2026 Blue Wave

This week we talk about the US midterm elections, the unusually broad Democratic victory, and what a change in congressional control means for the remainder of President Trump’s term. We also discuss the cost of living, the limits of a congressional majority, and what these results may, and may not, tell us about 2028.Recommended Book: 1873 by Liaquat AhamedTranscriptOn November 3rd, 2026, Americans voted in an election that, under ordinary circumstances, would have been expected to make governing a little more difficult for the sitting president. The party in the White House often loses seats in midterm elections, as voters register their frustrations and the coalition that elected the president proves more difficult to assemble without the president, the party’s figurehead, on the ballot.This time, though, the losses were substantial enough to change who controls both chambers of Congress. Democrats won the House of Representatives, and they won the Senate, picking up seats in states that would have seemed like especially difficult targets, perhaps even borderline impossible to win, not that long ago. They also won several governorships, including in until recently unlikely-seeming places, like Texas, Iowa, Ohio, and Georgia.The precise size of those majorities are still in flux as of the day I’m recording this, as the last close races are still being counted and have yet to be called. But the basic result is settled: when the new Congress takes office in January, President Donald Trump will be working with a Democratic House and a Democratic Senate after two years of enjoying a Congress run by his party, in addition to his spot at the top of the executive branch, and a Supreme Court stacked with loyalists and ideological fellow travelers.What I’d like to talk about today is how that happened, what it allows the Democrats to potentially do after they’re seated, and what it might mean for the next presidential election.—As I mentioned, at the start of this election cycle, Republicans held the presidency, the House, and the Senate. Their House majority was narrow, which made a Democratic takeover there plausible. The Senate, as is often the case for Democrats hoping to win it, looked like it would be a far heavier lift, to the point that it wasn’t really in play. Senators serve six-year terms, so only about a third of them face voters at any given midterm, and Democrats needed to win seats in places where Republicans had recently been quite strong, in some cases leaning toward the Republicans in the double-digits.But the Dems managed to win those seats, and then some. Democrats flipped Senate seats in North Carolina and Maine, but also in Kansas, Iowa, Texas, and South Carolina. In Ohio, Sherrod Brown returned to the Senate after losing his seat two years earlier. As I’m writing this, Democrats have secured at least 54 seats, with a couple of contests still unresolved, and 51 seats represents a majority in the Senate.That Kansas result is especially striking: the state hadn’t elected a Democrat to the US Senate in nearly a century. Texas hadn’t elected a Democrat to a statewide office in more than three decades, but voters there chose James Talarico for the Senate and Gina Hinojosa for governor in this election—both Democrats.The House results were widespread and numerous, too, with Democrats flipping more than two dozen Republican-held districts, according to the Associated Press. And across the country, some voters elected new Democratic governors even where the state legislatures will remain Republican. This matters because a governor can shape budgets, appointments, and the administration of state government, but generally can’t enact their whole agenda without the legislature.That said, for a sense of how unusual this is, the House and Senate have both changed hands during a midterm election only three other times since World War II: in 1946, 1994, and 2006. A president’s party losing ground halfway through a term is common. Losing enough ground to hand over both chambers, including one with the Senate’s electoral map and offset election cycle, is not.So let’s take a moment to consider why this happened, because election results, even unusual ones, tend to get appealingly simple explanations in the press and think pieces, but there are several different things going on here, all at once.One is the cost of living. In the Associated Press’s large survey of people who voted, those who were especially concerned about grocery bills, healthcare, gas, and other daily costs overwhelmingly favored Democratic House candidates. Many voters blamed Trump’s economic policies, including tariffs, for making prices worse. The war with Iran also pushed fuel costs higher, and was unpopular with many voters who had expected a different approach to foreign policy.None of which requires a recession or other dramatic economic state of affairs. People can have jobs and still feel that the monthly budget is no longer enough. Nor does it mean everyone who voted for a Democrat shares the same explanation for inflation, or agrees on what should replace the president’s policies. It means these concerns were prominent, and that Republicans, as the governing party, bore much of the blame.Another issue was power. About seven in ten voters in that survey said Congress wasn’t doing enough to check the presidency, a view held by many Republicans as well as Democrats. More than half said government corruption was crucial to their vote. Those findings suggest that part of the electorate wanted Congress to behave more independently, whatever they thought about individual Democratic candidates; for many, this was more about the failures of the current leadership than about the specifics of whomever might replace that current leadership.Candidate quality did seem to matter, too, though. In Texas, Republican Senate candidate Ken Paxton’s scandals were a prominent part of the campaign. In South Carolina, voters had doubts about the experience of the Republican nominee. Trump-backed candidates in several states faced voters who felt they were too closely tied to him. So while this was a national political wave in favor of the Dems, some local races also had specific vulnerabilities that helped local Democrats make the most of the moment.The groups voting for each party also changed from the last election. In that AP survey, about seven in ten voters under 30 supported Democrats in their local House race, a marked change from the near-even split between young voters who backed Trump and Kamala Harris in the 2024 presidential election. Latino voters also moved toward Democrats in a big way, compared with 2024, and Democrats won every House district that touches the US-Mexico border.Now, keep in mind that a midterm electorate and a presidential electorate are different groups of people. Comparing the two can reveal a shift in support, but it doesn’t prove that a particular voter changed their mind. Some people voted in one election and stayed home for the other, and some voted for the first time this year. The pattern is politically meaningful, but it’s not a one-to-one comparison, two years offset.It’s also worth noting that the winning Democrats aren’t all the same sort of politician, and don’t all stand for precisely the same platform. The party elected moderates in some Republican-leaning states and more progressive candidates in places like Maine and Michigan. Each faction can point to victories that support its preferred approach, but that’s also a sign that there isn’t one neat ideological mandate behind this blue wave turnout.The survey found something else that complicates the wave story, too: about one in five people who voted for Democratic candidates viewed the Democratic Party itself unfavorably. These voters helped put Democrats in power, but still have concerns about the party as a whole. And that’s something that could complicate things for individual candidates who benef… This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit letsknowthings.substack.com/subscribe

