What We’re Reading (Week Ending 05 October 2025) - 01 Oct 2025
Reading helps us learn about the world and it is a really important aspect of investing. The late Charlie Munger even went so far as to say that “I don’t think you can get to be a really good investor over a broad range without doing a massive amount of reading.” We (the co-founders of Compounder Fund) read widely across a range of topics, including investing, business, technology, and the world in general. We want to regularly share the best articles we’ve come across recently. Here they are (for the week ending 05 October 2025):
1. Nature’s cruel lesson for bag holders – Thomas Chua
On the screen, a cheetah was stalking a gazelle through tall grass, and I found myself holding the stretch longer, mesmerized. My body frozen in tension—part from the stretch, part from watching this life-or-death chess match.
Then the gazelle’s head shot up. The cheetah immediately stopped. No hesitation. Just turned and walked away.
As I eased out of the stretch and settled onto the mat, I realized something: successful investors act like that cheetah, but most people? They do the exact opposite.
Here’s what fascinates me about predators: they never chase a losing cause. The moment their cover is blown—the second the distance starts widening—they quit. No ego. No “sunk cost” thinking. No “but I’ve already come this far.”
They just walk away.
But when you look at the markets, you see the opposite everywhere…
…The thing about us humans is that our prefrontal cortex—the part that makes us “smarter” than animals—actually makes us worse at knowing when to quit. We tell ourselves stories. We rationalize. We create elaborate justifications for why this time is different.
But the cheetah? It just knows: energy spent on a failed hunt is energy that can’t be used on the next opportunity.
This isn’t about giving up easily. It’s about recognizing when persistence becomes stupidity. When determination becomes delusion.
The sunk cost fallacy tricks you into thinking backward—valuing what you’ve already invested over what you could gain elsewhere. But successful investors, like predators, always look forward.
They ask: “From this moment, right now, is this my best opportunity?”
If the answer is no, they walk away. No drama. No hesitation.
Like nature’s hunters, the smartest investors know exactly when to abandon the hunt. They don’t chase dead ends. They stalk fresh opportunities.
2. America’s top companies keep talking about AI – but can’t explain the upsides – Melissa Heikkilä, Chris Cook and Clara Murray
The FT has used AI tools to identify these mentions of the technology in US Securities and Exchange Commission 10-K filings and earnings transcripts, then to categorise each mention. The results were then checked and analysed to help draw a nuanced picture about what companies were saying to different audiences about the technology.
SEC filings require companies to disclose risks to the businesses, and are necessarily more cautious than the sales pitches made by executives on earnings calls. But the increasing array of risks described in filings appears not to be weighing on executives in the public pronouncements.
374 of the S&P 500 mentioned AI on earnings calls in the past 12 months — with 87 per cent of the calls logged as wholly positive about the technology with no concerns expressed…
…The FT sought to categorise the expected positive benefits of the technology. Most of the anticipated benefits, such as increased productivity, were vaguely stated and harder to categorise than the risks.Companies anticipated being able to optimise workflows through automation, and hope to achieve market differentiation through their use of AI. Some hoped to be able to use the technology to improve the personalisation of their products.
Filings do reveal that the companies able to give clear AI upsides include those that serve the rising AI-driven data centre boom. Energy companies First Solar and Entergy cited AI as a demand driver.
Freeport-McMoran, which has a stockpile of copper, stated that “data centres and artificial intelligence developments” would support the metal’s price. The company also said the technology can help with material characterisation and mineral extraction.
Equipment manufacturer Caterpillar reported that its energy business was benefiting from supporting “data centre growth related to cloud computing and generative artificial intelligence”…
…As the number of companies discussing AI has grown, fewer businesses are expressing positive views about the technology than they did in 2022.
The most commonly cited concern was cyber security, which was mentioned as a risk by more than half of the S&P 500 in 2024.
3. The 100 faces of China’s retiree wallet – Nina Chen
The heterogeneity within generations is not unique to China, nor is it exclusive to the elderly. But what sets China’s current seniors apart is the unprecedented structural complexity born of rapid social upheaval and institutional ruptures. This has made them the most heterogeneous and fragmented generation in the country’s modern history.
A closer look shows that different age cohorts in China were shaped by profoundly different circumstances. Take the post-1950s cohort: as children, they endured the Great Famine, carrying lasting memories of hunger through their formative years. In adolescence, their schooling was interrupted by the Cultural Revolution. At an age when they should have been applying their knowledge and skills, they were sent to the countryside for “re-education” through manual labor. Many missed out on the economic dividends of China’s later boom simply because they lacked access to formal education.By contrast, the post-1960s cohort came of age in a very different world. They benefited from the reinstatement of the college entrance exam and the rapid expansion of education. Their youth coincided with the early years of reform and opening, full of energy and opportunities. By middle age, they had experienced soaring property prices, volatile stock markets, the rise of the internet, and widening wealth gaps. The difference between these two generations is not one of degree, but of kind—a qualitative rupture rather than a quantitative stretch…
…In many reports and media narratives, seniors are depicted as embracing new trends and eager to spend: the spotlight often falls on smiling tourists, silver-haired influencers in stylish outfits, and the curated lifestyles of upscale retirement communities. Such portrayals carry strong visual appeal but obscure the underlying consumption attitudes of the majority.
