In the early 20th century, a hefty Sears mail-order catalog was delivered to rural American households. Farmers sat at the dining table flipping through it, finding clothes, watches, sewing machines, farm tools, furniture, pianos, and even complete house kits that could be shipped by railroad. Those living in remote small towns could finally buy goods from across the country at prices close to those in the city. The prices in the catalog were open and transparent, making it difficult for local stores to continue arbitrarily raising prices based on information asymmetry. Sears once solved the toughest problem in American retail: how to deliver a vast selection of goods to consumers scattered across a vast territory. For nearly a century afterward, it evolved from a mail-order company into the city, then followed cars and suburban populations into shopping centers. By the 1970s, Sears' scale once approached 1% of the U.S. economy, and its headquarters, the Sears Tower, became the tallest building in the world at the time. In October 2018, this symbol of American retail filed for bankruptcy protection. At that time, the company had fewer than 700 Sears and Kmart stores left, with about 68,000 employees. More than a hundred years ago, it bypassed traditional stores relying on catalogs and railroads; more than a hundred years later, Amazon used websites, warehouses, and delivery networks to do what Sears knew best all over again. 1. Railroads, postal service, and industrialization turned America into a market. In 1886, railroad freight agent Richard Sears began selling watches by mail. At that time, American railroads were rapidly extending westward, manufacturing was mass-producing clothing, tools, and household goods, but rural areas lacked

On August 28, 2026, the French quantum computing company Pasqal debuted on Nasdaq. The stock price surged as much as 73% intraday and still closed up about 40%. The company raised approximately $360 million through a merger, with a listing valuation of about $2 billion. Pasqal's revenue in 2025 was only €16.5 million, currently deploying seven quantum computers, with factories in France and Canada producing up to about 13 units annually. Limited revenue and immature technology yet a $2 billion valuation indicate that capital is buying into a distant future in advance. Meanwhile, IBM plans to invest over $10 billion in quantum computing over the next five years; the U.S. government is preparing to invest about $2 billion in nine quantum-related companies and acquire partial equity; the UK announced a quantum plan worth up to £2 billion, with £1.2 billion allocated for purchasing large-scale quantum computers. Quantum computing has moved from university laboratories into government industrial policies, tech giant capital expenditures, and global security competition. From a physics experiment to a national-level industry: In 1981, physicist Richard Feynman proposed that classical computers find it difficult to effectively simulate the quantum world, and perhaps machines following quantum rules should be used to study it. In 1994, Peter Shor proposed a quantum algorithm that theoretically can quickly factor large integers. The security basis of some public key encryption used in modern networks is precisely that classical computers find this task difficult. Since then, the significance of quantum computing has expanded in two directions.

In the fall of 2001, Salesforce's bank account was nearly depleted before the next payday. The internet bubble had burst, many clients went bankrupt, and venture capital firms collectively tightened their checks. Salesforce was burning about $1 to $1.5 million per month, and the founding team visited investment firms one after another, but were almost all rejected. At that time, many people thought it was crazy to put enterprise customer data outside the company's firewall. Investors even suggested Salesforce remove ".com" from its name, abandon the cloud model, and revert to traditional software. Salesforce stuck to its original technical path but had to solve a more practical problem first: where the cash would come from. The key move that ultimately saved the company was surprisingly simple. It began encouraging customers to switch from monthly fees to annual fees, prepaying for 12 months of software in advance. Salesforce thus received cash upfront, and customers locked in services at a lower cost. This annual fee billing later became one of the most important cash flow structures in the SaaS industry. Twenty-five years later, Salesforce just announced that its AI and data products' annual recurring revenue is close to $4 billion, with its stock price rising 22.6% in a single day. From nearly being unable to pay salaries to enabling global enterprises to subscribe annually to AI agents, this company has redefined how software is charged and is changing the structure of white-collar work. Leaving Oracle, Benioff wanted to end traditional software. In 1999, Marc Benioff left Or