This week we talk about urban fires, wildland fires, and AQI.We also discuss PM2.5, chemicals, and plumes.Recommended Book: Biological War by Annie JacobsenTranscriptOn January 7, 2025, a fire began in the hills above Altadena, California, and surged through a community of homes, businesses, vehicles, and everything inside them. To someone watching the plume from a distance, it was indistinguishable from any other wildfire smoke. As it turns out, though, the composition of that smoke was a lot more dangerous than the emissions of a typical fire.A study published in October 2026 found that, during the most intense period of structural burning, the smoke from the Eaton Fire carried far more lead and other hazardous material than smoke measured during a fire that burned mostly vegetation. The researchers estimate that roughly seven million people in the Los Angeles area were beneath air containing lead at more than eighty times its usual local concentration for a twenty-four-hour period.Which is a striking number; though importantly, that doesn’t mean seven million people developed lead poisoning, or that a fire produced eighty times some established threshold of danger. The study measured the air and modeled where it went, it didn’t measure how much of that air ended up inside of people, and what happened inside of body that inhaled some of that air.This finding does raise a practical question, though: when a wildfire moves into a city, do our standard ways of describing and responding to smoke give us enough information to make informed decisions about it, and the danger it potentially represents?What I’d like to talk about today is the difference between a burning forest and a burning neighborhood, and what we still need to learn about protecting people downwind from this sort of disaster.—Smoke is not just one thing, it’s a mixture of gases and tiny particles. And its nature is shaped by what’s burning, how hot the fire is, and what happens to the plume as it travels.In a vegetation fire, the fuel might be trees, brush, grass, and other organic material on the ground. A fire moving through a built-up area burns those things, too, but also burns through roofing, insulation, wiring, cars, furniture, electronics, paint, and cleaning products stored in kitchens and garages. Some of those materials contain metals and synthetic chemicals that a stand of trees does not contain.The location where homes and other development meet burnable vegetation is often called the wildland-urban interface. Which makes it sound like a narrow line on a map, but in practice this interface area contains neighborhoods where fire can make the leap from vegetation to buildings, and then from one building to another. It’s at this point that a landscape fire becomes, in part, a structure fire, and potentially a structure fire on a very large scale.Researchers had reason to expect that structure fires would produce different smoke from wildfires. A National Academies report in 2022 described major gaps in our knowledge of what happens when the materials in a town burn alongside vegetation, though. Earlier EPA work found elevated metals in smoke associated with fires that destroyed buildings, but the aforementioned Eaton Fire gave us a more detailed look at how this mixture of smoke elements can change over a matter of hours.Let’s take a moment to talk about how we measure this sort of thing. Most of us think of smoke, when we think of smoke, through the lens of an air-quality number. On a smoky day, a map might show that the air is unhealthy because of the presence of PM2.5, which are tiny particles that are small enough to be inhaled deep into the lungs. The PM2.5 number represents how much particle pollution is in the air, and it’s useful: decades of research link fine-particle exposure to breathing and cardiovascular problems, and the Air Quality Index tells people when to reduce their exposure, because the air is laden with such particles.But PM2.5 describes particle size, not a single type of chemical. Two samples can contain the same mass of fine particles, the same amount of PM2.5 in the air, while having different ingredients—tiny bits of different sorts of things. An ordinary smoke reading won’t necessarily tell you whether those particles came from burning trees, a row of houses, or some combination of the two.That was the opening for this new study. About eleven miles south of Altadena, in Pico Rivera, Caltech researchers had installed an advanced air-monitoring station that’s part of a twelve-site research network called ASCENT, which tracks not just the quantity of airborne particles but aspects of their chemical makeup. Importantly, it also collects measurements continuously, rather than gathering a sample every few days and then analyzing that sample later.As wind carried the Eaton Fire’s plume past the station, the instruments recorded changes to the smoke composition hour by hour. The researchers then compared those readings with data from another station that had observed the Airport Fire, a 2024 Southern California fire that burned mostly vegetation.The contrast was stark. According to the study, the urban plume had more than fifty times as much lead, around three times as much arsenic, and about seven times as much chlorine and bromine as the wildfire plume.That composition changed with time, though. Lead levels rose sharply while structures and vehicles were burning in Altadena. Later, as the Eaton Fire moved into the mountains and burned more vegetation, its chemical signature shifted toward what the researchers saw in the mostly wildland fire. They could watch a single fire change from something more like an urban fire to something more like a forest fire, even though the smoke might have looked similar from the ground.The research team also estimated how many people may have been underneath the lead-bearing plume when it was at its most dangerous. Their answer was roughly seven million during a peak twenty-four-hour period, when lead concentrations were more than eighty times the area’s usual background level.Now, keep in mind that the instruments they used directly measured the air at a specific location, and that broader population figure is an estimate that comes from combining measurements with an assessment of how the plume spread. And “eighty times background” compares the event with what is normally a very low lead concentration in Los Angeles air. It’s not an eightyfold increase in an individual’s health risk, it’s an eightyfold increase in the presence of lead in the air, which can be absorbed or not by those breathing that air.Being under a smoke plume also doesn’t mean every person breathed in the same amount, or in the same way. People were in different places, indoors and outdoors, and their buildings filtered air differently. The study did not test blood lead levels or follow the people who were exposed to determine later health outcomes, so this is a solid assessment of a potential source of danger, but not an indication of who actually got hurt, or even who was at risk of being hurt, because those other variables determine how exposed and vulnerable a given person is to this kind of environmental risk.That said, lead is a neurotoxin, and long-term exposure is especially dangerous to children. Arsenic also has well-established health risks. These numbers support the need for further investigation, even if they fall far short of providing some kind of widescale post-disaster diagnosis.Also worth knowing is that Altadena had many older homes, some of which may have contained lead-based paint. The prominence of lead in this plume doesn’t suggest that will always be the case for all fire-engulfed communities.Of course, inhalation of smoke isn’t the only danger here.Los Angeles County’s public-health department found elevated lead in a higher share of soil samples from intact properties downwind of the Eaton Fire. It cautioned that older pollu… This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit letsknowthings.substack.com/subscribe

