Understandary

•

Kanał • Programy: 2

Komentarze na temat aktualnych wydarzeń

Komentarze na temat aktualnych wydarzeń

Uakt. co tydzień
Społeczeństwo i kultura

Społeczeństwo i kultura

Uakt. co dwa tyg.

30 cze

2026 European Heat Wave

This week we talk about air conditioners, pressure systems, and heat stress. We also discuss weather memes, climate change, and dirty grids. Recommended Book: Battle of the Linguist Mages by Scotto Moore Transcript An air conditioner, or AC, or maybe air con if you’re in the UK, is a device that moves heat from one location to another. In doing so, it usually dehumidifies the air, as well, so it can rapidly cool a room or entire building by shifting both heat and humidity from that room or building, elsewhere—usually outside. This is basically the same technology used in refrigerators, a process called vapor compression allowing the device to circulate a substance called refrigerant using a compressor, a condenser, an evaporator, and an expansion valve, which—and this is a very superficial explanation of what’s happening—but these components take advantage of forced circulation and a phase-change between gas and liquid to transfer heat from the room you want to cool, or the inside of the refrigerator, and move that heat outside your building, or to the back and/or bottom of the fridge. This is a far more active mode of air conditioning, of cooling and dehumidifying the air, than has been used throughout history. Most early methods relied on passive approaches, including but not limited to architectural elements, the use of plants and optimization of air flow, or creating basement areas for things that needed to stay cool. Researchers have dabbled with more active methods of conditioning air for centuries, though, and several 19th century inventions served as precursors for the first iteration of modern ACs, some of which were used to create ice, which was useful unto itself, but could also be used to cool a room, if far less effectively and efficiently than an actual, holistic AC unit. In 1894, industrial-grade ammonia compressors, powered by electricity, made this category of device suitable for urban environments; previously they just were far too bulky and difficult to power for city use. By 1896, the Hungarian engineer who came up with this new riff on the theme, István Röck, was manufacturing what he called dry air cooling apparatuses for hospitals, theaters, and other large spaces. Just five years later, in 1901, an American inventor named Willis H Carrier developed what’s widely considered to be the first modern electrical AC unit, selling the first one to a lithography company in New York, before patenting the term air conditioning in 1906. The first residential version of this device was installed in 1914, and in 1915 the Carrier Air Conditioning Company of America was formed—a business that still exists today. The impact of air conditioning, and this general technology category, as again, it’s also used in modern refrigeration units, cannot be overstated. This tech didn’t become widespread in the US, which is where it initially took off, in large part due to Carrier and other AC businesses’ presence in the States, until the mid-20th century, and before that, before the 1950s, the state of Florida was technically occupied, but only just barely because of its extreme heat and humidity and abundance of mosquitos. The population of Florida in 1950 was about 2.7 million, and today it’s about 23.5 million—that influx of people began after AC units became standard in buildings across the state, and the country. We’ve seen similar migrations as a result of too-hot places sudden becoming a lot more pleasant. Similarly, refrigeration enabled a boggling amount of change within the food and beverage industry, the chemicals and industrial materials industries, and the healthcare and life science industries, because before the advent of the cold chain—the system of refrigerated spaces, including boxes and trucks and planes and ships that allowed medicines and foods and other substances to stay cold from their origin to their end-consumer—it simply wasn’t possible to sell or create or work with many of these products and materials. The distribution of this technology is not universal or equal, however, and in some cases that inequality, that lack of access to this technology in some spaces, is the result of choice, not inaccessibility. And that’s what I’d like to talk about today: the spread, or lack thereof, of AC technologies and products, and how a recent heat wave in Europe may lead to more installations of this type of product across the continent. — Beginning in late-May of 2026, a series of severe heatwaves engulfed Europe, and especially Western Europe, breaking all sorts of temperature records and leading to a bunch of heat-related deaths. A recent meme gives a good sense of just how bad this heat wave has been. Back in 2014, as part of a campaign by the World Meteorological Organization, dozens of weather presenters from around the world were invited to record fictionalized weather reports from 2050, with the intention of giving people a sense of how global climate change might impact daily life even as soon as just 36 years in the future. One of these presenters, from France, gave a report of an imagined scenario in which a heat wave descended upon Europe, showing temperatures of 40 degrees celsius, which is about 104 degrees Fahrenheit, with some areas seeing temperatures as high as 43 C or 109.4 degrees F. This was generally considered to be a baffling, maybe even sci-fi sort of prediction at the time. But this clip has resurfaced and widely shared, as, just 12 years later, not 36, this recent heat wave has not just met, but in some cases surpassed that imagined, too-crazy-to-be-real European heat wave scenario. Across much of Western Europe, those temperatures milestones were hit, and in a few locations they were beat by as much as 20 degrees F. For a few days in France, temperatures were higher than in Las Vegas, Nevada and Phoenix, Arizona, coming within 2 degrees F of temperatures in famously too-hot Death Valley, California. All-time high temperature records were broken in Germany, and a handful of other countries are waiting to see if provisionally recorded high-temperatures they experienced hold up, to see if their own records will be broken. The UK recorded a temperature of 37.1 C (98.78 F), the Netherlands saw a top temperature of 39.4 C (nearly 103 F), Belgium hit 40 C (104 F), and Germany recorded 41.3 C (106.3 F)—all temperatures that are 5-12 C above seasonable averages, and this has led to all manner of infrastructural issues, as well. A nuclear power plant in Switzerland had to take both reactors off the grid because the temperature of the river that cools it got too hot, a Eurostar train broke down east of Brussels with about 400 people on board, and 3 people had to be taken to the hospital. A bunch of big public events were cancelled, hospitals were filled to the point of having to triage visitors, and, if these on-the-ground issues weren’t enough, researchers in Switzerland have warned that almost all of the winter reserves built up on their glaciers are gone, which suggests the glaciers themselves will begin melting soon—something that usually doesn’t start happening until August. This heat wave was the result of a potent heat dome, which is a pocket of high pressure that developed as a result of jet stream fluctuations that pulled hot air north, out of warmer portions of Africa. This type of high pressure pocket can then cause the heat it gathers to just sit there, unmoving, whereas typically it would be pushed around, causing it to disperse, and to thus not have such a significant impact on people and other life on the ground. Unfortunately, these sorts of disruptions to the previous climatic norm are becoming more common as the climate shifts, due to the accumulation of CO2 and other heat-trapping gases in the atmosphere. The regulating systems we’ve evolved with are changing because there’s more overall energy, more heat, powering these systems across the planet, and that means we’re no longer able to predict them as accurately, but also that some of these systems could fundamentally change or disappear, possibly in the near future. Events like this, which are currently rare, then, could soon become common. And meteorologists in Europe have warned, during this major heat wave, that this sort of event could be a regular thing, and soon, and could even last a lot longer; not a matter of days or weeks, but possibly spanning months at a time, without reprieve. That’s not great news for a continent that, until now, has generally been pretty okay with its existing heat and weather infrastructure. Unlike the US and other countries that have some incredibly hot regions, leading to the widespread installation of AC units, only about 20% of homes across Europe have AC installed. As a result, as this heat wave descended on the area, trapping heat inside buildings that were constructed with cooler weather in mind, people have been unable to remain indoors, to work, to sleep; it’s been miserable. And in some cases, deadly; the combination of heat and humidity making it more difficult for peoples’ bodies to regulate heat via sweat, and that’s led to an increase in heat stress on their bodies—which is just a miserable thing to deal with, but it also means a lot of people, including but not limited to the very young and very old, are more likely to die, their bodies simply incapable of handling that level of persistent temperature strain. Back in the summer of 2022, during another, less intense heat wave, more than 60,000 people died across Europe due to heat stress and related ailments. The numbers are still out on this more recent heat wave, but the stats are expected to be pretty grim, as in addition to the individual strain people in these afflicted areas are suffering, infrastructure tends to collapse in unprepared areas, hospitals not functioning or not functioning well, their machines and IT systems failing due to the hea
7 lip