In June 2026, Canada's Deep Sky completed North America's first batch of independently verified direct air capture carbon removal deliveries, with buyers including Microsoft and the Royal Bank of Canada. Its Alberta pilot site has a designed annual capture capacity of about 3,000 tons, still very small in scale, but it accomplished an important milestone: companies finally obtained carbon removal products that can be verified, registered, and tracked for storage locations. In the same month, Frontier, involving companies such as Stripe, Google, Shopify, Salesforce, and Anthropic, added $915 million in long-term purchase commitments, bringing the total planned scale to $1.8 billion. Microsoft disclosed that in fiscal year 2025, it signed carbon removal agreements totaling about 45 million tons with 21 global suppliers. It is important to distinguish here: most of the 45 million tons belong to contracts for delivery over many future years and do not mean that this CO2 has already been removed from the atmosphere. The electricity demand driven by AI is pushing up emissions from tech companies, and cleaning up these emissions is beginning to form a new industry chain. From "emitting less" to cleaning up historical emissions Early carbon markets mainly traded emission reduction and avoidance projects, such as forest protection, improving energy efficiency, or replacing fossil fuels with renewable energy. This model has faced many controversies: whether projects really happened because of carbon credits, whether the same forest was double-counted, and whether it might be burned down by wildfires decades later—all could affect credit quality. Carbon removal

In early 1991, Oracle released a very strange report to the market. The company's revenue still grew by 29%, yet it recorded a quarterly loss of $6.7 million. Banks immediately cut Oracle's credit line from $170 million to $80 million, and the stock price fell about 80% from its peak. A few months earlier, Oracle had already laid off 400 employees, equivalent to 10% of its U.S. workforce. The board even discussed whether to fire founder Larry Ellison. This tech giant, which later supported databases for global banks, telecommunications, governments, and large enterprises, was then very close to bankruptcy. Thirty-five years later, Oracle is once again at the center of controversy in the capital markets. In fiscal year 2026, Oracle invested about $55.7 billion to build AI cloud infrastructure, with free cash flow turning negative $23.7 billion; meanwhile, cloud infrastructure revenue is growing rapidly, and unrecognized contractual obligations have reached $638 billion. What almost destroyed Oracle back then was treating future contracts as current revenue too early. Today, Oracle faces a new challenge: investing huge amounts of cash upfront to build data centers, then waiting for future AI orders to materialize. The database revolution made Oracle run too fast. In 1977, Larry Ellison, Bob Miner, and Ed Oates founded Software Development Laboratories in California, which later became Orac

Toys "R" Us, a happy childhood memory for so many! Including me... Every year before Christmas, American children would receive a thick toy catalog. Some would circle the toys they wanted, others would fold the pages, then place the catalog where their parents could easily see it. On weekends, walking into Toys "R" Us, rows of Barbie dolls, LEGO bricks, remote control cars, and video games stretched from the entrance deep into the store. For generations of Americans, Toys "R" Us was almost synonymous with childhood itself. It pioneered the large toy specialty store model, relying on a vast selection and a nationwide store network, pushing traditional department store toy counters into the corners. By 2018, this retail giant that had accompanied American consumers for decades decided to liquidate its U.S. business, with 735 large stores closing one after another, affecting about 30,000 jobs. Children still need toys, and the global toy market has not disappeared. What disappeared was an industry giant locked in debt, suburban megastores, and old retail rules. 1. It turned the toy counter into a kingdom. The story of Toys "R" Us began in 1948. After World War II, the United States entered the baby boom. Founder Charles Lazarus realized that durable goods like cribs and strollers could only be sold once to a family, but toys would generate new demand as children grew. He gradually reduced children's furniture and gave store space to toys. In 1957, the first store officially named Toys "R" Us opened in Maryland. It adopted a very novel approach at the time

On August 3, 2026, Mastercard officially completed the acquisition of stablecoin payment infrastructure company BVNK. The deal was valued at up to $1.8 billion. BVNK has the capability to connect fiat and stablecoins across more than 130 countries, along with the hardest-to-replicate local licenses and compliance channels. Two days later, Circle announced its Q2 results: USDC reserve income reached $668 million, a year-over-year increase of 5%; related costs such as distribution and trading amounted to $412 million. Meanwhile, Visa's stablecoin settlement pilot has expanded to nine blockchains, with an annualized settlement scale of about $7 billion. By August 25, new commercialization signals appeared: in July 2026, global stablecoin card spending surpassed $1 billion for the first time. Payment company RedotPay predicts this scale could reach $50 billion annually by 2028. This is a corporate forecast, and whether it will be realized still depends on users, merchants, and regulatory progress. Stablecoins are moving out of exchanges and into corporate finance, bank cards, payroll, cross-border trade, and everyday wallets. From crypto trading tools to global payment systems Stablecoins initially solved problems within the crypto market. Bitcoin's price volatility is too high, and traditional banks cannot process on-chain transactions around the clock. Traders needed a digital dollar that could hold funds and settle at any time. Stablecoins like USDT and USDC thus became the foundational assets between exchanges and blockchains. Subsequently, the market discovered they could also handle the most troublesome aspects of traditional finance