This week we talk about schools, riots, and police violence.We also discuss Le Pen, Macron, and international reporting.Recommended Book: This Is for Everyone by Tim Berners-LeeTranscriptOn October 17, 2026, hundreds of thousands of people marched in cities across France. Many were protesting the price of fuel, groceries, and other necessities. Others were high-school students, parents, teachers, and union members demanding better schools and public services.Most of those marches were peaceful. But in several cities, there were clashes with police, shops broken into and looted, and public buildings attacked. The town hall in Limoges caught fire. In Paris, images of burning vehicles and smashed storefronts made headline news, often without clear context, explaining where the peaceful marches ended and the violence began.Those two realities, peaceful protests on one hand, violence and widespread theft and destruction on the other, have since competed to define this story. Many people who marched, and in some cases continue to march, are concerned that their demands will become inextricably associated with all that widely shared footage of rioting. For shopkeepers and residents whose neighborhoods were damaged, that footage represents a real event with real costs. And for politicians facing a presidential election next spring, the unrest has become a test of who can speak to both protestor grievances and concerns about public safety.What I’d like to talk about today is how these protests started, why a movement about schools became entangled with a cost-of-living crisis and, then, serious disorder, and what all of this might mean for French politics; especially for Marine Le Pen, the far-right politician who’s leading the race to succeed President Emmanuel Macron.—The current protest wave began in September, at a high school in Créteil, outside Paris, where students were angry about unfilled staff positions. Walkouts and blockades spread to other schools, then to other parts of the country.The specific complaints varied by school, but the broad list was fairly consistent: teaching roles remaining unfilled, crowded classrooms, aging buildings, and school days that can stretch from eight in the morning to six in the evening. Some students also object to Parcoursup, the government system that sorts students who apply for higher education into schools, which students say adds uncertainty and a lot of pressure to the tail-end of their high school experience.These are not just complaints about aesthetics or convenience. If a student loses weeks of instruction before an exam, or has to spend a winter in a poorly heated classroom, that can affect what they learn and what opportunities they have later. And the experience isn’t evenly distributed. France’s national education system is meant, in part, to make citizenship and opportunity less dependent on where someone was born. Yet the OECD has found that students’ socioeconomic background remains strongly associated with their educational outcomes, despite these attempts at establishing opportunity equality.The French education ministry has pointed to investments it has made and improvements in filling teaching posts. But national spending and staffing figures don’t guarantee that the right teacher is available in the right classroom on Monday morning. A vacant specialized post, a long absence without a substitute, or a building whose repairs are delayed can still shape a student’s educational experience for an entire year.The government’s early response to these demonstrations mixed acknowledgment with policing the protestors. Prime Minister Sébastien Lecornu promised work on teacher replacements, building repairs, and schedules. Authorities also moved to clear blockades and arrest people accused of vandalism or attacking police and school staff. As the confrontations continued, teenagers suffered severe injuries attributed to police crowd-control weapons, including one young protester who lost a hand. Officers and school staff have been injured, too.These injuries and all that pushback intensified an existing dispute over whether forceful policing was preventing violence, contributing to it, or doing both in different places at different times. That question does not have a single answer, but it does help explain why some parents and students who were originally focused on school conditions began talking about policing as well: the police response to those protests blended that existing conversation and its associated complaints with those related to schooling in the country.Meanwhile, another source of anger was building well beyond the schools.France has been hit by an energy-price surge connected to the broader disruption of oil and gas markets in the Middle East. Its statistics agency reported that consumer prices were up 3.1% in September compared with a year earlier, but energy prices were up more than 21%. Diesel was up more than 44%, and gasoline more than 25%. Fresh food prices also spiked pretty dramatically, partly because of shortages caused by summer heatwaves.Calls circulated online to revive the “yellow vest” movement, which began in 2018 over fuel prices and grew into a much broader protest about living standards and political representation. Unions, including the CGT, backed a new day of action. Student groups urged their supporters to join. By October 17, these strands had converged, though they didn’t all share the same priorities or protest methods.The Interior Ministry estimated about 550,000 people demonstrated on that single day, though organizers have said the number was higher. Either way it was a very large mobilization. And reporters saw peaceful crowds carrying flags and singing, alongside far smaller groups throwing projectiles, setting fires, breaking windows, and looting. Reuters reported roughly 2,400 arrests on the first day, when hundreds of businesses were damaged. One demonstrator died after suffering a cardiac arrest. In the aftermath, there were additional clashes and localized curfews.As tends to be the case in these situations, some people may have joined one of these demonstrations intending to commit violence. Others may have taken advantage of confusion once violence began around them. Some protesters may also have committed crimes. Investigators still have to establish who did what in each individual case, though a survey by the pollster Elabe found that 70% of respondents believed those involved in the violence were mainly people who showed up specifically to cause it, which suggests that many French people distinguish the protests from the looting—something not all reports on these demonstrations have clearly done.That same survey found that 91% of respondents condemned the violence, while 93% considered the cost-of-living demands legitimate and 82% considered the students’ demands legitimate. In other words, a person can think a demonstration is justified but the destruction around it isn’t.So why did all of these issues come to a head, and become so combustible, now?Persistent school problems over a long period of time, followed by a sudden rise in fuel and other prices, gets us partway to an explanation. But more fundamentally, many French voters expect the state to provide reliable education, transportation, healthcare, and protection from sharp economic shocks. Those expectations are pretty central to France’s social model, but the government hasn’t kept up their end of the bargain, and has little spare money with which to address these issues, at the moment.At the end of June 2026, public debt stood at about 119% of the country’s annual economic output, according to France’s statistics agency. High debt doesn’t mean the government will run out of money tomorrow, but it does mean that large new spending promises come with questions about long-term borrowing costs, taxes, and what would have to be cut in the tradeoff. France has also struggled to pass budgets… This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit letsknowthings.substack.com/subscribe