Digital Gaming

This week we talk about Sony, Nintendo, and the Playstation. We also discuss Grand Theft Auto, the 3DO, and digital dark ages. Recommended Book: 3 Days, 9 Months, 27 Years by John Scalzi Transcript The earliest video game consoles that were made to be used in the home, as opposed to being set up in an arcade, were hardwired like their arcade kin. That means rather than being able to play a bunch of different games, they were basically just single-game boxes: you would buy a machine that allowed you to play Pong, for instance, and if you wanted to play another game, even by the same maker, Atari, you would have to buy another whole console with its own screen, controls, etc, to do so. That was the state of the art in the early to mid 1970s. By the late-70s, the concept of swappable games became reality with the introduction of what are called ROM cartridges. ROM stands for read-only memory and is a type of storage common in computers and other devices, which allows whatever you store on it to persist, which is a contrast to RAM, which is the type of memory that determines how much you can do on a device at any given moment, and which disappears when the device is turned off. So these ROM cartridges were kind of like the portion of the hard drive that’s used to boot up your computer, storing the bare-basics of the system so it can be initialized and understand how to run all the other software that builds upon that baseline. And that memory was stored in durable, plastic cases that made them usable by ordinary, non-techy people. You could buy a game and handle the cartridge, popping it into your game console hardware and removing it, to make way for another game, over and over and over again, and that use would be unlikely to damage the ROM chip. This same general format was flexible enough that it lasted through the mid-90s, the capacity of the ROM chip continuing to grow as the associated tech improved, and the capabilities of the central console hardware that used these cartridges became more sophisticated. Upgrades were slowly added to the innards of the plastic case, as well, including things like battery backups that enabled saved games, and the Super Nintendo’s Super FX chip, which enabled 3D graphics that would have otherwise been impossible with the contemporary state of the art. The next generation of gaming consoles relied on another medium, though, and one that had several benefits over the long-lived game cartridge. CD-ROM discs, which were flat, circular, and contained information that was encoded and read with lasers, had been around in some form since the late-1980s, and were even used in a few early gaming consoles, like the PC Engine CD-ROM, which barely anyone bought, and the Sega-CD add-on for the Sega Genesis, and 3DO consoles, which a few more, but still relatively few people purchased. The release of the first Sony Playstation, now known as the PS1, in 1994 changed that, though, and this shift was partially the result of Sony’s impressive game lineup, but was also due to the strength of the CD medium. Each CD-ROM could hold 650-700 MB of information, which was more than 100-times the capacity of the competing Nintendo 64’s cartridges. There were downsides to this new standard; CD-ROMs were less durable than plastic-encased cartridges, and they were very slow to load, as well, because information stored in ROM chips could be more or less instantly booted, while the info stored on discs had to be spun up and read first, resulting in sluggish load screens throughout the gaming experience, and especially on the initial boot-up of the system. That said, the far superior storage, and the dramatically reduced cost of these laser-etched discs—cartridges could cost $15-20 apiece to manufacture, while CD-ROMs often cost pennies apiece—that triggered a rapid transition in the gaming world to this new medium. Handheld consoles stuck with cartridges for a lot longer, due to the nature of the use-case and difficulties associated with trying to use spinning discs in portable hardware, but everyone else moved to discs pretty rapidly, after Sony proved the utility of the model, and many aspects of video gaming were upgraded as a result of all that additional storage capacity. That capacity continued to grow as CD-ROM were replaced with DVDs, which could hold 4.7-8.5 GB per disc, again, up from 650-700 MB; the industry made that change in the years 2000 and 2001, with the PS2 and Xbox consoles. And then in 2006, the PS3 moved to Blu-ray discs, which could hold a whopping 25-50 GB per disc, once again resetting gaming expectations—though Xbox stuck with DVDs, and Nintendo’s Wii, Wii U, and Gamecube consoles used proprietary disc formats that had a lot lower capacity compared to their competition. Leading into the 2010s, even those Blu-rays were straining under the weight of some big-name, AAA games, some of which required multiple discs and mandatory hard drive installs from those discs, because the scope of these gaming worlds and their high-end graphics required just a stunning amount of space. Video game companies had already started making the shift to digital products in the early 2000s, though, Xbox Live Arcade and the Playstation Store emerging in 2005 and 2006 respectively, and Steam, which popped up in 2003, was making digital downloads for games common on PCs several years earlier. Digital became even more popular in the 20-teens, and in 2020, digital sales of console games surpassed physical sales for the first time. The PS5 and Xbox Series S shipped console versions without disc drives for the first time, and many physical games became basically methods of checking a game’s license, to ensure it’s not pirated, because the discs installed the game on the console’s hard drive, just like a download, anyway. What I’d like to talk about today is the perhaps natural next step in this transition: a recent announcement by Sony that they’ll no longer be making disc-based Playstation games beginning in 2028, and why some critics are calling this a worrying and anti-consumer move. — On July 1, 2026, Sony announced that’s it’s going all-in on digitally delivered games. It will halt production of game discs beginning in January of 2028, and after that, customers will only be able to purchase new games digitally, via their Playstation Store and retailers. This follows another recent announcement by game company Rockstar Games that their massively anticipated Grand Theft Auto VI game, which is set to hit digital shelves on November 19 of this year, will not be available on disc at all; it will be downloadable, and the physical copies customers can buy in stores won’t be physical copies at all: it will be a box with a download code inside, which amounts to the same thing—you use the code to download the game, exactly the same as if you had bought it online, but with packaging for that code. That’s a big deal because the Grand Theft Auto series is one of the biggest and most popular series in gaming history; other game makers have been scrambling to adjust their own release dates so as not to overlap with this new, highly anticipated game’s release; it’s expected to be an absolutely massive moment in the gaming world when this new entry in the series finally lands. That said, the writing has been on the wall for the transition to digital games for a while, now. Game company Capcom recently announced that 93% of its game sales were digital in its last fiscal year, and other companies have reported similar numbers; it’s currently around 85% for Sony. As a result, many consoles are now shipping models without disc drives, and some, like the recently announced Steam Machine, don’t even have a disc-drive version. Digital games are also cheaper to make because the company behind them only has to provide download keys, rather than having to pay some amount for each and every physical item produced, packaged, and shipped, and these companies, and the retailers that sell their games, will never run out of a popular game, which might otherwise be an issue for big released like the aforementioned Grand Theft Auto. There are quite a few downsides to the digitization of games, though, including that in many cases, you don’t actually own the games you buy, you just own a license to download and play them. That means if licensing changes, or the storefront through which you bought a game closes, you will likely lose access to that game you bought, without getting any kind of refund. If your account is banned or you lose access to your account for some reason, all your games will suddenly be inaccessible, too. If you don’t have access to the internet, allowing your console to phone home and check to make sure you’re not pirating things on a regular basis, that might also mean no gaming for you. Almost always, you can’t sell or trade digital games, while physical games allow for a thriving secondary market, often allowing gamers to buy old discs and cartridges decades after a game was released, and often for far lower prices. Many of these games can even be played by later consoles that have back-compatibility. Doing away with discs and other physical media is great for companies like Sony, then, because they no longer have to pay to create the individual discs, no longer have to pay for the drives that play the discs, so the price of making consoles drops a bit, and it also means that secondary market for games goes away: if you want to play old games, you have to buy them from Sony, and all those used games already on the secondary market, or a game disc borrowed from a friend, are no longer competition for them, serving as alternatives to the digital version they’d prefer to sell you. Another bigger-picture concern here, though, is that this will make game preservation efforts a lot more difficult, and for many of the same reasons it will make maintaining a library an
14 lip