For most seniors, the guiding principle is frugality: “money must be spent where it matters.” Family resources are first directed toward projects deemed vital and long-term—children’s housing, weddings, or cars—seen both as investments in future security and as obligations rooted in traditional family responsibility. Frugality is regarded as a virtue, so spending always requires a clear justification. Tangible goods with lasting value are far more acceptable than abstract or experiential services, with subscription models in particular often dismissed as “non-essential.” Seniors are highly price-sensitive, prioritizing cost-effectiveness over brands or aesthetics…
…This generation lived through China’s abrupt transition from a production-oriented society to a consumer-driven one. Having grown up under scarcity and a planned economy, they later faced a sudden explosion of marketization and commercialization in adulthood. Without systematic guidance—whether from family or school—on new retail channels, advertising formats, rules, and risk awareness, most lack strong consumer judgment. Social isolation and emotional vulnerability further make them particularly susceptible to highly personalized, “caring” marketing.
This dynamic often leaves them swallowing losses in contradictory ways. Some engage in compensatory spending when finances allow, yet remain drawn to bargain hunting. They might skip a RMB 79.9 buffet but stockpile dozens of RMB 9.9 trinkets online. When warned about scams, they retort, “You just don’t understand.” They defend dubious products with, “It has a factory address.” And to children who question their “adopted sons and daughters” from livestreams, they reply, “At least they call me more than you do.” The cautionary phrase they once told their children—“Everything online is a scam”—has boomeranged back to them.
Seniors thus represent both an underserved market and a lucrative yet highly fragmented, information-poor, and emotionally fragile consumer base. Until high-quality elder-focused supply matures, low-cost, easily replicated, high-margin “elder exploitation” businesses will fill the gap. Often, all it takes is a livestreamer repeatedly calling viewers “grandpa” or “grandma” to trigger enthusiastic purchases…
…Within the senior consumer base, those with limited resources and capabilities make up a large share. Behind the silver economy narrative lies a stark truth: the majority of seniors remain low spenders, with consumption disproportionately shaped by a small, visible minority…
…Supplements: Over 80% of seniors do not take them. Among those who do, more than 80% spend less than RMB 3,000 annually. Although supplements are often viewed as the quintessential product for older consumers, a survey on the Living Conditions of Urban and Rural Seniors data shows otherwise: only 16.6% of seniors report using supplements. Of these, 60.6% spend less than RMB 1,000 per year, 24.0% spend RMB 1,000–2,999, 6.5% spend RMB 3,000–4,999, and just 8.9% spend over RMB 5,000…
…Elder care: More than 80% of seniors cannot afford standard retirement home costs. According to CEIC data, in 36 major cities, the average monthly fee for self-sufficient seniors exceeds RMB 2,600—including accommodation, meals, and basic care—and is even higher for semi-dependent or disabled residents. Survey data show that only 15.8% of seniors can afford RMB 3,000 or more per month, meaning over 80% remain priced out of institutional elder care…
…Over 80% of seniors did not travel in the past year. Among those who did, more than 80% spent less than RMB 5,000 annually. According to the 2021 Survey on the Living Conditions of Urban and Rural Seniors, only 9.1% traveled in 2020, and even with some growth in recent years, the share is still estimated at under 20%. Among the small group who do travel, the majority spend under RMB 5,000 a year—pointing to a market still dominated by short, budget-friendly trips…
…First, draw the finest slice, not the biggest circle. Before a single yuan is spent, nail down exactly who you’re serving: how many grandmothers and grandfathers within a ten-year birth band, how much they can actually pay, and how often they’ll open their wallets. Over-count the grey tide and you’ll build a palace for a village—then watch inventory rot and margins drown.
Second, once you’re inside the right yard, seniors likely stay. Familiarity beats flashy ads; trust is a lifelong contract. Win them once and they’ll keep the same travel agency, the same pill brand, the same breakfast stall—year after year—turning your customer-acquisition cost into a one-time entry fee.
Third, profits may also come from the quietest voices—but only if you’re willing to do the hard, on-the-ground work. Village grandpas without apps and grandmas without data plans don’t appear on dashboards, but their needs are vast and competition is thin. Reaching them is hard: you must squat on the lane curb, piggy-back existing clinics, and price for a pocket that holds only folded bills. Yet thin-margin, high-frequency sales—subsidised just enough by local government—add up quickly when idle assets are put to work. I searched online and found some uplifting examples:
- In Ningxia’s Tongyi village, a derelict fish pond and drying yard were simply re-leased; anglers’ tickets and night-market stall rents now fully finance an 18-bed nursing home that charges residents zero fees.
- In Gutian, Fujian, a ¥3 lunch canteen covers its costs with a tiny on-site grocery counter plus monthly on-site pharmacy sales, and the same micro-format has already been copied in 48 neighbouring villages.