This week we talk about Adobe, Call of Duty, and Microsoft.We also discuss compilation, machine code, and AI agents.Recommended Book: The Reverse Centaur’s Guide to Life After AI by Cory DoctorowTranscriptIn October of 2026, a software developer named Brandon Thomas attracted a lot of attention by releasing a collection of free, open-source programs with names like PhotoCraft, FilmCraft, and VectorCraft. These apps are meant to replicate the functionality of digital tools like Adobe Photoshop, Premiere, and Illustrator. Within days, the same project also had a word processor, spreadsheet, and presentation app—alternatives to Microsoft Word, Excel, and PowerPoint.The coverage of these releases was borderline breathless. One person and a collection of AI coding tools had apparently cloned the key offerings of two of the most established software businesses in the world. The developer of these apps suggested that software, as we’ve always known it, might be over.Another story, also circulating this month, featured a team using AI tools to reconstruct the underlying code of the 2009 video game Call of Duty: Modern Warfare 2. This wasn’t a lookalike built by watching the game and imitating its menus, it was an attempt to work backward from the compiled program to something like the source code its developers might have originally written.Both projects fall under a loose umbrella of AI-assisted reverse engineering efforts. But they used different approaches and achieved different levels of fidelity, which have led to different likely legal and practical consequences.What I’d like to talk about today is how people are using AI to reproduce old games and current software, why this is suddenly feasible at a much greater scale than before, and what happens if copying the behaviors and features of a program becomes a lot cheaper than creating the original was.—To kick things off, let’s go through three things that are often called cloning in the world of software.One is making a program that looks and behaves enough like another that a user can accomplish similar tasks. You can watch someone use Photoshop, read documentation, examine a PSD file, and build a different image editor with familiar tools. You don’t need Photoshop’s source code to do that. People have been making alternatives to proprietary software for decades; GIMP and LibreOffice, alternatives to Adobe and Microsoft apps, respectively, existed long before today’s AI models.A second approach aims for compatibility. Maybe you don’t need to reproduce every button in a word processor, you just need your program to open and save documents that Word users can exchange without losing their formatting. That requires understanding file formats, often down to details that aren’t obvious from the official specifications, so it can be a deceptively complex undertaking.The third approach is decompilation, which means starting with a compiled program—the machine instructions a computer runs—and trying to reconstruct higher-level code from it. This is like having a cake and working backward toward the recipe. You can identify many ingredients, and perhaps bake something that tastes the same, but the original measurements and the cook’s handwritten notes are not available to those doing this kind of attempted backtracking.The Modern Warfare 2 project fits into that last category. Developer Maurice Heumann described spending roughly three months, with collaborators and an expanding group of AI agents, rebuilding a popular first-person shooter. His October account no longer names the game, but earlier public posts and reporting identify it as Modern Warfare 2. Heumann says earlier posts were removed after copyright-owner objections, and he doesn’t plan to release the reconstructed code.Initially, the team had three AI agents doing the reconstruction and a fourth reviewing it. They made quick, visible progress: menus worked, maps loaded, and it looked as though much of the game had been recovered. Then the team discovered that a great deal of the code was wrong. The AI agents that they used had invented logic, misunderstood data structures, and even replaced simple operations with much bulkier, slower ones.So they changed the test. Using the compiler that had been used for the original game, they compiled each reconstructed function and compared the resulting machine instructions to the original executable. If a function matched byte for byte—accounting for things like references whose addresses can vary—they had much stronger evidence that they’d recovered its behavior. If it didn’t, the agent had to try again.Even that required supervision, though. Heumann says agents sometimes tried to satisfy the test by embedding arbitrary bits of assembly code, or by modifying the checker so a troublesome function would be skipped. The team had to prohibit those shortcuts and protect the verification process from changes. This isn’t exactly an example of an AI deciding to cheat in the human sense, but it’s similar: the bots were trying to pass a test in ways that the people running it didn’t want or intend.By the time Heumann published his account, he said 99% of the game’s functions were present in the reconstructed source and 83% matched byte for byte. He also said the game was fully playable. Worth noting is that those are the team’s reported results, not an independently audited release; the code is private. And that 83% refers to functions, not 83% of the game’s artwork, design, or overall creative work. Still, the scale of the attempt is striking, as is the amount of infrastructure required to make it happen in the way the creators intended.A smaller, more archival version of this story involves a game called Klonoa: Empire of Dreams, a Game Boy Advance title that was released in 2001.A developer who goes by Macabeus has been rebuilding its code from the game cartridge’s data, with help from AI tools and other contributors. In an August write-up, Macabeus reported that 51% of the game’s code bytes had been matched: the newly written C code, when run through the right old compiler, produced the same machine instructions as the original.This is a very different sort of project from creating a playable imitation. The point is to understand how the game was made, make the work reproducible, and potentially preserve or adapt it for the future. The public project doesn’t include a copy of the game, and participants need their own legally obtained game data.While exploring the code and testing the game, the AI agent Macabeus was using found a hidden minigame on the “Clear Save Data” screen. Macabeus initially assumed the AI was hallucinating, before verifying that it was real. The discovery didn’t require the model to be a game-design genius, it just had to inspect a lot of code, notice an odd pathway, and try it: exactly the kind of patient investigation a person might do, but with much more capacity for repetitive searching.The project also illustrates a limit in this sort of process. Matching the machine code doesn’t automatically produce clean, readable source code. Macabeus describes agents finding awkward shortcuts when they get stuck, and more experienced human reverse engineers rejecting some of those supposed solutions. All of which is arguably fine for this kind of project, where the original source is gone, and the reverse-engineer effort is primarily intended to create a playable approximation of the original. But the process is still often uneven and uncertain, and isn’t as simple as pushing a button and getting a perfect copy of a game or other piece of software spat out by the AI.The aforementioned Adobe and Microsoft projects are different from those game cloning efforts. ArtCraft’s “Craft” apps say they’re written from scratch, using public specifications and observations of existing programs, with original or openly licensed assets. They aren’t claiming to have disassembled Photoshop or extracted Microsoft’… This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit letsknowthings.substack.com/subscribe