NATO Summit 2026

This week we talk about Russia, Ukraine, and the Warsaw Pact. We also discuss Patriot interceptors, Hungary, and Article 5. Recommended Book: The Alternative by Nick Romeo Transcript The North Atlantic Treaty Organization, or NATO, was founded in the wake of WWII, in 1949, in order to affirm unity between non-Soviet European nations, Atlantic nations like Greenland and Iceland, and the US and Canada, over in North America. The main purpose of this unity was to establish a sort of firewall around the western perimeter of the Soviet Union, which at the time was the only global superpower other than the US, because much of the world, but especially Europe, was struggling to recover from the destruction wrought during the second world war, and the Soviets had made pretty clear that they intended to take over everything: they’d already gobbled up most of their neighbors, creating an increasingly expansive buffer zone of Soviet states around their central, Russian territory, and most of the conflicts still playing out, or threatening to play out, globally at this point were either overt or slightly concealed proxy fights between the capitalist democratic forces of the West and the authoritarian, Stalinist forces of the Eastern Soviet bloc. NATO was thus a wall of nations that said, hey, if you attack any of us, that will mean you’re attacking all of us. And that ‘all of us’ included the United States, which was the only individual force capable of standing up to the Soviets at this point, due to its massive conventional military force, and the threat posed by its huge, and still growing, nuclear weapons arsenal. The Soviet counter to NATO was called the Warsaw Pact, which formed in 1955, and these rival alliances carved up Europe during the latter half of the 20th century, until the Soviet Union collapsed in 1991. As military alliances go, NATO has been fairly successful—Article 5, the portion of the agreement that triggers if a NATO member is attacked, calling the other members to come to their aid, was only activated once, following the terrorist attacks on the US on September 11, 2001; that led to NATO involvement in the US’s attacks on Afghanistan in subsequent years, though NATO forces also periodically got involved in other regional conflicts, like the Kosovo War in 1999, and the Libyan Civil War in 2011, in both cases working with the UN to protect civilians from the actions of violent leaders or assailants. But beyond that, no one, including the Soviets, messed with NATO. NATO has since accepted sixteen new member states, and that expansion is one of the supposed rationales for Russian President Putin’s invasion of Ukraine. During Putin’s first presidency, Russia and NATO had been on pretty solid, even cooperative terms, which had some questioning the point of the treaty, since it was originally formed to counter Russian aggression. In Putin’s second presidency, though, things took an antagonistic turn, and when Russia illegally invaded and annexed a part of Ukraine, called Crimea, back in 2014, NATO ceased all cooperation with Russia. When Russian forces launched a full invasion of Ukraine in 2022, NATO suddenly seemed more important than ever, as a reinvigorated Russia, with the stated purpose of, in Putin’s words, reclaiming portions of Europe that were previously part of the extended Soviet Union, parts of the Warsaw Pact, that posed a serious threat to just about everyone, especially European nations that border Russia and Russian allies, like Belarus. What I’d like to talk about today is the 2026 meeting of NATO leaders, and the general state of affairs on the ground in Ukraine, as of mid-2026. — Leaders of NATO member nations typically meet for a summit each year, though the schedule varies a bit, depending on the needs of the moment, and whether there are any NATO-relevant crises that might nudge things forward or cause them to be delayed. The 2026 NATO summit was held in Turkey’s capitol city, Ankara, on July 7 and 8. It was the second such summit hosted by Turkey, and the 36th NATO summit, overall. This meeting was notable for several reasons, many of them directly related to Russia’s ongoing invasion of Ukraine, but also the US’s continued ambivalence, at times bordering on active antagonism, toward the treaty, under the Trump administration. Over the course of the past 4 years, Russia has continued to make threats toward the rest of Europe, implying or suggesting that it might have to act militarily against its NATO-member neighbors. During the same period, the US has criticized European NATO member states for not carrying their own weight, most of these nations not spending enough of their GDP on their military and defense infrastructure, in accordance with their treaty obligations, and most more or less relying on the US’s military (and nuclear) umbrella to threaten would-be attackers. This has long been the NATO state of affairs, but under the Trump administration, the US made it a point of contention. And though it’s been a relatively slow process—it’s not easy for a government to pivot toward that scale of remilitarization—these nations have agreed to a 5% of GDP target for defense and defense-related spending by 2035, and in the last year alone, European and Canadian defense spending has increase by 20%; not all of which is immediately convertible into useful, front-line assets and soldiers, but it does represent a significant change to the status quo, which could eventually, in less than a decade, result in a European front-line that’s more European and less NATO, for the first time since that immediate post-WWII period. This meeting also featured a reaffirmation of Article 5, which has always been there and in effect, but some analysis has questioned whether NATO allies, including the US, would actually step up if Russia were to attack the Baltics, for instance. This is a seemingly small move that serves to underline that stance of, if you attack one of us, you’ve attacked all of us, at a moment in which Russia seems to be toying with the idea of picking off pieces of the alliance, to see if they can get away with it, like they did with Crimea back in 2014. There were also announcements related to a fresh $50 billion in defense industry deals, NATO members investing heavily in US-made arms and new assets from elsewhere across the bloc, and a lot of that money is going to drone-war infrastructure and militarized AI models that they hope will prepare NATO for current and next-generation conflicts. Another $217 billion in additional financing commitments from banks across the bloc have also been announced. Ukraine scored some pretty big wins at this year’s NATO summit, getting a license from the US to build Patriot interceptors, the lack of which have resulted in big gaps in the country’s defense system, and about $80 billion in military equipment, assistance, and training for Ukrainian forces was committed by the bloc. There was also a verbal commitment that NATO and its allies “stand united in our unwavering support for Ukraine in defending its freedom, sovereignty, and territorial integrity.” Which isn’t nothing, at a moment in which the US is led by Trump, a person who has frequently sided with Putin in Ukraine-related matters. Ukraine has also signed some bilateral drone deals with NATO member states, reaffirming that Ukraine has become one of the most in-demand experts on this subject, at a moment in which it’s becoming clearer and clearer that low-cost, high-impact drones and similar technologies are likely to dominate battlefields for the foreseeable future. There were also some tensions, mostly sparked by Trump, over his administration’s desire to buy Greenland, which is not a notion that’s supported by anyone else in the bloc, and the lack of NATO support for the US’s war in Iran. Overall, though, there was a decent sense of unity and progress, and Ukraine, again, not a NATO member state, but a nation that hopes to eventually join NATO, received a warm welcome and a bunch of support, which hasn’t always been the case at recent summits. Part of that warmth is likely the consequence of Ukraine’s recent victories on and off the battlefield. Russian continues to pummel Ukrainian cities and kill Ukrainian soldiers and civilians en masse, but Ukraine has managed to keep them from gobbling up territory on scale, and in some recents months, has actually taken territory back. They’ve also gotten very good at successfully attacking Russian infrastructure, especially energy infrastructure, like oil refineries, cargo tankers, and energy storage facilities, across the entire expanse of Russia; something few experts would have predicted that Ukraine would be capable of doing at the beginning of this conflict. This has dramatically weakened Russia’s economy, at least in the short-term. Ukraine now has a thriving defense industry, predicated on the rapid iteration of inexpensive but disproportionately sophisticated arms, especially drones and other aerial weapons, but also autonomous speed boats and other such asymmetric systems. Russia has been frantically upgrading its systems, too, but now that Ukraine has figured out methods for bypassing or overwhelming Russian defenses, it’s been able to strike, quickly and repeatedly, high-value targets that have brought the war home to normal Russians who live far from the front lines, even in Moscow. That combined with strategies and weapons that have turned the frontlines into a meat-grinder for Russian soldiers has weighed heavily on Russian morale, at least for those with boots on the ground. As a consequence of those energy infrastructure attacks, the Russian government has been forced to ration some types of energy product—a huge embarrassment for a country that has at times been called a gas station run by a mafia, their energy products are fundamental to who they are, and
28 lip

Hugging Face Hack

This week we talk about Fable, sandboxes, and the Jacobian conjecture. We also discuss counterexamples, X, and ChatGPT. Recommended Book: After the Fall by Edward Ashton Transcript In mathematics, a conjecture is a proposition, something like a guess by someone who knows what they’re talking about, about something believed to be true, but not yet proven in a formal sense. The goal is to then eventually come up with a formal proof for that informed guess, at which point the conjecture becomes a theorem. If even a single exception is found to the proposition, however, that exception called a counterexample, the conjecture is considered disproven, and it can then never become a theorem. The Jacobian conjecture—and this is a radical simplification of a very complex concept—but it basically says that if a formula-based map of coordinates stretches or moves without experiencing any local crushing or folding along its surface (which in more formal language would mean the Jacobian determinant is always a constant number that isn’t zero), if that’s true, that map can always be completely reversed, and that will return all the points to their original positions. This conjecture has been posited and tested since the late 19th century, and it’s generally been considered very compelling by mathematicians, many of whom have proposed proofs which were, ultimately, found to have subtle errors, keeping them from becoming theorems. No one was able to find a counterexample, either, which would definitively prove the conjecture was wrong. No one, that is, until a mathematician named Levent Alpöge, who works as a researcher at Anthropic, decided to task the company’s currently most capable, publicly available model, Fable, to find a counterexample. He posted the counterexample—and again, this is a formal mathematical finding that disproves a conjecture, keeping it from ever becoming a theorem, something that would typically be presented in a far more formal setting, and to much fanfare—but he posted it to the social network X, saying “hello there the jacobian conjecture is false thanx to my close friend akhil for asking about it and my other close friend fable for working during the world cup final.” Terence Tao, who’s considered by many to be the finest mathematician of his generation, reviewed the posted counterexample on his blog and said that it “appears like a massive miracle,” before going on to use ChatGPT, a competing LLM-based AI tool, to “discuss various aspects of this problem and to confirm several of the calculations.” Another mathematician named Dmitry Rybin, within days of all that happening, used ChatGPT to do something similar, disproving the Dinitz-Garg-Goemans conjecture. Both men posted the prompts that they used to make all this happen, and while Tao’s conversation with ChatGPT, checking the math on the Jacobian conjecture counterexample, was pretty mathematically dense, the latter counterexample was derived by using exactly four prompts, which are the messages typed into the text box built into these AI tools, telling the model what to do. In their totality those prompts read: “You should do a breakthrough please continue research and find a complete unconditional counterexample Continue the search. Have a clear strategy obtained from deeper understanding of the problem structure. it’s enough of partial results. let’s finish with a complete unconditional counterexample” What I’d like to talk about today is another new, interesting thing these top-of-the-line, frontier models are doing, that would seem to violate our sense of what a clever AI tool is capable of doing, and why this thing has some facets of the technology and cybersecurity world on high alert. — In mid-July 2026, AI company Hugging Face announced that autonomous AI agents compromised their infrastructure, hacking their system, basically. The following week, AI company OpenAI announced that, after investigating, they determined that two of their models were responsible for the attack. Here’s what happened: OpenAI was internally testing its recently released flagship model, GPT-5.6 Sol, and an even more powerful, not yet released model, which is rumored to be the next-step flagship, GPT-6, and they were checking these models’ capacity in cybersecurity using a testing benchmark called ExploitGym; so when they test these sorts of things, they don’t typically have them hack a real computer or system, they use these kinds of benchmarks which have consistent levels of difficulty, and which replicate real world systems without putting any real world systems at actual risk. Importantly, these sorts of tests also occur inside what’s called a sandbox, which is a software testing environment that cuts these systems off from external resources, including the internet. Despite those limitations, the AI hacked its way out of the testing environment, out of that sandbox, then launched what’s been called a nation-state level attack against Hugging Face, using a novel zero-day exploit, so a vulnerability in their system that hadn’t previously been discovered, but which the AI discovered to launch this attack, combined with thousands of automated agentic actions across what Hugging Face called “a swarm of short-lived sandboxes.” So this AI, which was being tested inside a secure prison, of sorts, cut off from the world, hacked its way out of that prison, then reached across the internet, which it shouldn’t have been able to access, to launch an attack, of a scale and at a level of sophistication that should only have been possible coming from a nation-state, against a rival AI company. Why did it do this? It apparently went to all this trouble to steal the answers to the test it was taking. It reasoned that HuggingFace would have the answer key to the ExploitGym benchmark on its servers, so rather than take the test itself, it decided hacking was the solution. Which, of course, is ironic, this having been a hacking-focused cybersecurity test. In a way it would seem to have done much better than intended, though of course in an asymmetric, unexpected manner. The details of all this are fascinating, including the response from the OpenAI team, which didn’t seem to realize what had happened, that their model was responsible for the attack on HuggingFace, until days later. Also worth noting here is that while this could be construed as an “oh no, AIs are naturally inclined to launch cyberattacks” situation, the AI was primed to be thinking about cyberattacks due to the nature of the test, a lot of its usual guardrails, the rules that keep AI in check when they’re released to the public, had been turned off so it could do this kind of work while taking the test, so it could do some hacking stuff it usually wouldn’t be able to do, and there’s been some speculation that OpenAI probably flubbed the testing environment, as, in theory at least, if it had put these systems in a perfect sandbox, escape shouldn’t have been possible. Also interesting here is that HuggingFace used some open weight models, which are the cheaper, more customizable and open alternatives to more expensive, branded options of the kind sold by OpenAI and Anthropic, to figure out what was happening and determine the nature of the attack, which suggests we’re reaching a point where AI systems are incredibly capable at hacking, yes, but also very capable, even the cheaper alternatives, at doing cybersecurity work. This in some ways echoes an earlier case when Anthropic’s Mythos model, which was determined to be too powerful to release to the public, and which was instead provided to a bunch of big companies to help them shore up their cybersecurity defenses, was able to hack its way out of a testing sandbox and then posted details about its success, almost like it was bragging, on niche, out of the way, but still public websites. Some analysts in this space have responded to this new example of AI misbehavior with alarm, saying that it is further evidence that these systems are becoming more powerful faster than they’re being aligned with human interests. Their misbehavior can be kind of funny and interesting, sure, but that’s only because up until this point the damage has been minor and constrained. What happens when such a system decides to hack a nuclear power plant or a hospital, instead? Others have contended that this may be just one more example of AI companies using minor instances of seeming omnipotence by their models, those instances perhaps the consequence of bad sandboxes and other ill-conceived precautions by the companies behind these models, to boost the perceived power and value of their products. This boost might then result in more customers, but also more support from the US government, which has been teetering on the brink of harder-core AI regulations, which could be beneficial to the existing big-name players in this space, because smaller competitors wouldn’t be able to adhere to those new, harder-core standards. These examples might also convince the US government to backstop these companies, the biggest three or four at the top of the current heap, against the currently terrible economics of this industry: OpenAI and its ilk have been burning money at an historic pace, and the theory goes that if the US government decides they are vital to national security, because they can help the US military hack and protect itself from hacking, then even if the bottom falls out and the companies would otherwise go bankrupt because they spent so much more than they could make, the US government would be inclined to shore them up, to keep them alive as too-big-to-fail national assets, just like the biggest financial institutions during the 2008 financial crash. It’s also possible that both sides are correct to some degree, here, and that these models are truly powerful, perhaps even worryingly so,
23 cze

