What We’re Reading (Week Ending 10 August 2025) - 10 Aug 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 10 August 2025):
1. How we’re making data centers more flexible to benefit power grids – Michael Terrell
That’s why we’ve been working to bring flexible demand capabilities into our data center fleet, which enables us to shift or reduce power demand during certain hours or times of the year. These capabilities, often referred to as demand response, have several advantages, especially as we continue to see electricity growth in the US and elsewhere. It allows large electricity loads like data centers to be interconnected more quickly, helps reduce the need to build new transmission and power plants, and helps grid operators more effectively and efficiently manage power grids.
We’re pleased to report on our progress in the implementation of these capabilities, including two new utility agreements with Indiana Michigan Power (I&M) and Tennessee Valley Authority (TVA). These agreements represent the first time we’re delivering data center demand response by targeting machine learning (ML) workloads. This builds on our successful demonstration with Omaha Public Power District (OPPD) where we reduced the power demand associated with ML workloads during three grid events last year — paving the way for us to pursue opportunities at other locations…
…Advancing Google’s 24/7 carbon-free energy ambition requires a holistic approach, to both procure clean energy and support the grid through demand-side solutions. Flexible demand is an important piece of this portfolio — it can be deployed quickly, helping bridge the gap between short-term load growth and long-term clean energy solutions, and delivers immediate benefits.
The first data center demand response capabilities we developed involve shifting non-urgent compute tasks — like processing a YouTube video — during specific periods when the grid is strained. Through our ongoing partnerships with Centrica Energy and transmission system operator Elia in Belgium, and Taiwan Power Company in Taiwan, we’ve leveraged this capability to help grid operators maintain reliability during those periods of the year when demand is the highest.
As AI adoption accelerates, we see a significant opportunity to expand our demand response toolkit, develop capabilities specifically for ML workloads, and leverage them to manage large new energy loads. By including load flexibility in our overall energy plan, we can manage AI-driven growth even where power generation and transmission are constrained.
2. Why China is building the world’s largest hydropower station in Tibet – Amber Zhang
On July 19, 2025, Chinese Premier Li Qiang stood in the remote southeastern Tibetan city of Nyingchi and announced the official commencement of the Medog hydropower station—what he termed a “project of the century”…
…The mega project is a hydropower station on the lower reaches of the Yarlung Tsangpo River, a plan of such breathtaking scale that it redefines the very concept of a mega-project. With a projected investment of 1.2 trillion yuan (approximately $167-170 billion) and a planned annual electricity output of 300 billion kilowatt-hours (kWh), the facility is designed to generate nearly three times the power of the iconic Three Gorges Dam, China’s previous “project of the century”.
Its output alone would be enough to power the entire United Kingdom [*(2023 statistics)] and is equivalent to 20% of China’s total residential electricity consumption in 2024 [*], enough for 300 to 400 million people…
…The Yarlung Tsangpo project marks the boldest chapter yet. It is located in Medog County, a remote corner of southeastern Tibet, at a dramatic geographical feature known as the “Great Bend” of the Yarlung Tsangpo River. Here, after flowing eastward across the Tibetan Plateau, the river makes a hairpin turn around the sacred Mount Namcha Barwa and plunges south toward India. In just 50 kilometers (31 miles), the river drops between 2,000 and 2,350 meters (over 6,500 feet) [*], through the world’s deepest canyon—three times deeper than the Grand Canyon in the United States.
It is this staggering vertical drop that has long been viewed by engineers as the single most promising site for hydropower generation on Earth, with a water energy density estimated to be seven times that of the Three Gorges [*].
To exploit the potential of the “Great Bend,” China is employing the “run-of-the-river” design, or, figuratively speaking, the “cut-the-bend” approach. Instead of constructing a single massive dam with a vast reservoir—which would be impractical and even more hazardous in this terrain—the project will consist of a series of five smaller “cascade” hydropower stations. These dams will divert a portion of the river’s powerful flow into a network of four enormous tunnels, each stretching approximately 20 kilometers (12.5 miles) and bored directly through the Himalayan mountains [*].
This “run-of-the-river” approach, which utilizes advanced dam-less diversion technology, means water is not consumed but rather borrowed with a resource utilization rate of up to 85% (*). It plunges down these tunnels, gaining immense velocity, to spin turbines located at a much lower elevation at the bottom of the canyon. After generating power, the water is discharged back into the river just before it crosses the Line of Actual Control into India. This design allows China to harness the massive potential energy from the 2,000-meter elevation drop while minimizing the size of the required reservoirs…
…For instance, the engineering challenges require technical capabilities that China has only developed in recent decades.
Beyond basic infrastructure like building roads and bridges, two key technologies have made this project feasible:
The first is tunnel boring machines (TBM)—used to dig long tunnels through mountains, similar to how a pangolin burrows through soil. These machines were once monopolized by German manufacturers and were prohibitively expensive. But after China localized their production, they became widely available and cost-effective.