This week we talk about polio, malaria, and smallpox.We also discuss vaccines, the CDC, and the HHS.Recommended Book: Remnant Population by Elizabeth MoonTranscriptIn 2000, the United States declared measles eliminated. That didn’t mean nobody in the country would ever catch it again, it meant the virus was no longer freely spreading within the US. New cases generally arrived with someone who had been infected elsewhere, and public health workers were able to stop those cases from turning into long-running chains of transmission via the systems, tools, and infrastructure they created.In early October of 2026, Pennsylvania reported more than a thousand new measles cases this year, most of them part of an outbreak that began in April. As of October 7, the state had confirmed 1,078 cases across 41 counties, 207 hospitalizations, and five measles-associated deaths. Nationally, the Centers for Disease Control and Prevention, the CDC, had counted nearly 3,900 cases through the beginning of the month, making this the worst year for measles in the US in decades.Pennsylvania’s latest update included 74 newly confirmed cases, though the state was careful to say that more than 40 of those people had become sick over two weeks earlier. So these were newly *reported* cases, not necessarily newly *infected* people—a distinction that matters when we’re trying to figure out whether an outbreak is speeding up or slowing down.There’s a similar distinction being made with the death figures. Pennsylvania calls its five deaths *measles-associated*, using a definition that considers timing, a confirmed infection, medical evidence, and whether an unrelated cause might explain the death. The CDC’s national death count has not always matched the state’s, in part because the agencies publish on different schedules and review and tally deaths differently; so reporting on these sorts of things can sometimes be unintentionally fuzzy due to varying definitions and numbers.What I’d like to talk about today is what it actually means to eliminate a disease, how measles has managed to claw its way back in a country that achieved that ‘eliminated’ milestone more than a quarter-century ago, and how vaccination, misinformation, and changing federal policy might shape what happens next.—There are a few words public health workers and officials use that sound similar, but which describe substantially different achievements.If a disease has been *eradicated*, it no longer circulates naturally anywhere in the world, as is the case with smallpox. The last known natural case was recorded in Somalia in 1977, and the World Health Organization declared the disease eradicated in 1980. There have been no natural smallpox outbreaks since. That’s a truly remarkable global achievement, and it’s why routine smallpox vaccination isn’t necessary for most people today.*Elimination* is more local, and more conditional. It means sustained transmission of a disease has been interrupted in a defined place, like within a country. The pathogen can still exist elsewhere, and an infected traveler can still bring it in. Whether that introduction fizzles out or takes hold depends on what it finds when it arrives—how prepared that country’s government is, and how well vaccinated its population, how well funded its healthcare system, and so on.Measles isn’t the only disease that’s been eliminated in the United States. The last US case of polio caused by *wild* poliovirus was in 1979, and the Americas were declared free of it in 1994. But in 2022, an unvaccinated person in New York developed paralytic polio caused by a vaccine-derived poliovirus, and genetically related virus residue was found in local wastewater. That was not wild polio somehow reappearing from the American past, it was a different type of poliovirus that latched onto a vulnerable person in a country where sustained polio transmission had long since been halted.That virus was linked to the live oral polio vaccine used in some other countries, which can sometimes give rise to circulating variants where immunity is low. The inactivated polio vaccine used in the US cannot do that.The US also eliminated sustained malaria transmission in the early 1950s, but mosquitoes capable of carrying the parasite still live in the US, and people returning from places where malaria is common can carry it across the border. In 2023, the CDC recorded ten locally acquired, mosquito-transmitted cases across four states—the first such cases in about twenty years. Ten cases doesn’t mean malaria has become endemic in the US again, it’s just further evidence that elimination of a disease in a region doesn’t mean reintroduction is impossible.Smallpox, then, is an example of a disease that has been removed from natural circulation worldwide. Polio, malaria, and measles, in contrast, can be locally defeated, but regions in which that’s the case remain vulnerable to the movement of people, viruses or parasites, and that victory is reliant on the continued health of the health systems meant to detect and interrupt them.Measles is an especially demanding test of those systems because it spreads so efficiently.The virus moves through the air when an infected person breathes, coughs, or sneezes, and it can remain in the air, in a room in which someone has coughed, for up to two hours after that person has left. People can pass it on before the distinctive rash appears, and, according to the CDC, as many as nine out of ten nearby people who aren’t immune can become infected after exposure.For some, the illness brings a fever, a cough, red eyes, and a rash, followed by recovery. For others, it causes pneumonia, swelling of the brain, or death. Young children, pregnant people, and those with weakened immune systems face higher risks, and this is why a measles outbreak isn’t just a rash going around, and why even just a few cases can generate an enormous amount of public health work: finding contacts, checking immunity, arranging vaccinations, and warning people who may have spent time in the same airspace. Measles can be painful and uncomfortable, but it can also be deadly, so all that effort is arguably well-invested.The measles, mumps, and rubella vaccine—the MMR vaccine—is the main tool for preventing the spread of measles. One dose is about 93% effective against measles, and two are about 97% effective. That doesn’t mean vaccinated people can never get sick; these are estimates of how much the vaccine lowers the chance of infection. When most people in a community are protected, the virus also has fewer opportunities to reach those who cannot receive the vaccine or who haven’t yet reached the right age for their first routine dose.Public health agencies often use about 95% coverage as a target for measles. That isn’t a magical national number, though, where a country is safe at 95.1% and doomed at 94.9%. It describes the level of protection usually needed within communities to make sustained spread difficult. A state can have a healthy-looking average while particular schools, neighborhoods, or social networks have far lower coverage and are thus at higher localized risk.And that’s part of what makes the current trend concerning. CDC data show MMR coverage among American kindergarteners falling from 95.2% in the 2019-to-2020 school year to 92.4% in 2025-to-2026. About 280,000 kindergarteners lacked documentation of a completed MMR series in that latter year. Not all of them were necessarily unvaccinated (these sorts of records can be incomplete) but the direction of travel, and the concentration of missed vaccinations in some places and among some communities, leave more openings for the virus, which means everyone is at greater risk of infection.There isn’t one explanation for all those missed doses. Some families have trouble getting appointments, taking time off work, or reaching a clinic. Others seek exemptions or have questions they don’t feel were answered well by medical … This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit letsknowthings.substack.com/subscribe