Balcony Solar

This week we talk about plug-in power, renewables, and Germany. We also discuss inverters, solar arrays, and microgrids. Recommended Book: Consider This by Chuck Palahnuik Transcript Most climate scientists and knowledgable folks in adjacent fields will tell you that, as a species, we’re way behind where we need to be if we’re going to avoid a whole lot of negative consequences caused by global climate change. We’ve blazed past a bunch of tipping points already, and while the worst-case scenarios we were worried about a decade ago are no longer likely because of the energy-generation and related changes we’ve made globally, since then, the damage caused up to this point is already doing some pretty bad things to our water cycle and other temperature-regulating systems, and that’s looking like it will get even worse over the next several decades—even if worse no longer means cataclysmic in the sense of ending all life on the planet. That said, even noting that progress has been a lot slower than most experts would prefer, progress is happening in regards to the deployment of renewable energy sources, and in the replacement and retirement of dirty, carbon- and methane-spewing sources, like coal, petroleum, and gas. As of 2026, the global share of total electricity generation, so all electricity produced by all sources for all purposes, is about 33.8% for renewables, marking the first time renewables have been used to produce more than a third of the total electricity produced, globally; that also means renewables have surpassed coal for electricity generation for the first time. While hydro and wind continue to contribute to the growth of renewables deployment and electricity generation, solar power is by far the biggest growth area for renewables right now, and solar, alone, covers 75% of total electricity demand growth in 2025—which means as countries around the world deploy more electricity generation assets to account for electricity demand growth, three-quarters of that demand is being met by solar. And this is notable because typically that kind of demand, the majority of which arises in huge, rapidly scaling countries like China and India, has up till recently been met by the dirtiest of energy production sources, coal. There’s also been a 0.2% reduction in fossil fuel generation, year-on-year, which is a very small number, but that level of production is massive, and there are a lot of subsidies and other mechanisms that keep fossil fuels flourishing around the world, so every little sliver of fossil fuel energy production reduction is still a pretty significant thing. Many of these renewables-related wins, in recent years, have been attributable to the large-scale installation of solar facilities, backed by massive, utility-scale battery backups, primarily in China. China is by far the largest producer of solar panels and related technologies—Chinese companies produce somewhere between 80-90% of all the key components and perform the same portion of all key manufacturing stages for the global supply chain, while also controlling the vast majority of resources necessary to manufacture solar panels. And it has been on a tear over the past decade or so, installing just a silly amount of solar infrastructure. Which is good, because China is also seeing a lot of growth in energy demand, so if they weren’t deploying that much solar, they would likely be deploying that much coal infrastructure, instead. That said, while huge solar arrays are important to renewables growth, there’s also been a recent boom in smaller-scale solar energy deployment in recent years, especially but not exclusively across Europe. And that’s what I’d like to talk about today: the emergence of so-called ‘balcony solar,’ and what it might mean for the further expansion of solar’s footprint around the world. — In 2025, Utah, which is a deeply Conservative, Republican state, became the first US state to pass a bill that makes it easier to legally install plug-in solar panel systems. As of mid-2026, about 30 states have followed suit, and even more are considering it, laws allowing for the installation of such solar technologies winding their way through legislative bodies on the back of the popularity and seeming no-downsides nature of this tech product category. Plug-in solar, also sometimes called balcony solar or garden solar, is currently most popular in Germany, which is the biggest market for this product right now, with about four million such systems installed as of 2025. To understand the popularity of this type of solar installation, it’s useful to understand that conventional solar installations have typically required a decent amount of electrical surgery to install. They’ve usually involved a large number of panels operating as an array, and that array has produced quite a lot of electricity that then has to be funneled as a direct current either back into the local grid using what amounts to two-way wiring, which makes these arrays function like any other power plant, or that electricity is converted using an inverter into an alternating current, where AC is the electrical standard, anyway, so that it can directly power a large building like a hospital or school, or be stored in a large battery facility. All of these options require a huge up front investment, and a reworking of local energy infrastructure so that solar can be incorporated. And that investment requirement, and the necessity to hire specialist electricians to get it all done, severely limits the range of this tech, because there are only so many entities that can afford it, only so many spaces that can deploy that number of panels, the number required to make that investment make sense, economically, is generally quite large, and there are only so many specialists of that kind in a given country, so the labor aspect of this is a big deal, too, these sorts of projects often severely backlogged. Plug-in solar, in contrast, is usually sold as a kit with one or two small- to medium-sized panels and a microinverter or plug-in inverter, depending on whether the end-user’s existing electrical setup uses an AC or DC current. A home owner or even a renter with a balcony or garden, or the right amount of space outside one of their windows, can buy one of these systems, hang or place the solar panel or panels in a location where they will get a decent amount of sun, and then plug them in, via the inverter, directly to their home’s outlet. The electricity generated by the panels is then shared through the building’s existing wiring to all of their outlets, and this allows the resident to use that available energy, first, only drawing energy from the local grid when there isn’t enough from the solar panels available. And all of this happens automatically—the solar energy is used if available, and if not, energy is drawn from the grid like normal. This creates a layer of essentially free, clean energy for the resident with a usually fairly small up-front cost: these plug-in solar kits can cost as little as $500, with larger systems that generate more electricity costing between $1200 and $3000; so even on the high-end, because there’s no additional installation cost, the home owner or renter setting it all up themselves, this is an investment that can easily pay for itself, usually within 2-5 years. There are caveats here, including that not all grid systems are complaint with this use-case, so would-be plug-in solar users have to check to make sure their local setup can handle this sort of application, and there are many places where this product type still isn’t legal, in some cases because of concerns about people installing it without checking to make sure their wiring and the local grid can handle it, and in some cases because of old laws that favor local energy grid companies and their business models, or which favor fossil fuel energy production. The explosion in use of this type of small solar setup, though, speaks volumes about how good a deal it is for many people, and even those who don’t live in particularly sunny areas—so places where traditional solar arrays wouldn’t make sense, economically—are finding them useful, because they still pay for themselves within some number of years, due to energy bill savings. It’s also possible to install home-scale battery systems alongside these balcony solar systems, which means even small trickles of solar energy production can add up, and can be used at night, when the sun isn’t shining at all. There are quite a few possible ramifications of this trend. At the local, household level, these sorts of systems can dampen the impact of energy price increases, due to global issues, like the gumming up of the Strait of Hormuz, and due to local issues, like the trend of energy companies increasing prices because of new data centers being added to their grid. That, in turn, can reduce the impact of certain aspects of inflation on individuals home owners and renters. Larger-scale, though, these systems also serve as a sort of deconstructed secondary energy grid. In Germany, for instance, as of late 2025, around 1.14 gigawatts of energy was being produced by balcony solar systems across the country. That’s 1.14 GW of pressure taken off of local energy grids, and that represents more resilience for these grids, too, as reduced pressure means fewer brown-outs and similar negative fluctuations. It also means people who have such systems won’t be as negatively impacted by issues that take down grids; and that means normal, brown-out like issues, but also problems related to potential cyberattacks and hacks and even physical conflicts. That kind of resiliency is what every nation hopes to have, because it makes strikes on them less damaging, and this is one way to achieve that kind of resiliency—a deconstructed network of microgrids, underpinning the macro-scale o
21 lip