In the early planning stages of the Medog hydropower station, several construction proposals were considered. Now, only one remains viable: a “run-of-the-river” approach, which involves digging a tunnel over 30 kilometers long to connect both ends of the river’s U-shaped bend, using the more than 2,000-meter drop in elevation to generate electricity. Such an idea would have been unthinkable in the past—but with TBMs, it has become a realistic option.
In fact, there’s already a prototype for this kind of construction: the Jinping II Hydropower Station on the Yalong River. Its surrounding terrain is nearly identical to that of the Medog section. There, engineers cut through both ends of a similar U-shaped bend, building four water diversion tunnels—each 17 kilometers long. The station has been operational for six years and has proven stable.
With this prior experience as a foundation, taking on the challenge of the Himalayas no longer seems so daunting.
The second breakthrough is ultra-high-voltage (UHV) power transmission. Tibet is vast and sparsely populated, and local demand is far below the project’s potential output. Most of the electricity will need to be transmitted to major power-consuming provinces in the east—or exported to Southeast Asia. The only viable solution for such long-distance transmission is UHV, one of the few technologies in which China is globally recognized as a clear leader. Years of experience from the “West-to-East Power Transmission” program have proven that UHV is both mature and reliable…
…Also, the Yarlung Tsangpo Grand Canyon is one of the most inaccessible places on Earth. Until the completion of the Paizhen-Medog Highway in 2022, the area lacked reliable road access and a power supply, making the logistics of transporting millions of tonnes of materials like steel and an estimated 40 million tons of cement a monumental undertaking.
The resulting power output is staggering. The project is designed with an installed capacity of 60 to 70 gigawatts, producing 300 billion kWh of electricity annually. This is enough energy to meet the needs of nearly 300 million people, making it by far the most powerful hydroelectric facility on Earth.
3. The Imitation Game: Defending against AI’s Dark Side! – Aswath Damodaran
A few weeks ago, I started receiving a stream of message about an Instagram post that I was allegedly starring in, where after offering my views on Palantir’s valuation, I was soliciting investors to invest with me (or with an investment entity that had ties to me). I was not surprised, since I have lived with imitations for years, but I was bemused, since I don’t have an Instagram account and have not posted on Facebook more than once or twice in a decade. In the last few days, those warnings have been joined by others, who have noted that there is now a video that looks and sounds like me, adding to the sales pitch with promises of super-normal returns if they reach out, and presumably send their money in. (Please don’t go looking for these scams online, since the very act of clicking on them can expose you to their reach.)…
…To get a measure of what the current AI scams that are making the rounds get right and wrong, I did take the time to take a closer look at both the Instagram post and the fake video that are making the rounds….
…The good news is that this AI scam gets my language and look right, but it is sloppily done in terms of content and capturing who I am as a person. The bad news is that it if this scammer was less lazy and more willing to put in some work, even with the current state of AI, it would have been easy to bring up the grades on content and message. I will wager that the Damodaran Bot that I mentioned earlier on in this post that is being developed at NYU Stern would have created a post that would have been much more difficult for you to detect as fake, making it a Frankenstein monster perhaps in the making. The worse news is that AI technology is evolving, and it will get better on every one of these fronts at imitating others, and you should prepare yourself for a deluge of investment scams…
…It remains an uncomfortable truth that the people most exposed to these scams are the ones who have read little or none of what I have written, and I wish there were a way that I could pass on the following suggestions on how they can protect themselves against the other fakes and scams that will undoubtedly be directed at them.
1. “Looks & sounds like” not good enough: Having seen the flood of fake AI videos in the news and on social media, I hope that you have concluded that “looks and sounds Iike” is no longer good enough to meet the authenticity test. This remains AI’s strongest suit, especially in the hands of the garden variety scammer, and you should prepare yourself for more fake videos, with political figures, investing luminaries and experts targeted.
2. Steer away from arrogance & hype: I have always been skeptical of the notion that there is “smart” money, composed of investors who know more than the rest of us and are able to beat the market consistently, and for long periods. For the most part, when you see a group of investors (hedge funds, private equity) beating the market, luck is more of a contributor as skill, and success is fleeting. In a talk on the topic, I argued that investors should steer away from arrogance and bombast, and towards humility, when it comes to who they trust with their money, and that applies in spades in the world of AI scams. Since most scammers don’t understand the subtlety of this idea, screening investment sales pitches for outlandish claims alone will eliminate most scams.
3. Do your homework: If you decide to invest with someone, based upon a virtual meet or sales pitch, you should do your homework and that goes well beyond asking for their track records in terms of performance. In my class on investment philosophies, I talk about how great investors through the ages have had very different views of markets and ways of making money, but each one has had an investment philosophy that is unique, consistent and well thought through. It is malpractice to invest with anyone, no matter what their reputation for earning high returns, without understanding that person’s investment philosophy, and this understanding will also give you a template for spotting fakes using that person’s name.
4. Avoid ROMO & FOMO: In my investing classes, I talk about the damage that ROMO (regret over missing out) and FOMO (fear of missing out) can do to investor psyches and portfolio.