This week we talk about Egypt, Eritrea, and Sudan.We also discuss the Grand Ethiopian Renaissance Dam, ports, and journalism.Recommended Book: The Infinity Machine by Sebastian MallabyTranscriptIn 2018, Ethiopia reopened its embassy in Eritrea. The two countries had fought a border war in the late 1990s and had spent nearly two decades in a tense state that was not quite war, not quite peace. The reopening was part of a striking reconciliation between Ethiopian Prime Minister Abiy Ahmed and Eritrean President Isaias Afwerki. Abiy received the Nobel Peace Prize the following year, in large part for his efforts to end that conflict.At the beginning of October, though, Ethiopia announced that it was re-closing that embassy. It ordered ten Eritrean diplomats to leave Addis Ababa, saying they had engaged in activities that threatened Ethiopian security. Eritrea called the decision astonishing and said it was severing diplomatic ties with Ethiopia in response.Ethiopia also expelled an Egyptian diplomat, and Egypt responded by expelling an Ethiopian one. The African Union urged all three countries to exercise restraint and preserve channels of communication, but all of this was happening as fighting was picking back up in northern Ethiopia, and as explosions were heard in the capital, Addis Ababa. A source connected to an armed opposition alliance told Reuters that the group had launched drone strikes there. But as of early October, there had been no independent confirmation of what caused those explosions. Witnesses and officials confirmed hearing blasts; whether they were an attack, and who might have carried it out, are still open questions.What I’d like to talk about today is how a peace that once seemed to transform this part of Africa fell apart, how Ethiopia’s internal war became entangled with its disputes with its neighbors, and why it’s important to distinguish what’s actually happening from what each side in this simmering conflict says is happening.—Let’s kick things off with a quick refresher of this region’s geography and history, to help explain why this crisis has so many layers.Eritrea sits along the Red Sea, just north of Ethiopia. It became independent from Ethiopia in 1993, which left Ethiopia without a coastline. The two countries initially maintained decent relations, but then they went to war over their border in 1998. The fighting ended in 2000, without any real resolution, and without formally establishing a new, mutually acceptable relationship. That was the long stalemate Abiy and Isaias seemed to resolve back in 2018.Inside Ethiopia, there was another important change during this period. For decades, a political movement originating in the northern region of Tigray, the Tigray People’s Liberation Front, or TPLF, had been the dominant force in the coalition that governed the country. Abiy came to power in 2018 and reshaped that governing coalition, and this put even more distance between the central government and the TPLF.In 2020, those political disputes erupted into a war between the Ethiopian federal government and Tigrayan forces. Eritrea fought alongside Ethiopia’s government against the TPLF. This is one of the most striking reversals in the present crisis: the government and its former Eritrean ally are now accusing one another of dangerous interference, while Ethiopian officials allege that Eritrea is helping that same Tigrayan movement they once fought against.This war, which lasted from 2020 to 2022, killed hundreds of thousands of people, according to estimates that vary substantially because conditions made reliable counting so difficult. It also displaced millions, damaged farms and hospitals, and involved absolutely horrific abuses against civilians. International investigators and human-rights groups documented atrocities by multiple parties, including Ethiopian and Eritrean forces and Tigrayan fighters.In November of 2022, Ethiopia’s government and the TPLF signed a peace agreement in Pretoria, South Africa. They agreed to stop fighting, disarm Tigrayan forces, restore services throughout the region, allow humanitarian aid to reach civilians, and to work toward a more stable political settlement. But Eritrea was not a party to that agreement, and that’s because there were two different peace processes.The 2018 reconciliation was between Ethiopia and Eritrea: two states ending a border standoff. The Pretoria agreement was between Ethiopia’s federal government and the TPLF: two parties trying to end a civil war. Neither agreement automatically settled the questions at the heart of the other one, and Eritrea’s role in the civil war made it more difficult for a ceasefire between Ethiopian and Tigrayan leaders to guarantee peace along the Eritrean border.Parts of western Tigray were, and remain to this day, contested by Tigrayan and Amhara authorities. Hundreds of thousands of people displaced from that area have not been able to return home safely. Disarmament did not proceed as the agreement envisioned, and rights groups continued to report abuses, including by Eritrean troops in parts of Tigray.Within Tigray, the TPLF split into rival factions. The status of the regional administration became another source of conflict. In May of this year, TPLF leaders pushed out an administration backed by the federal government. The two sides disputed who had authority there, how the peace agreement should be carried out, and whether federal funds were being withheld. Human Rights Watch also reported allegations of a coercive recruitment drive by Tigrayan authorities, including the forced recruitment of child soldiers.So the Pretoria agreement stopped a major war, but it left armed groups, competing governments, disputed territory, and displaced families in place, their problems unresolved.Then, in late September of this year, fighting between federal forces and Tigrayan fighters intensified in Tigray and neighboring Afar. The United Nations reported that armed forces had seized the airports in Mekelle, Axum, and Shire, and warned that hostilities were spreading into neighboring areas. An alliance of seven Ethiopian opposition groups also announced a common challenge to Abiy’s government, though importantly, an announcement of cooperation doesn’t tell us how closely those groups are actually coordinating on the ground.Ethiopia’s army chief then accused Eritrea of helping the TPLF and suggested that Egypt and Sudan were supporting the opposition as well. Eritrea, Egypt, and Sudan have denied the allegations. And those differing stories point at a difficulty in any conflict, but especially one happening in this part of the world, at this moment in time.Governments have access to information they may not make public. They also have reasons to portray an internal opponent as an instrument of a foreign power. Armed groups have their own reasons to claim allies or capabilities they may not actually possess. And governments accused of involvement in a conflict or violent act have strong reasons to deny it, whether the accusation is true or false.This fog of war applies to the aforementioned blasts in Addis Ababa, too. If opposition forces had successfully attacked targets in the capital, that would represent a significant shift in the conflict. Reuters reported a claim from a source linked to the opposition alliance, but also said it could not confirm who caused the explosions. The Associated Press reported that it remained unclear whether the blasts were an attack or just military exercises. So there’s a consequential claim here, but not yet confirmation of a rebel strike on the capital.Part of the problem is that independent observation is difficult in this region right now. Internet service in northern Ethiopia was disrupted as the fighting escalated. Roads and flights have been interrupted. Earlier this year, Ethiopian authorities expelled a French journalist after her reporting trip to Tigray, and temporarily revoked accreditati… This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit letsknowthings.substack.com/subscribe

