Books - Q4 2025

• Born to Be Wired: Lessons from a Lifetime Transforming Television, Wiring America for the Internet, and Growing Formula One, Discovery, Sirius XM, and the Atlanta Braves (3/5) Everyone is talking about this autobiography by TMT entrepreneur John Malone. The WSJ review puts it well: "revelatory personal history and an exhaustive—sometimes exhausting—exploration into the intricacies of Mr. Malone’s various business transactions and relationships." John Malone was very early to the idea that you should focus on cash flow generation rather than net income. (Some say that he invented the EBITDA metric!) Cable companies were perfect for this because the capex to hang the cables and such could be depreciated over a much shorter period than their useful lives. He also liked to use as much debt financing as possible because interest is tax deductible. He also had a flywheel concept (like Bezos): "More scale equals more savings, which gave us more buying power to buy more systems and build more scale, which equaled more savings - a virtuous growth cycle." Byrne Hobart quips, "there are good times, and there are bad times, but there are no times when AT&T is not making poor capital allocation decisions." Something interesting about Malone - he was a big believer that you should partner with competitors in joint ventures, rather than bidding against them (e.g. for acquisitions). Split the surplus with your competitor rather than giving it all away to the seller in a competitive auction.
• Rethinking the Economics of Land and Housing (3/5) Always good to be rethinking things. We posted a link earlier to a Georgist economist who had a conspiracy theory that neoclassical economics was designed and promoted by landowners and their hired economists to divert attention from the way that landowners capture an unfair share of society's increasing wealth. "Once land is owned privately, much of the additional value created by society is captured by landowners, even if they have done nothing to earn that value." The debate between neoclassicist and Georgist economists comes down to whether land is just capital or a distinctive factor of production. I would bet on John Bates Clark. Points out: "physical space is also highly desired and not subject to diminishing returns - as people get richer, they want more space, " and "as economies mature the demand for land relative to other consumer goods increases," because it is a positional good. That is good for land investments if cornucopianism is right. A little bit of heterodox economics goes a long way, but if the Georgists were hopping mad about the benefits of owning land, that tells us that land is a bottleneck that we should be owning. The trick would be to find more attractively priced kinds of land, and note that natural resources are "land" in Georgist thinking. 
• The Scaling Era: An Oral History of AI, 2019–2025 (4/5) This gives really good insight into what the tech nerds who are going to spend a trillion dollars on data centers full of GPUs ("compute") are thinking. (Fair warning: it is written in Redditor nerdspeak full of jargon, and it can be painful to read.) A great point by François Chollet who is a skeptic of LLM intelligence: "If you look at LLMs closely, it's pretty obvious that they're not really synthesizing new programs on the fly to solve the task that they're faced with. Instead, they're reapplying things that they've stored in memory. For instance, LLMs can solve a Caesar cypher, transposing letters to code a message. That's a very complex algorithm, but it comes up quite a bit on the internet. They've basically memorized it. What's really interesting is that they can do it for a transposition length of three or five, because those are very common numbers in examples provided on the internet. If you try to do it with an arbitrary number like nine, it's going to fail. It does not encode the generalized form of the algorithm but only specific cases." Leopold Aschenbrenner is one of the Capex Maniacs: "The next model doesn't just require code, it involves building a giant new cluster. It involves building giant new power plants. Since ChatGPT, this extraordinary techno-capital acceleration has been set into motion." "By 2030, you get the trillion-dollar cluster using 100 GW - over 20 percent of US electricity production. That's 100 million H100-equivalents." There is an essay called The Bitter Lesson by Richard Sutton: "[G]eneral methods that leverage computation are ultimately the most effective, and by a large margin." "AI researchers have often tried to build knowledge into their agents; this always helps in the short run and is personally satisfying to the researcher; but in the long run it plateaus and even inhibits further progress; and breakthrough progress eventually arrives by an opposing approach based on scaling computation by search and learning." The author believes in the Gods of Straight Lines: "The GPUs are getting steadily better. The training runs are using steadily more compute. All the lines on all the graphs are perfectly, eerily straight (on a log-log scale!) and can be extrapolated with ease. Next year is simply this year, plus a known rate of change multiplied by delta-t. Scale is all you need."