- With ROMO (regret over missing out), where you look back in time and regret not buying Facebook at its IPO price in 2012 or selling your bitcoin in November 2013, when it hit $1000, you expose yourself to two emotions. The first is jealousy, especially at those who did buy Facebook at its IPO or have held on to their bitcoin to see its price hit six digits. The second is that you start buying into conspiracy theories, where you convince yourself that these winners (at least in the rear view mirror) were able to win, because the game was fixed in their favor. Both make you susceptible to chasing after past winners, and easy prey for vendors of conspiracies.
- With FOMO (fear of missing out), your overwhelming concern is that you will miss the next big multi-bagger, an investment that will increase five or ten fold over the next year or two. The emotion that is triggered is greed, leading you to overreach in your investing, cycling through your investments, as most of them fall short of your unrealistic expectations, and searching for the next “big thing”, making you susceptible to anyone offering a pathway to get there.
Much as we think of scammers as the criminals and the scammed as the victims, the truth is that scams are more akin to tangos, where each side needs the other. The scammer’s techniques work because they trigger the emotions (fear, greed) of the scammed, to respond, and AI will only make this easier to do. Looking to regulators or the government to protection will do little more than offer false comfort, and the best defense is “caveat emptor” or “buyer beware”.
4. How has macroeconomic research misjudged China? – Robert Wu and Dongfan Ma
From 2000 to now, China’s economic structure has undergone at least three major transitions:
- 2000–2010: Export and processing-led growth was the dominant force.
- 2010–2020: Real estate and household leverage became the drivers, with Total Social Financing (TSF) as the key indicator.
- Post-2020: Traditional models started to fail, and new growth drivers began to emerge.
Yet, most macroeconomic analyses remain stuck in phase two—still using TSF and real estate sales as core references. What we observe now is that these indicators have lost predictive value. Their correlation with PMI and corporate earnings is quickly fading…
…Processing trade began tapering off after 2010. But from 2016 onward, general trade has grown steadily and rapidly.
This distinction matters: processing trade mostly reflects contract manufacturing for others, while general trade signals the rise of China’s own manufacturing capabilities, brands, and integrated industrial chains. In 2024, China’s total export volume was already three times the size of real estate investment. Back in 2019, the two were roughly equal.
In other words, China’s new economic engine is no longer real estate, nor low-end contract exports—but rather the international expansion of Chinese brands…
…We’ve studied many outbound brands—MINISO, Pop Mart, Xiaomi, innovative pharmaceuticals, EV makers, short-form video platforms, games—and what we see is not a fleeting opportunity but a fundamental shift. It reflects the rise of talent, capabilities, and global competitiveness.
The era of debt-fueled growth is over. Today, growth comes from improvements in corporate strength, from truly competitive products, and from globalized operations…
…Still worried about China’s government debt? Concerned that the debt expansion of 2010–2020 is no longer sustainable? TSF growth slowing down? PPI still falling? Property prices not yet bottomed? Premium liquor sales still sluggish?
We have data and research to show that many former economic pillars and core assets are undergoing a transition. These variables are no longer fatal risks to the Chinese economy or its markets.
5. Wall Street’s Big, Bad Idea for Your 401(k) – Jason Zweig
Money managers are in a desperate race to stuff illiquid, so-called private-market assets into funds anyone can buy, including your 401(k). They say we all can earn high return and low risk with nontraded “alternatives” like private equity, venture capital and private real estate…
…Bluerock Total Income+ Real Estate is an “interval fund.” This is a structure that generally allows investors to buy as many shares as they wish at any time—but only to sell limited amounts at predetermined intervals, typically 5% of shares per quarter…
…Because private assets don’t trade, it’s the fund managers—not the market—that determine what they’re worth. That enables the managers to report much fewer and lower fluctuations than public funds do. Then they get to declare that private funds are low risk.
That’s ridiculous. In the real world, risk is the chance of losing money, which has nothing to do with how often prices are reported…
…Owning an alternative fund is a lot simpler than selling it. When you own it, you might take the manager’s valuations for granted, even if that’s a bad idea. When you sell it, the valuation matters—a lot. That’s a risk.
Until now, investors have been able to sell their shares back to each of these two funds at “net asset value,” or what the manager claims they’re worth. Even if other investors might disagree with some of those valuations, the manager has stood behind them.
That works until the number of people looking to sell swells and the managers can’t raise money because they are holding illiquid or distressed assets…
…The answer for these two funds, and for the alternative-asset industry writ large, is to move assets to public markets. There, the price will be set by what other investors—not the managers—believe the assets are worth. If that’s less than NAV, that’s mainly a problem for investors in the fund, not the managers…
…Most of the Bluerock fund’s holdings are stakes in other private real-estate portfolios. If it lists and ends up trading at a discount to net asset value, that might signal that the public market doesn’t believe the private valuations on dozens of these funds.
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 Alphabet (the parent company of Google). Holdings are subject to change at any time.