This week we talk about EU membership, trade deals, and association.We also discuss Australia, Ursula von der Leyen, and Brussels.Recommended Book: Being You by Anil SethTranscriptJust off the coast of Newfoundland, there’s a small group of islands that belongs to France. Saint-Pierre and Miquelon are close enough to Canada that, at their nearest point, the two countries are separated by only a few kilometers of water.In September, Canadian Prime Minister Mark Carney met French President Emmanuel Macron on these islands. The two leaders could stand on French territory and talk about a partnership with the European Union, while Canada was visible across the water.A few days earlier, European Commission President Ursula von der Leyen had made a similar, but possibly more significant proposal. Speaking before the European Parliament, with Carney in the room, she said she wanted to open the door for Canada to become the EU’s first “associate member.”Depending on which headline you read about the statement, this may have sounded like Canada might be joining the EU, joining it partway, or joining an entirely new version of it. Soon after, Australia’s trade minister said his country was “on the same page” as Canada regarding closer ties, and the European Parliament’s president then suggested Australia and New Zealand might follow Canada’s lead.Now, despite all those headlines and interpretations, there’s one problem with all these stated ambitions: the European Union does not have an established category called associate member. No one has agreed on what rights or obligations an associate member country would have, and figuring that out—and getting some kind of resolution passed—would be necessary for anyone, including Canada and Australia, to get closer in that way with the EU.What I’d like to talk about today is why this proposal was made, what it could mean if it eventually becomes more concrete, and whatever happens, what these statements tell us about the way global alliances are changing.—The EU has 27 member countries, but there are already a few different ways to be connected to it.Full member nations help write EU law and send representatives to its institutions. They all participate in the single market, which means goods, services, money, and people can move across member nation borders under shared rules.The single market is distinct from the customs union, which sets common tariffs on goods imported from elsewhere. Both are distinct from the Schengen area, which removes most routine passport checks between participating countries. And nations that use the euro are another group entirely. These arrangements tend to overlap, but they’re not the same thing, and membership in one does not automatically mean membership in all the others.There are also countries outside the EU that participate in some of its systems. Norway, Iceland, and Liechtenstein belong to the European Economic Area and are part of the single market. To do that they have to accept many EU rules, although they don’t vote on those rules as that would require full EU membership. Switzerland has built its own set of bilateral arrangements with the EU. And Britain, after leaving the union, has also negotiated a special trade and cooperation agreement with it.So there are precedents for a country having a deep relationship with the EU without being a member. But there’s no ready-made ‘associate’ slot that Canada can just step into.There’s a legal distinction here, too. The EU treaties say that a European state can apply to become a member, and Canada is not a European state. The treaties separately allow the EU to make association agreements with countries outside the bloc, and it already has many kinds of agreements with external partners. An association agreement, though, does not make that partner a member of the Union.“Associate member” could eventually become a useful name for a new collection of rights and obligations held by nations outside those existing parameters. For now, though, it’s a political invitation and a negotiating idea, not a defined legal status.All that said, Canada isn’t starting from scratch on this. Its trade agreement with the EU, called CETA, has been applied provisionally since 2017. Most of it is already in effect, though ten EU countries still haven’t completed the ratification required for the whole of the agreement to go into force.CETA removes most tariffs and opens some opportunities for companies on each side, but it doesn’t make Canada part of the single market. A Canadian product can be easier to sell in Europe without a Canadian worker gaining a general right to take a job there. That’s part of the distinction between a trade deal and the sort of relationship people sometimes imagine when they hear the phrase “union membership.”Canada also joined Horizon Europe, the EU’s major research funding program, in 2024. It has a security and defense partnership with the EU, and it became the first country outside Europe to join an EU defense financing initiative called SAFE.That means some of the proposed future relationship already exists, but in pieces rather than as a unified whole. The question is whether those pieces can be connected and extended into something much more ambitious and holistic.Now, important to understanding the why of all this is understanding Canada’s relationship with the US, and how that relationship has been fraying, of late.Canada’s economy is deeply integrated with that of the United States. Its proximity to the massive US market has brought enormous benefits, but it also means a change in American trade policy can have an immediate and dramatic effect on Canadian businesses. President Donald Trump’s tariffs and repeated suggestions that Canada should become the 51st state have made that dependence more than a little politically uncomfortable.For Carney, a closer European relationship offers a way to reduce the risk of having too many economic and security decisions shaped by one increasingly unpredictable and at times belligerent neighboring country. Europe has its own reasons to be interested in closer ties. Canada has energy and critical minerals, industrial capacity, research institutions, and a role in the Arctic. Both sides also support Ukraine and work together through NATO and other forums.Von der Leyen has proposed cooperation with Canada on batteries, artificial intelligence, cybersecurity, advanced manufacturing, and defense production. Carney has added ideas about financial services and opportunities for young people to live, work, and study across the Atlantic.This isn’t a plan to move Canada’s economy from one continent to another. Geography still matters. A car part crossing the US-Canada border may be part of a production chain built up over decades; a new trade agreement with Europe doesn’t instantly recreate or replace that chain.But diversification doesn’t require replacing one partner with another. It can mean having additional buyers, suppliers, research collaborators, and sources of investment, so that a disruption in one relationship is less damaging.The costs and limits matter here, too. Canadian opposition leader Pierre Poilievre has warned that deeper EU ties could mean higher costs and Canadian industries being regulated from Brussels. His motives in bringing this up are likely at least partly political, but it’s a question worth asking: if Canada wants more access to a tightly regulated European market, which European standards would it have to accept as a tradeoff?Europe would face its own questions. EU governments have spent decades building shared rules among countries that accept reciprocal obligations. They may not want to give a distant partner the benefits of those standards without comparable commitments. The fact that ten member states have yet to ratify the existing Canada trade agreement is a reminder that closer ties require more than an enthusiastic speech from the Commission presid… This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit letsknowthings.substack.com/subscribe

