guardian (December 27, 2024).
79. Eliezer Yudkowsky and Nate Soares, If Anyone Builds It, Everyone Dies (Little, Brown, September 16, 2025).
Trust us to get your leaders to be at their best!
Comments