Thursday Night Links

• Thank you. It's great to be here again and to see so many familiar faces. I've been looking forward to this weekend—what better place to mark my 100th day as Chairman? For the fine hospitality, everyone here is in debt to President Jeff Schmid and his colleagues at the Federal Reserve Bank of Kansas City. Jeff, our thanks to you all. Jeff and the other planners have some recreation options lined up for later today. And I'd advise you to be very careful with your choices. As I learned years ago, you can take two different kinds of hikes on the trails around Jackson Hole. I can sum up my hikes with former Vice Chairman Don Kohn in two words: I survived. These steely marathon death marches revealed a side of Don I wasn't ready for. There's another kind of hike—one I associate with Chairman Ben Bernanke, my old colleague. With Ben, it's a much more leisurely pace, an easy stroll along the wandering trails at the Rockefeller Preserve. [Warsh]
• The cement plant, then known as River Cement Company, took advantage of the vast reserves of Kimmswick and Plattin limestone that follows the banks of the Mississippi River.  These formations of limestone have been referred as the “White Cliffs of Selma” in writings of Mark Twain.  Also, in Mark Twain’s “Innocents Abroad,” he compared the bluffs to the Pyramid of Cheops. [Buzzi Unicem USA]
• Democratizations lower stock valuations and raise risk premia substantially across several proxies in data covering 90 countries over 200 years. These results cannot be explained by increased macroeconomic risk nor do other periods of high political or regime transition risk have the same effect. Exogenous variation coming from a change in Catholic church doctrine confirms that risk premia rise with the probability of a successful democratization. Redistribution risk can explain these results. In the data, redistribution follows successful democratizations: the size of the public sector and measures of economic competition rise, and income inequality and measures of corruption fall. Moreover, democratizations with higher redistribution risk see a substantially larger rise in risk premia than other democratizations. A redistribution-based model of democratic transitions with asset prices and incomplete markets can fully explain the results. [Max Miller]
• The acquisition will be funded through a $9 billion nonvoting minority equity investment in ONEOK's existing business. The investment carries an internal rate of return (IRR) that is capped at 7.0% for the first nine years of the investment, which is lower than ONEOK's cost of publicly traded equity. Distributions in excess of the capped IRR will reduce the minority equity capital balance over time, which increases the economic value attributable to ONEOK common shareholders. In addition to funding the acquisition, ONEOK intends to extinguish approximately $5 billion of existing indebtedness, immediately reducing expected pro forma 2027 leverage to approximately 3.25 times debt-to-EBITDA. [Oneok, Inc.] 
• As we discussed last quarter, we have seen increased interest from data center developers here recently. And I'd say that interest has been pretty varied. It ranges from developers looking to purchase maybe a few hundred acres for the actual data center footprint, other developers looking for several thousand acres for data center co-located power and perhaps even a buffer zone to address community concerns. But needless to say, the site requirements and the due diligence for data center development are even more stringent than what we see for solar development. So these types of opportunities are invariably going to take longer to materialize. With that said, we're certainly focused on identifying these types of opportunities within our land base and really capturing some of that momentum that we see in data center development. We put together a cross-functional team internally to identify what we think are higher potential sites within the portfolio. We've also listed some outside experts to help us with identifying those opportunities as well as marketing the potential side. So optimistic that we'll see this type of use contribute to our portfolio of HBU opportunities. But still a little early to speculate on just orders of magnitude in terms of pricing or number of acres that might ultimately be sold into that use. [Rayonier Inc.]
• Shareholders usually do not care so much about a particular company since they have invested widely across the economy. Instead of dying company A making a bold bet and potentially collapsing, much better for company A to just keep plodding along giving good but not amazing returns while they simultaneously invest in company B which is new and innovative and doing all sorts of crazy risky things. The shareholders can diversify across these risks while the management and leadership of company A cannot. This idea is the core of Clayton Christensen’s disruption theory. That incumbents have many advantages of scale, size, resources, etc, but their actual existing successes often become hindrances in the long run because they also entail a lock-in to the current way of doing things. The opportunity costs for doing anything else are just too high. And so, in a nonstationary environment, eventually there open up opportunities for newer firms, without high opportunity costs, to seize these opportunities, grow, and become the incumbents in their turn. These opportunities often (but not always) emerge at the lower end of the market, which the incumbents do not compete in as fiercely because the opportunity costs of further maximizing their existing business vs undercutting themselves to cater to the lower end are too high. [Beren Millidge]
• Poor is the person without Slack. Lack of Slack compounds and traps. Slack means margin for error. You can relax. Slack allows pursuing opportunities. You can explore. You can trade. Slack prevents desperation. You can avoid bad trades and wait for better spots. You can be efficient. Slack permits planning for the long term. You can invest. Slack enables doing things for your own amusement. You can play games. You can have fun. Slack enables doing the right thing. Stand by your friends. Reward the worthy. Punish the wicked. You can have a code. Slack presents things as they are without concern for how things look or what others think. You can be honest. You can do some of these things, and choose not to do others. Because you don’t have to. Only with slack can one be a righteous dude. Slack is life. [Zvi]
• On a positive note it is likely the case that the lives of many former aristocrat descendants are not worse than they would have been in the counterfactual with no Industrial Revolution. While their relative power has massively declined, they now have access to modern amenities, modern medicine, much more accessible travel, and in general it is likely just better to be a minor rich person today with a family house out in the country than a powerful lord several hundred years ago. The positive case for the singularity would look something analogous to this — as a human your economic power relative to other agents would be vastly diminished compared to today, but at the same time you would have access to many amenities unimaginable to billionaires today — such as a post-scarcity world for almost any good, biological immortality and digital backups, the ability to comprehend and understand the universe at a much deeper level, access to truly immersive VR simulations, and the possibility of interstellar travel. [Beren Millidge]
• We get clean and predictable scaling laws when we have removed all bottlenecks that would prevent further scaling. Once all bottlenecks are removed, then the generality of neural networks provides us a direct and predictable way to convert compute into amortized compressions of increasingly rare features of the training dataset. The hard part, of course – and the reason why figuring out ‘scaling’ took so long historically – is that removing enough bottlenecks to enter this ‘scaling regime’ is hard. [Beren Millidge]
• This latter point is the core argument of the bitter lesson. At small scales we do not have the capacity to properly model the entire distribution, so using a bias that encodes part of the structure ‘by hand’ can speed up training, increase sample efficiency, and generally improve performance. However, no real-world data distribution perfectly matches our hand-coded biases, so in the long run with sufficient compute and data it is better to encode the most minimal prior that sufficiently covers the hypothesis space. [Beren Millidge]
• Surprising detail is a near universal property of getting up close and personal with reality. You can see this everywhere if you look. For example, you’ve probably had the experience of doing something for the first time, maybe growing vegetables or using a Haskell package for the first time, and being frustrated by how many annoying snags there were. Then you got more practice and then you told yourself ‘man, it was so simple all along, I don’t know why I had so much trouble’. We run into a fundamental property of the universe and mistake it for a personal failing. [John Salvatier]