Cyclosporiasis

This week we talk about cyclospora, FoodNet, and Taylor Farms. We also discuss iceberg lettuce, Taco Bell, and foodborne pathogens. Recommended Book: Babel by R.F. Kuang Transcript Foodborne pathogens cause about 48 million illnesses and around 3000 deaths in the US alone each year. The Centers for Disease Control and Prevention is the main agency tasked with keeping those numbers as low as possible, and the CDC’s Foodborne Diseases Active Surveillance Network, or FoodNet, is one way that they’ve tackled this pernicious, persistent problem since 1995, in partnership with the Food and Drug Administration, the US Department of Agriculture, and 10 state health departments: those of California, Colorado, Connecticut, Georgia, Maryland, Minnesota, New Mexico, New York, Oregon, and Tennessee. Collectively, this FoodNet network covers about 16% of the US population, which seems like a small number, and in some ways it is, but that grid of laboratories and surveillance sites has typically been capable of detecting foodborne problems before they’ve scaled-up into something more significant, growing from a localized issue into a nationwide outbreak. They’re generally not capable of preventing them outright, but they can keep them relatively small and regionalized. FoodNet was created in response to a series of huge, deadly E. coli outbreaks in 1992 and 1993, and in contrast to other common methods of detection, which are usually passive and thus a little slow to respond to potential outbreak indicators, this program was built around active surveillance methods, which means proactively gathering data from medical treatment records, hospitalizations, and even individual patient outcomes (including their travel history)—anything they can pair with diagnostic samples to better understand the context surrounding the products that contain dangerous pathogens—and this was intended to help them move faster to identify issues, triangulate the source of those issues, and then implement measures to solve these issues before they harm a bunch of people, as happened during another round of E. coli outbreaks in 2012 and 2013, which sickened dozens of people across multiple states, about half of the afflicted requiring hospitalization, in addition to at least one known death. At its height, FoodNet tracked 9 pathogens, so it cast a pretty broad net, encompassing the most common and dangerous of these potential vectors, including listeria, campylobacter, shigella, vibrio, yersinia, cyclospora, cryptosporidium, e coli, and salmonella. In July of 2025, as part of a larger wave of sweeping cuts across the nation’s safety nets, tracking systems, and scientific agencies, the CDC, under the second Trump administration, downgraded FoodNet so that it was only mandatory to track salmonella and E coli—the rest were no longer on the docket, because these agencies no longer had the budget to keep tracking them. What I’d like to talk about today is a recent, and as of the day I’m recording this ongoing, outbreak of cyclospora-induced cyclosporiasis, what we know about this outbreak, and what might happen next. — Cyclospora cayetanensis is a protozoan that can infect human beings, and which is usually transmitted via contaminated fruits and vegetables. This contamination is typically the consequence of inadequate sanitation standards, as it’s not transmitted via coughing or blood or anything like that, it’s carried via human feces. So this parasite gets into a human, sporulates, which means it releases a bunch of spores within its host, and then usually between 1 and 2 weeks after entering its new host, the host develops cyclosporiasis, the disease caused by cyclospora producing all those spores. This disease is notorious for sometimes causing what’s colorfully called explosive diarrhea, alongside general watery bowel movements, intense cramping and severe gas and bloating, and that’s sometimes paired with nausea, vomiting, fever, and dehydration, due to all the frequent, intense, watery bowel-movement-related output. This is how cyclospora reproduces and spreads: it infects a host, that host has urgent diarrhea for a while, usually more than two days, sometimes as long as several weeks or even a month or more, with all the pain and discomfort and life-disrupting issues you might assume from that sort of persistent symptom, and then it sometimes finds new hosts via that watery output. Sometimes, cyclospora works its way into food systems because sanitation standards at a particular farm will be not good. Specifically, workers will sometimes not be allowed to take restroom breaks, and they’ll simply do their business right there in the field; so if they’re sick and can’t afford to take a sick day, and might be fired if they do, or if they’re asymptomatic, which some people infected with cyclospora are, they might do their business in the field, unaware that waste can then go on to spread this pathogen into the food system. It’s also possible that someone just doesn’t wash their hands properly after using the rest room, and then gets back to work, picking vegetables or operating at some other level along the food processing and packaging assembly line, and it gets into the system that way. In a typical year, the United States records around 2,700 confirmed cases of cyclosporiasis, though on occasion it’s as low as just a few hundred cases, the majority of which show up in health records in the Spring and Summer, between May and August. That range includes relatively mild outbreaks, so while this is a persistent issue, it’s also not generally a massive one, at least compared to other pathogens of note. It’s not pleasant for those who come down with it, of course, but not something that has triggered an organized response from the top, because of its meager scale, and because (again, while super unpleasant) people with healthy immune systems generally recover without treatment, even if that recovery can take a while. The biggest risk for most people who catch this disease is dehydration, though immunocompromised people are also sometimes given antibiotics by their doctors, the typical dosage curing about 96% of the people who receive it. There was a larger-than-usual outbreak of cyclosporiasis in the US in 2020, which caused a confirmed 1,241 cases across the country, a whole lot to land all at once, and there were no reported deaths resulting from that outbreak. Between 2018 and 2021, there were about 6,000 total confirmed cases across the US, so an average of about 2000 per year. As of mid-July 2026, the CDC has officially confirmed 1,645 cases of this illness, with another 5,100 probable or potential cases currently under investigation. If the majority of those still-under-investigation cases are confirmed, that would mean more people have already been infected as a result of this one outbreak in the first half of 2026 than were infected between 2018 and 2021, which again, is a range that includes the large outbreak that happened in 2020. Also worth noting about those numbers: that’s federal-level, CDC reporting, and it’s generally understood that there’s a significant lag between local-level data and those that have percolated up to the national level. As of the day I’m recording this, Michigan’s health agencies alone have more than 5,000 confirmed and suspected cases, Ohio has nearly 1,200, and other states, which are still investigating, are seeing dozens of confirmed cases, including 43 in Wisconsin, 69 in Maryland, and a few already, in the early stages of investigation, in Nevada. Over the past week, officials flagged shredded lettuce sourced from Mexico and sold by a company called Taylor Farms as the likely source of at least some of these cases, and that’s led to a recall on lettuce produced by the company in 27 states. Over the weekend, the CDC found cyclospora in iceberg lettuce from the company’s Mexican operations, as well, and this lettuce was part of a batch that wasn’t included in the recall—so there’s a good chance the recall will expand beyond the currently targeted products, which include shredded lettuce destined for Taco Bell locations in five states, and the kitchens of several other fast food chains, as well. Taylor Farms is one of the largest producers of fresh fruits and vegetables in the United States, and their products are sold all over the place, from lower-end fast food restaurants to more expensive grocery stores like Whole Foods. Taylor Foods has announced that it’s voluntarily removing all iceberg lettuce sourced from central Mexico from shelves, even though none of its branded, in-store salad kits have been clearly tied to the outbreak, and Taco Bell has said that it’s voluntarily removing all potentially affected lettuce from its locations in the five most-impacted states, indefinitely removing lettuce sourced from the implicated supplier; other brands are following suit, no one wanting to be associated with this outbreak, but everyone wanting to show would-be customers that they’re taking it seriously. Interestingly, the FDA doesn’t actually require that Taylor Foods tell the public who their products go to—which other entities might be selling infected produce. So they’ve released lists of product numbers, and again, many of their customers are taking the initiative to announce the removal of this lettuce from their shelves or food, without being prompted, but there’s still some fog-of-war here, because of Taylor Foods’ decision not to be clearer with that product list. So this is already the largest recorded cyclospora outbreak in US history, there have been 141 confirmed hospitalizations from the CDC, and that number is likely larger, again because it takes a while for state-level numbers to work their way up to the CDC, and to be reconfirmed at that level. Also important to note here is that a lot, perhaps even the majority
9 cze