- In Caoxian, Shandong, 200 shared e-tricycles (¥1 per 3 km) pay themselves off in 18 months through side ads and parcel deliveries, proving that even a village road can become a revenue-producing asset.
4. AI isn’t replacing radiologists – Works in Progress and Deena Mousa
Radiology is a field optimized for human replacement, where digital inputs, pattern recognition tasks, and clear benchmarks predominate. In 2016, Geoffrey Hinton – computer scientist and Turing Award winner – declared that ‘people should stop training radiologists now’. If the most extreme predictions about the effect of AI on employment and wages were true, then radiology should be the canary in the coal mine.
But demand for human labor is higher than ever. In 2025, American diagnostic radiology residency programs offered a record 1,208 positions across all radiology specialties, a four percent increase from 2024, and the field’s vacancy rates are at all-time highs. In 2025, radiology was the second-highest-paid medical specialty in the country, with an average income of $520,000, over 48 percent higher than the average salary in 2015.
Three things explain this. First, while models beat humans on benchmarks, the standardized tests designed to measure AI performance, they struggle to replicate this performance in hospital conditions. Most tools can only diagnose abnormalities that are common in training data, and models often don’t work as well outside of their test conditions. Second, attempts to give models more tasks have run into legal hurdles: regulators and medical insurers so far are reluctant to approve or cover fully autonomous radiology models. Third, even when they do diagnose accurately, models replace only a small share of a radiologist’s job. Human radiologists spend a minority of their time on diagnostics and the majority on other activities, like talking to patients and fellow clinicians…
…Over the past decade, improvements in image interpretation have run far ahead of their diffusion. Hundreds of models can spot bleeds, nodules, and clots, yet AI is often limited to assistive use on a small subset of scans in any given practice. And despite predictions to the contrary, head counts and salaries have continued to rise. The promise of AI in radiology is overstated by benchmarks alone.
Multi‑task foundation models may widen coverage, and different training sets could blunt data gaps. But many hurdles cannot be removed with better models alone: the need to counsel the patient, shoulder malpractice risk, and receive accreditation from regulators. Each hurdle makes full substitution the expensive, risky option and human plus machine the default. Sharp increases in AI capabilities could certainly alter this dynamic, but it is a useful model for the first years of AI models that benchmark well at tasks associated with a particular career.
There are industries where conditions are different. Large platforms rely heavily on AI systems to triage or remove harmful or policy-violating content. At Facebook and Instagram, 94 percent and 98 percent of moderation decisions respectively are made by machines. But many of the more sophisticated knowledge jobs look more like radiology.
In many jobs, tasks are diverse, stakes are high, and demand is elastic. When this is the case, we should expect software to initially lead to more human work, not less. The lesson from a decade of radiology models is neither optimism about increased output nor dread about replacement. Models can lift productivity, but their implementation depends on behavior, institutions and incentives. For now, the paradox has held: the better the machines, the busier radiologists have become.
5. China’s AWS of Manufacturing – Thomas Chua
Guangzhou and its neighbors—Shenzhen, Dongguan, Foshan—form the Pearl River Delta manufacturing cluster. Decades of development have created an ecosystem so dense that suppliers, manufacturers, and assemblers for almost any product sit within hours of each other.
I took this trip to visit some of these wholesalers and it’s amazing how much they can do.
Walking through the wholesale markets, I saw many products that retail in Singapore and on online platforms selling at a fraction of the price. Take compression boots, for example—under $200 here. A similar device with a different brand slapped on in a Singapore mall near my house? Around $1,000.
Five times the price. Similar product.
I’ve known friends who’ve come here with specifications for products, whether clothing retail or electronics, and they’re able to establish their products and start selling abroad very quickly. All without ever having to sink heavy investments into building a factory or dealing with hiring anyone to produce these items…
…The Laifen hair dryers in hotels across China cost around $50. Dyson? $600.
The performance difference? About as noticeable as the taste difference between Pepsi and Coke.
We’ve also seen DJI and Insta360 run laps around GoPro in their offerings. If businesses can’t innovate fast enough, they’re going to be left behind.
This changing landscape created lots of new value, with some accruing to these newer, more nimble businesses. Consumers capture a nice chunk of the value as competition intensifies…
…On the first day, I had to use DeepSeek for my daily tasks—research, responding to my tour guide in mandarin, planning.
I’d tested DeepSeek during its Sputnik moment in January 2025 and found it comparable to ChatGPT. But now, having to use it due to the firewall, I realized just how rapidly Claude and ChatGPT have advanced. These models improve incrementally day by day—you don’t notice until you’re forced to switch between them.
The pace of AI development is staggering.
I ended up getting LetsVPN to access Claude and ChatGPT again—reliable for short China trips if you need Western services…
…During my daily hour at cafes, coffee in hand, doing my reading, I noticed something.
Many people around me were perpetually on LLM tools. Not just occasionally checking—constantly working with them.
DeepSeek and ChatGPT being the two most common.
Disclaimer: None of the information or analysis presented is intended to form the basis for any offer or recommendation. We currently have a vested interest in Amazon (parent of AWS) and Meta Platforms (parent of Facebook and Instagram) Holdings are subject to change at any time.