• The Third Kind of Knowledge: Memoirs & Selected Writings (3/5) This is a collection of writing and memoirs by Robert Fitzgerald (1910-1985) who made probably the best translation of the Odyssey. As we saw with Paul Fussell, it is hard to understand what a big deal poetry and literary criticism were for men of this generation. An important corollary there is that there was no demand for poets during the Great Depression and it made this generation of Ivy League would-be poets into communist sympathizers. (Fitzgerald graduated from Harvard at the exact low of the Great Depression and had to work for a second rate newspaper in New York and then for Fortune magazine. He writes of Spain as "a legitimate Republic [that] had been attacked by a military and Fascist uprising.") Fussell, who almost died in WWII, said that Harvard men got to be "admirals' aides" in the war, and that is exactly what Fitzgerald, who went to Harvard, was! Something else that's "older generation coded" is being married multiple times. Fitzgerald had three wives. Other highlights: "General Marshall at Princeton some years ago said that no young man should think himself educated until he had read Thucydides - because, I take it, nowhere better than in Thucydides can you learn the best and worst to be expected from war and politics." "The Odyssey is about a man who cared for his wife and wanted to rejoin her. In the resonance of this affection, and by way of setting it off, the poem touches on a vast diversity of relationships between men and women: love maternal and filial, love connubial and adulterous, seduction and concubinage, infatuation superhuman and human, chance encounters lyric and prosaic." "A word about 'translation'. The Odyssey, considered strictly as an aesthetic object, is to be appreciated only in Greek. It can no more be translated into English than rhododendron can be translated into dogwood. You must learn Greek if you want to experience Homer, just as you must go to the Acropolis and look at it if you want to experience the Parthenon." "Late in '44, I was assigned to CINCPAC at Pearl Harbor, and when that command moved to the Marianas, to Guam, I went along on staff, to do various menial jobs. From, say, February to October of that year, I had nothing to do when off-duty but to read. I took three books in my footlocker. One was the Oxford text of the works of Virgil, one was the Vulgate New Testament, and the other was a Latin dictionary... my first real exposure to the Aeneid was hand to hand, with nothing but a dictionary - no instructor, no scholarship, nothing but the text itself, and the choice, evening after evening, of doing that or going to the Officers' Club and getting smashed." A big reason that he translated the Odyssey was that he could work remotely: "the best place to get help with the household and the children was overseas." His expenses were paid for six years by a Doubleday book advance and a Guggenheim fellowship to do the translation. He was Catholic and he did have six children!
• Underground: Stories to Understand Past, Present, and Future of Mining (3/5) Something to think about - sodium carbonate (soda ash) is used to produce lithium carbonate from lithium brines. You would think that demand for NRP's Sisecam Wyoming soda ash should grow with increasing global lithium production. "China is very advanced in the battery supply chain, in all segments, and much better than the West." Was surprised that the author mentioned the documentary Empire of Dust! Highlights: "During a family vacation on the French Riviera, Adolf [Lundin] took his sons, Lukas (12 years old at the time) and Ian (10 years old) to a cafe. Adolf looked across the table at his sons and announced that it was time to make a decision. 'Which of you will be my mining engineer and who will be in charge of the oil? You have 10 minutes to decide.' Lukas opted for the former, and Ian for the latter." Mentions an essay by Vaclav Smil: Peak Oil: A Catastrophist Cult and Complex Realities. Author: "We will take the opportunity to argue forcefully that resources are not running out globally and that we will never run out of oil." Talks about Malthusian thinking and mentions Julian Simon but not Herman Kahn or "cornucopianism." Mentions Alphamoon! (Tin producer in the DRC. Africa is rich with minerals.) Cecil Rhodes' Confession of Faith: "I contend that we are the finest race in the world and that the more of the world we inhabit the better it is for the human race."