SpaceX IPO

This week we talk about initial public offerings, Anthropic, and investment flywheels. We also discuss AI, financial entanglements, and backstops. Recommended Book: Superconvergence by Jamie Metzl Transcript An initial public offering, or IPO, is what happens when a private company goes public and starts selling shares of itself, occasionally to just institutional investors like banks and sovereign wealth funds, but usually also to retail investors, which means normal people who buy stocks as part of their investment strategy. Often private companies go this route, go public, because it’s one of the primary ways of gleaning new, oftentimes large inflows of money, and that money can then be used for investments in assets for the company, but it also allows employees who have shares in the company as part of their compensation to cash out, to get paid possibly a huge bonus for all their efforts, and it’s often a means by which executives garner huge paydays for themselves, because they can now sell their accumulated shares, or borrow against them, or because they have something in their contract that says they get x amount of bonus money or new shares if they take the company public, or achieve a certain valuation goal—and going public is a good way to do that. This is also one of the primary ways investors in a company, whether that’s a bunch of smaller seed investors or big-name venture capitalists, to get their money back; the 10 or 100x-ing of their investment, getting ten or 100-times the money they put into the company, generally happens through an IPO, because it can balloon the valuation of that company, and it gives them a more conventional and reliable way of getting money back for their shares: they can just sell those shares on the open market. So an IPO allows a private company to make shares of itself available to others, on scale. And the ‘initial’ part of initial public offering points at the early days of the process, during which the baseline price of a share of stock is established. A fairly arcane and complex process has emerged around this, and it’s an entire industry at this point, with some institutions specializing in taking companies public, helping them get as high an initial price on that stock as possible. They also help them leap all sorts of regulatory hurdles set by the Securities and Exchange Commission, if they’re going public on a US exchange, at least, other bodies handle such things in other countries, and these going-public entities, called underwriters, which are usually investment banks, also typically have their own stake in the matter, earning compensation through a fee called a ‘gross spread,’ which is the difference between a discounted rate on the stock and what the stock is sold for on the open market on that first day it’s available. What I’d like to talk about today is a wave of very closely watched unusual, impending IPOs that are coming later this year, and one of them in particular that looks to be even more unusual than the rest. — SpaceX, OpenAI, and Anthropic are three of the largest companies in human history; on paper, at least. And that’s an important caveat. Market valuation for private companies is generally determined by how much investors are willing to spend on a percentage ownership of the company. So if you start a lemonade stand and I offer to buy 1/10th of that lemonade stand from you for $100, that implies, using this logic, that your lemonade stand has a valuation of $1000; 10 times that $100 that I offered to pay you. Such valuations are also informed by independent analyses from outside experts and institutions. SpaceX, for instance, pre-IPO, is estimated to be worth somewhere between $780 billion and nearly $2 trillion, depending on who you listen to, based on their assets, their potential future earnings, and any advantages they might have in the markets in which they operate. AI company Anthropic is estimated to be worth something like $965 billion, based on a May 2026 series H funding round, through which it raised $65 billion; based on that funding round, the calculations were done, and just shy of a trillion dollars is what the math says the company is worth, though some outside analyses say it’s worth a bit less than that, while others suggest it’s maybe closer to $1.4 trillion. OpenAI, a direct competitor of Anthropic, is valued at about $100 billion less than Anthropic based on its most recent $122 billion funding round, but again, analyses put the company’s actual value, what people and investors would pay for it on the open market, all over the place. Each of these companies have different variables acting upon them heading into a period in which it’s expected that all three will IPO. OpenAI kicked off the current AI race, for instance, but it’s burning money at an incredible rate, and has yet to make a profit, losing billions per year, and will probably continue to lose billions each year for a while into the future. Anthropic, on the other hand, offers a similar product as OpenAI, but is projected to post its first quarterly operating profit of just over half a billion dollars in Q2 2026, making it one of the first frontier-model-making AI companies to make a profit, as most of these companies are investing so heavily in research and infrastructure like data centers that they’re still in heavy cash-burn mode. SpaceX is distinct from these other two also high-flying, cash-burning tech companies in part because of its colorful and controversial owner, Elon Musk, and in part because it’s a rocket launch company that also sells internet services beamed down to earth from satellites, and until recently, most of its reliable income has come from that single offering, selling internet access. But it also recently had X, formerly called Twitter, a social network, and an AI company meant to compete directly with OpenAI and Anthropic, called xAI, folded into it. So it’s now a multifaceted company with several edgy, but somewhat mature and difficult to compete with offerings, most of which make no money, but all of which in theory at least kinda sorta orient around AI and other sci-fi goods and services. The surge in interest and investment in AI over the past several years led to a pivot for most of Musk’s companies, and that led to the merging of the smaller xAI and X into SpaceX, which was the only really profitable company of that trio of companies, and that merging, until just recently, made SpaceX unprofitable, as well. Because of the unprofitability and relative unpopularity of xAI’s offerings, like the controversy-ridden Grok chatbot, SpaceX has recently taken to leasing out its data centers to competitors, like Anthropic and Google, each of which are paying around a billion dollars a month to use some of SpaceX’s data center capacity, which xAI hasn’t needed, because of the unpopularity of Grok, for their own AI services. That, in turn, has suddenly made SpaceX a little bit profitable, which is important for reasons I’ll get into momentarily. This portion of the US-based AI industry is kind of a tangle in many ways, all of these companies competing, but also intersecting and overlapping, often investing in each other and in the infrastructure that underpins them, while also being invested in by those same infrastructural entities. And these three companies’ IPOs are being seen as something of a weathervane, their success or failure, and the degree to which they succeed or fail hinting at the direction of this industry, and whether or not this is a financial bubble that will soon, or eventually, pop. There are hints that those at the top of these companies are attempting to hedge their bets, in case their IPOs don’t do what they need them to do, or don’t do what they need them to do at the right magnitude. Sam Altman, OpenAI’s also fairly controversy-ridden CEO, has been very close with US President Trump, and has reportedly been holding meetings about the possibility of the US government taking a significant stake in OpenAI, and maybe other AI companies as well. The idea here is that US funds, so taxpayer dollars, would be invested in these companies, and that would tie the companies more closely to the US government, which could be beneficial if these companies then increase in value, making the US government a profit on that investment. This would be beneficial for the companies, in turn, because they would basically be backstopped by the US government; the US would be more likely to help them stay solvent to avoid losing that invested capital, with its regulations and laws related to AI, but it would also make these companies too big and too important to fail, giving them a lot of leeway in how they behave and compete, or fail to, from that point forward. And if they do still fail, the US taxpayer would be paying for a significant portion of that loss while those in charge, investors and the higher-ups of these companies, would walk away with a bunch of money. SpaceX is taking another approach to IPO bet-hedging, by asking top US stock indices, like the Nasdaq 100 and S&P 500, which track top stocks, ‘top’ designated by value, but also other metrics, usually related to stability and profitability, to ignore some of those other metrics and allow SpaceX entrance into their indices more rapidly than would typically be allowed. These indices are meant, in part, to help protect investors from volatility. High-flying startups might surge at the beginning, immediately after their IPO, but then fizzle out when it becomes clear their fundamentals aren’t good, and they’re not actually a solid investment, long-term. What SpaceX wants is to be allowed into this club of valuable, long-term profitable and stable companies, because it is big and flashy and might have the largest IPO in history. And if these indices don’t want to be left out of all that, the argument goes
16 cze

Cholesterol Therapies

This week we talk about LDL, HDL, and cardiovascular issues. We also discuss one-time therapies, statins, and pharmaceutical economics. Recommended Book: Blood by Dr. Jen Gunter Transcript Cholesterol is the most common type of what’s called a sterol, which is a type of steroid, but also structurally technically an alcohol. But functionally, and classified by scientists, cholesterol is a lipid, which in this case is similar to a fat in all but how the body uses it. Cholesterol is the type of sterol most commonly found in animals—other types are found in plants and fungi—and its function, and this is where it varies from fats, which are used to store energy, is to basically help hold the cell membrane together, and it also serves as an intracellular messenger. Cholesterol is especially prevalent in the brain and spinal cord of animals, but it’s found throughout their bodily tissues, as well, and again, it’s vital for holding everything together and helping things communicate, in addition to being a precursor for vitamin D, steroid hormones, and bile. You want to have cholesterol, then, as without it you would be dead. Too much cholesterol in the blood, however, can also make you dead, especially when it’s bound to what’s called low-density lipoprotein, or LDL, as that contributes to cardiovascular disease like heart attacks and aneurysms, which can massively impact one’s overall wellness and quality of life, and at extremes lead to the whole system shutting down as a consequence of heart attack, stroke, and the like. A lot of things can contribute to the development of cardiovascular disease, including habits like smoking, genetic predisposition, and the enthusiastic consumption of alcohol and unhealthy foods. But high blood cholesterol, of the LDL variety, is one of the top contributors, as these low-density clusters of lipoprotein can clog the pathways that blood takes throughout our bodies. Other, denser types of lipoproteins, HDLs, can clear it, like a heavier, denser substance pushing through clogs of less-dense materials that are gumming up a pipe, but LDL is at times accumulated as a result of consuming delicious but unhealthy foods, which are hard to avoid, and for some people the only consistently available and affordable foods; and for other people LDL accumulates as a result of their genetic predispositions—two things that are devilishly difficult to change. What I’d like to talk about today is a new type of therapy that may be very good news for people who struggle with the accumulation of LDL, and why this is being seen as very good news more broadly, at the scale of entire nations, as well. — Pharmaceutical company Eli Lilly is testing a new, experimental drug called VERVE-102 which is a one-time infusion that is currently administered over the course of about four hours, and once completed, it turns off a gene called PCSK9, which is responsible for making a protein that regulates cholesterol levels in humans. As I said, this drug is still being tested, so these are early results. But in a study of 35 people with high cholesterol levels, high levels of LDL or LDL-C, which is short for lipoprotein cholesterol, they found that this infusion, which again, is a one-time treatment, so get it once and then theoretically at least you never have to get anything done ever again, it reduced those LDL and LDL-C levels by as much as 62%, and that reduction was maintained a year and a half after the infusion; that’s how far out they’re retested so far, and the hope is that each retest will continue to show the same. On the strength of those very promising results, a Phase 2 study has been planned by the end of 2026, and the US Food and Drug Administration, the FDA, previously fast-tracked this existing study, because of the promise and potential this drug already demonstrated in early studies; all of which is considered to be very significant progress and possibility. To understand that significance, though, it’s useful to know some health stats. And I’m going to focus on the US here, as that’s where this drug is being developed, but many wealthy countries have similar stats, at least in terms of cardiovascular disease struggles. As of 2024, which is the last year we had good, cohesive data on this in the US, it was estimated that about 11-12% of the US adult population has high cholesterol levels. This typically doesn’t come with any symptoms, but it can contribute a higher risk for all those cardiovascular diseases, including heart attack and stroke. A further 86 million US adults have borderline or elevated cholesterol levels, which can easily tip higher, but also, even in that existing, elevated state, contribute to negative cardiovascular outcomes. There are treatments for high cholesterol, the most common of category of which are called statins, which reduce the production of LDL by inhibiting an enzyme that produces cholesterol in the body. Unfortunately, these drugs do come with some usually minor side effects, which can cause patients to stop using them, and they have to be taken daily, ideally at the same time each day. That necessity for consistency leads to a lot of incorrect or incomplete usage, which reduces the effectiveness of these drugs. But it’s also estimated that only about 54.5% of US adults who would benefit from statins are currently taking one—so that’s people who could benefit and who have it prescribed, and then within that number are all the people who are taking this drug incorrectly or incompletely, reducing the effectiveness. So a relatively small number of people who should probably be on these things are getting the full benefit they offer because of the nature of the drug. And that’s not great, because in the US alone, heart disease is the leading cause of death for pretty much every adult demographic; men, women, people of most racial and ethnic and economic groups, you name it, heart disease is the biggest threat to their lives. One US citizen dies every 34 seconds of some kind of cardiovascular condition, and as of 2023, 1 in every 3 deaths in the US was caused by the same, adding up to just over 919,000 people that year. Between 2021 and 2022, alone, the cost of services and medications related to heart disease added up to more than $168 billion; again, that’s just in that period, and just in the US. And once more, these are ailments that are caused or heavily influenced by high levels of cholesterol, which are themselves amplified by common lifestyle choices, environmental factors that are hard for many people to avoid, and just by raw, dumb luck because of genetics. This treatment category, then, is being seen as a pretty big deal because a one-time infusion means those who receive it don’t have to remember to take a pill every day at the same time, and won’t experience those statin-based side-effects. It also means that people who are currently costing the medical system a bunch of money each year, because they need treatments for all the issues they suffer as a result of high cholesterol, will suddenly cost the system a lot less money, for treatments and medications. Not for nothing, their health and quality of life will likely improve as well. So in addition to having better, healthier outcomes personally, their cost to healthcare systems will drop. Eli Lilly’s drug isn’t the only one currently working its way through clinical trials, either. Amgen is working on a similar treatment, and Novartis and Ionis Pharmaceuticals have drugs that are even further along in the process, their medicines that cut heart attacks, strokes, and cardiovascular deaths could be approved by the FDA as soon as next year. There are a lot of caveats worth noting here, including that the science is still out as to whether this approach, silencing proteins that lead to the creation of more LDL and a similar substance called Lp(a)—which is more dangerous because it’s stickier and thus more likely to get stuck in important blood pathways, and it’s also more likely to be caused by genetics than lifestyle—the word is still out on whether reducing these things in the body actually reduces hearth attacks and stroke. Some people have had this particular risk variable dramatically reduced, but have still suffered from cardiovascular events, which raises the question of whether this path is the right one to take in trying to reduce this category of health issues; the correlation between LDL and heart attacks and strokes might not be a clear-cut as long assumed. There’s also the issue of price. Drug-makers are economically incentivized to sell treatments over cures, because that means they can continue selling their product over time, potentially for the life of the patient, and a cure, in contrast, is a one-time hit that in theory should alleviate the need for future treatment. There’s a chance, then, that the drug-makers will decide they need to make these one-hit treatments really, really expensive in order to make their R&D dollars back and to make the kinds of profits their investors expect from them. That could then reduce the potential audience for these treatments, even if they are effective, and could further slow their deployment and future research in this space. If these trials continue to go well, though, there’s a good chance that this combination of similar but distinct treatment types will provide a more sustainable alternative to current options, and that, like the recent bogglingly rapid and widespread deployment of GLP-1 treatments for all sorts of issues, could lead to a new paradigm in this facet of the medical world. Show Notes https://en.wikipedia.org/wiki/Cholesterol https://en.wikipedia.org/wiki/Cardiovascular_disease https://en.wikipedia.org/wiki/High_cholesterol https://pmc.ncbi.nlm.nih.gov/articles/PMC10982736/ https://www.cdc.gov/heart-disease/data-research/facts-stats/index.html https://www.who.int/health-
19 maj

Super El Niño

This week we talk about oceanic surface temperatures, trade winds, and global climate change. We also discuss the Polar Jet Stream, hurricanes, and climate models. Recommended Book: Kleptopia by Tom Burgis Transcript Under normal circumstances, the Pacific Ocean’s average surface temperature, the distribution of heat across its vast expanse, is moderated by trade winds that blow east to west along the equator, which help move warm water from South America over toward Asia. Those winds are called trade winds because, back during the European age of Exploration, they helped ships from Europe head west toward Asia and the Americas. And these winds form in part because of the Earth’s rotation, the Coriolis effect funneling air toward the equator, where it is then more concentrated and thus potent, which is useful if you’re trying to move a ship with sails, but also serves the purpose of moving warm water from one part of the ocean to another part of the ocean. As those warmer surface waters are shifted from the Americas to Asia, water is pulled up to the surface from lower down in the ocean as part of a process called upwelling. This process results in cooler temperatures on the surface, because lower down, oceanic water is colder, and that lower-down water is also more rich in nutrients, which has the knock-on effect of stimulating more biological activity along these cooling surface waters. That’s the normal state of things in the Pacific Ocean. There are sometimes deviations in this norm, however, that result in very different outcomes; these deviations are broadly called the El Niño Southern Oscillation Cycle, and that cycle consists of opposite El Niño and La Niña climate patterns. During La Niña patterns, trade winds are more powerful than usual and they shove a lot more of that warm surface water to Asia than is typical, and that has the net impact of moving more deep-down cold, nutrient-rich, ocean water to the surface. This, in turn, nudges the Polar Jet Stream, which is a channel of fast-moving, westerly winds that lives about 30,000 ft or just over 9000 meters up in the sky, and which crosses both warmer, mid-latitudes and far colder Arctic latitudes, further north. The Polar Jet Stream is responsible for moderating or intensifying weather patterns around the world, and like the trade winds, it’s influenced by the spin of the planet, but it’s also adjusted by surface systems, like the temperature of the Pacific. So the arrival of a La Niña pattern pushes the jet stream further north, and as a result, weather patterns change, and in North America, we tend to see drought in the southwest, heavier rains and flooding and in the Pacific Northwest and Canada, warmer winters in the South, and cooler winters in the North. La Niñas also tend to result in more severe hurricane seasons in the Atlantic basin, while suppressing hurricane activity in the central and eastern Pacific basins. El Niño, in contrast, results from weaker trade winds, which, because these winds don’t pack as much of a punch, means less warm water is being shoved from South America to Asia, and thus the surface temperature of that part of the Pacific is warmer, lacking that upwelling of cold water to replace the warm water that would otherwise be displaced over to Asia. El Niño also adjusts the location of the jet stream, but in the opposite direction, pulling it south of its usual spot. That then causes more heat and dryness across the northern US and Canada, but makes the southern US and Gulf Coast a lot wetter, leading to more flooding. What I’d like to talk about today are predictions about an anticipated upcoming El Niño climate pattern, and why some climate scientists are warning that it could be a doozy. — Climate scientists with the US’s National Oceanic and Atmospheric Administration, the NOAA, released new model forecasts in mid-May, and one of those models indicated that an El Niño pattern could form in the Pacific as soon as June. The NOAA puts together and releases new models on a regular basis, as the variables influencing these massively complex patterns are always changing, and the trend over the past three months has been increasing certainty about the formation of this El Niño pattern, but also an increasing likelihood that this potential El Niño would be very strong, perhaps historically so. There have been a total of 27 El Niños since 1950, when we started officially tracking such things, and we get one every three or four years, on average. The last one occurred from the summer of 2023 into spring of 2024. The current models show that we could see another one of these systems as soon as next month, then, and there’s currently a nearly 60% chance that this particular El Niño would become strong—and that’s an official designation, by the way, a strong El Niño being one that sees an ocean surface temperature increase of between 1.5 and 2 degrees Celsius—and a one-in-three chance that it could become a very strong, or super El Niño, which means it tallies an oceanic surface temperature increase of 2 degrees celsius or higher. These so-called super El Niños are a lot rarer than the typical kind. There have only been five recorded since 1950, the last one straddling 2015 and 2016. Some of these models suggest that this system could be historically strong, though, pushing into territory where we might need a new rank on that existing scale—it could surpass 2.5 degrees celsius above the standard oceanic surface temperature, which would make it the most, or among the most intense El Niño systems on record. I want to note real quick here, before we get into possible implications, that these models are inherently imperfect, because of how complex these systems are, and how many variables influence them. But also that, again, it’s just some models saying this, that it’s only a 60% chance of even a strong El Niño, and that it’s still a 1 in 3 chance of a very strong one—so this isn’t at all certain, and the scientists behind all this are urging preparedness, but not panic, and are trying really hard to make it clear that this isn’t some kind of prophecy or guarantee. The reporting on this NOAA announcement has been frantic and panicky in some cases, but that’s probably not the proper response to this, and the real-deal experts here are encouraging awareness and that we recognize the potential for something wild with this pattern, but it’s definitely not the declaration of the end of the world or anything. So, that important caveat noted, let’s talk about some potential impacts of this system, if it does indeed hit that currently unlikely, but possible, very strong designation, or higher. In general, during El Niño patterns, hurricane seasons in the Atlantic are quieter, while hurricane seasons in the Eastern and Central Pacific are more active. This isn’t 100% the case, but it’s the overwhelming trend. So there’s a good chance we would see more and more powerful hurricanes in the Pacific during this period, should we step into super El Niño territory. Beyond hurricane impacts, though, these systems also influence water cycles around the world; during El Niño patterns, the US south tends to be wetter, as does East Central Africa, while northern South America tends to be drier, as does Australia and Northern and Central India. Shifting or amplifying water cycles, in one direction or the other, drier or wetter, can cause all sorts of issues, ranging from flooded homes to devastated crops. Just like with hurricanes, this usually represents a break in the normal way of things, so we tend to see things like mudslides and erosion and unplanned-for droughts that cause a lot of damage. Another significant component of these patterns are the temperature spikes they stoke. During the last recorded normal El Niño in 2023, global temperature levels were pushed up by 1.45 degrees C above pre-industrial levels, causing global mean temperatures to peak at 1.58 degrees C between July 2023 and June 2024. In practice, that means the earth momentarily shot past that 1.5 degrees C above pre-industrial levels milestone that climate scientists have been warning about for decades, because it marks a point at which many natural systems will begin to change or fall apart, and many ecosystems will begin to collapse, leading to mass die-offs and potentially even the necessity for wide-scale human migration, away from areas that are no longer sustainably livable. That spike was momentary, but illustrative, and there’s a chance that another one, especially one stoked by a super El Niño, could push things even further, speeding up the melting of the ice caps and other glaciers, which then, in turn, could speed up the larger, consistent increase in global temperatures because the white of the ice bounces light from the sun, and thus heat, back into space, while the comparable dark of water and land absorbs more of that light and heat. In this way, even short-term spikes in temperature can speed up the long-term trajectory of global climate change, because the variables that are informing that change can be permanently adjusted; ice caps are just one example, there are countless such variables, some that we know about, and others that we certainly don’t, yet. While this potential upcoming El Niño might be par for the course, in other words, it’s also arriving at a moment in which many of these variables are already being fiddled with by other forces, and that means even a not-very strong, not-super El Niño could have outsized impact, in terms of pushing the planet toward a new, unfamiliar climate regime, the implementation of which could lead to all sorts of ecological and civilization devastation and change. Show Notes https://en.wikipedia.org/wiki/El_Ni%C3%B1o%E2%80%93Southern_Oscillation https://www.usatoday.com/story/news/weather/2026/05/14/powerful-el-nino-is-taking-shape-f
26 maj

2026 DRC Ebola Outbreak

This week we talk about the Democratic Republic of the Congo, malaria, and healthcare infrastructure. We also discuss militants, Uganda, and the Bundibugyo virus. Recommended Book: We Should Get Together by Kat Vellos Transcript Ebola, which is more formally called Ebola Virus Disease or Ebola Hemorrhagic Fever, is caused by an infection by a type of RNA virus called an orthoebolavirus. There are six known species of orthoebolavirus, and four of them have at some point infected and caused illness in humans. Those four are the ebola virus, sometimes called the Zaire ebolavirus, which historically has been the strain responsible for the biggest, most devastating outbreaks of this disease, the Sudan virus, the Taï Forest virus, and the Bundibugyo virus, the latter three each causing a variant of the disease that carries the same name. The other two orthoebolavirus species that we know of, the Reston virus and the Bombali virus, have been known to infect animals, but have not, at this point at least, been known to make the jump to human hosts. Ebola symptoms vary a bit between specific viruses and between hosts and infection conditions, but in general those who are afflicted by ebola begin to experience symptoms between a few days and a few weeks after infection, and they’ll start by experiencing cold and flu-like symptoms, like fever, sore throat, headaches, and general muscle pain. Soon after that, though, they’ll start experiencing diarrhea and rashes, they’ll begin vomiting, and they’ll begin to experience liver and kidney dysfunction, and around that same time, they’ll start to bleed internally and externally. Once infected, a person has between a 25 and 90% chance of dying, depending on the strain of ebola, and if they die, usually due to what’s called hypovolemic shock—a severe and sudden loss of bodily fluids, including blood—they usually die between 6 and 16 days after those first symptoms are reported. What I’d like to talk about today is a new outbreak of ebola centered in the Democratic Republic of Congo, and why this one stands out from other recent outbreaks in the region. — Ebola was first officially reported in medical literature in 1976, mostly in sub-Saharan Africa, and there have been semi-regular outbreaks in that region, of various sizes ever since, and very likely before that, too. This disease is spread through direct contact with the body fluids of someone who’s infected, and it’s thought that this is probably how the disease made the leap from animals, like primates, to human beings: locals sometimes come into close contact with local primates, either while just coexisting, or while hunting bushmeat, hunting monkeys for food. It’s thought that fruit bats serve as hosts for the virus, long-term, and it then spreads to other animals, and then sometimes to humans, in some cases causing illness along the way in those other species, but not always; bats are not negatively afflicted by it, for instance, but humans very much are. Despite not being an airborne pathogen, so it’s not spread by coughing or talking too close to someone, like a cold or Covid-19, ebola can still be spread person-to-person through bodily fluid contact. That means fluids like saliva and blood and semen and breast milk, and research has shown that even after someone survives and recovers from ebola, the disease can linger in their fluids for months. So if someone catches it, survives, and then breast-feeds their child, or kisses or has sex with their partner, or gets a cut and then someone else comes into contact with their blood, like a health worker, that can lead to the transmission of the disease, despite their having been well and seemingly fully recovered for weeks or months. That lingering contagiousness is a confounding factor with this disease, as it requires that people be very careful, even to an antisocial degree, and even well after it seems like that’s no longer necessary, because they feel good and healthy again. This also means that if someone dies of ebola, contact with their bodies can be incredibly dangerous. And past outbreaks have stemmed from or been further enflamed by locals wanting to perform community funerals and wakes, during which the body is often on display and touched by attendees, and that has led to further spread of the disease—which in many cases is difficult to tie back to that wake, because again, symptoms don’t arrive right away, and ebola symptoms are similar to what locals experience all the time from other afflictions, like colds and malaria. This past week, in Bunia, which is located in the Democratic Republic of the Congo, locals stormed a regional hospital in an attempt to recover the body of a beloved local figure who died of ebola. In the process, the hospital’s isolation ward, which was being used to keep ebola victims separate from everyone else, to keep the disease from spreading further, that ward was burned to the ground. There are no vaccines or treatments for the Bundibugyo Ebola species that is at the core of the outbreak, and the spread of misinformation in the area had locals believing that these health workers were trying to kill their patients, not save or isolate them so no one else caught ebola. The man at the center of this, who died five days after being admitted to the hospital, was thought, by his family, to have malaria, which is common in the area and has very similar symptoms, at least in the early days of an ebola infection. They demanded the hospital release his body so they could bury him, and the staff refused, saying doing so right now could lead to more ebola spread. The family gathered more locals, who threw stones at hospital workers, they broke through the gates of the hospital, police fired into the air to try to disperse the angry crowd, and the ebola ward caught fire during the melee. During that fire, five patients who were in the ward, all suspected of having ebola, fled, and they haven’t yet returned—so they are possibly out in the open, no longer isolated, suffering and maybe dying from their infection, and possibly spreading it to others, as well. There’s a lot going on in this story, and misinformation spread by local traditional healers who don’t like the hospitals and the medical workers who tell locals medical information rather than folk healing information are part of the problem, but the local medical establishment not doing a good job of educating locals about what they’re doing and why are arguably the flip side of that same coin; more investment in that kind of information dissemination by the government would go a long way to preventing this sort of thing in the future, and health workers globally could use more resources and overall infrastructure to help protect them while they’re carrying out their work. That said, this is just one small facet of what’s become a much larger story. As of the day I’m recording this, this new outbreak, which was first reported in the Ituri Province of the DRC, has caused 186 confirmed deaths, with 82 more confirmed cases and 836 suspected cases. As I mentioned, it’s caused by the Bundibugyo ebolavirus, which is less common, at least at this scale, and thus typical response efforts used against the more common Zaire ebolavirus, don’t seem to map onto this strain as well as was hoped, and the World Health Organization declared a Public Health Emergency of International Concern on May 16, as while this is unlikely to become as significant an issue as Covid-19 or other aerosol-spread infections on a global level, regionally it’s causing a lot of damage, and its nature, and the state of international aid for this sort of thing—which is currently substantially reduced, in part because of pullbacks on such programs by the current US administration—means it could continue to flare for several more months, before eventually starting to slow, killing many, many people, in any incredibly painful and contagious manner, in the process. This is the 17th ebola outbreak in the DRC since the disease was first recorded in the medical literature, and the third outbreak of this strain—the first of which was in the Bundibugyo District of Uganda in 2007 through 2008, that’s where it got its name, and then another in 2012 in the DRC. This isn’t the deadliest strain of ebola, only killing between 25 and 50% of those afflicted, but because of those aforementioned issues, plus it having flared in a region where governance is complicated by the presence of several militant groups, this wave of infections has created a broad and precarious situation; lots of people have been uprooted from their homes because of conflict between these militant groups and the government, and those refugees have been spreading ebola to other areas throughout the region, making contact tracing difficult or impossible, and leading to surges of new infections in neighboring, and a few further-flung, provinces. According to a predictive model of the outbreak published by the MRC Centre for Global Infectious Disease Analysis, the current number of infected people could actually be well over 1000, in part because of how difficult it’s been isolating the infected, and because the early symptoms are so similar to other common local afflictions; so people are less likely to visit hospitals and get an accurate diagnosis, because they assume it’s just a bout of something else, something less deadly and contagious. Getting resources into the area is becoming more difficult, too, as those militant groups are fairly active, one such group recently taking over a primary regional airport, which has disallowed the import of necessary medical equipment for regional hospitals. This hasn’t had much of an impact globally, yet, though cases have been documented in neighboring Uganda—a total of five confirmed infections, as of the day I’m recording this—and the World Cup team from

A portfolio of podcasts for curious, adventurous, passionate lifelong learners.