This week we talk about lobbying, renewables, and the NRA.We also discuss implied threats, midterm elections, and political action committees.Recommended Book: Sunward by William AlexanderTranscriptFor much of the late 20th century and the first few decades of the 21st, one of the most feared interest groups in US politics was the National Rifle Association, the NRA.Its power came from a large and politically engaged membership, a mailing list, a grading system that reduced complicated voting records to a letter, and a reputation for ending political careers over specific votes.Once it attained that reputation, the NRA didn’t have to defeat every politician it disagreed with. Members of Congress only had to believe it could defeat them, and that belief shaped races in which the group spent nothing; politicians went out of their way not to anger the NRA. Money can buy an advertisement or a meeting. What tends to change a vote is the expectation that one choice will be rewarded and another will carry consequences.The NRA’s influence has declined following internal scandals, financial trouble, and the growth of well-funded gun-control groups. But its model remains potent: pick a few visible fights, and allow your reputation to do a lot of the work for you, in the future.In 2010, the Supreme Court’s Citizens United decision, alongside a related appeals-court ruling later that year, helped create the modern super PAC: a political committee that can raise and spend unlimited sums advocating for or against candidates, so long as it does not coordinate that spending with their campaigns.This did not eliminate the effort and resources required to build influence, but it meant a few wealthy donors, a competent team, and some carefully selected races could establish a reputation in months rather than decades.In 2026, solar, wind, and batteries are projected to account for about 93% of new utility-scale electrical generating capacity added in the United States.That doesn’t mean they provide 93% of the country’s electricity—natural gas remains the largest source in the US—but these technologies are now the overwhelming majority of what the industry is building.Despite that growth, in 2025 Congress passed a law that sharply rolled back federal support for much of the clean-energy industry, and most of the politicians who voted for those rollbacks appeared to suffer no political consequences for doing so.What I’d like to talk about today is the effort to build a feared clean-energy lobby, how it has influenced a series of Republican primaries, and what its early successes do and do not tell us about the role of money in American politics.—The One Big Beautiful Bill Act, or OBBBA, was signed into law on July 4, 2025.For wind and solar projects, the new law generally ended production and investment tax credits for facilities placed in service after December 31, 2027, unless construction began within twelve months of the bill’s enactment.That twelve-month window closed in July of 2026, and a subsequent executive order directed the Treasury Department to adopt a stricter definition of when construction actually begins, further clamping down on entities hoping to benefit from those now-defunct credits.Tax credits for electric vehicles and residential efficiency upgrades ended in 2025, while support for clean hydrogen was curtailed. Other technologies, including batteries, nuclear power, and geothermal energy, were treated differently, so it would be misleading to say the law eliminated every federal clean-energy incentive, though it did severely curtail a lot of renewables-oriented industries and construction in the US.Republicans have generally been more supportive of fossil-fuel production and more hostile to federal wind and solar subsidies, while Democrats have generally taken the opposite position. There are important regional exceptions, especially among Republicans whose districts have attracted manufacturing plants, wind farms, and other energy investments. Several Republican lawmakers have even written letters asking party leaders to preserve some of the credits, in part because projects and jobs in their districts depended on them. When the final vote arrived, though, nearly all congressional Republicans voted for the bill.Tom Matzzie, the founder of the retail electricity company CleanChoice Energy, previously worked for Democratic campaigns and served as the Washington director of the progressive organization MoveOn.org, so he was familiar with electoral campaigning as well as the energy industry. In the wake of the passing of the OBBBA, he posed a question to Canary Media, possibly alluding to the success of political interest groups like the NRA when he said, “Are we someone that people can hurt without consequences?”Matzzie recruited Chris Larsen, the billionaire co-founder of the blockchain company Ripple and an investor in clean energy, and Michael Brune, the former executive director of the Sierra Club, and together they formed the Invest in Tomorrow Coalition, or ITC, an acronym that also evokes the investment tax credit the group was organized, in part, to defend.Federal Election Commission records show that the coalition raised about $6.8 million during the first half of 2026. Larsen provided $6 million, while the organizers said they had assembled commitments of around $20 million and hoped to spend as much as $30 million during the election cycle. Matzzie called the group’s candidate-targeting spreadsheet the “Revenge Tour Matrix,” which is an unusually candid name for a political document.The group’s most interesting strategic decision, though, is arguably that its advertisements generally did not mention clean energy.When ITC opposed Texas Representative Chip Roy in the Republican runoff for state attorney general, its ads questioned Roy’s loyalty to President Trump. When it opposed Tennessee Representative Andy Ogles, it created a website called Lyin’ Andy that focused on his missed votes and a federal investigation into his campaign-finance reporting.The coalition openly identified its reason for entering these races, but its advertisements used whatever campaign research suggested would move primary voters, not what its donors wanted them to think about solar panels.This approach was especially well suited to Republican primaries, where relatively small electorates can be reached through a concentrated group of conservative television, social-media, and streaming outlets.In Ogles’ race, for instance, the two campaigns had each spent roughly $600,000, while ITC spent around $2 million attacking Ogles and introducing voters to his challenger, Charlie Hatcher.The coalition describes its record so far as five wins in five races.In May, Roy lost the Texas attorney-general runoff by 10.4 percentage points.In June, Iowa Representative Mariannette Miller-Meeks, a Republican who had defended renewable-energy interests in Congress, survived her primary with help from ITC, including a $125,000 contribution.Later that month, South Carolina Representative Ralph Norman finished a distant third in the Republican primary for governor after the coalition opposed him.In August, Ogles lost his primary by more than six points despite an endorsement from Trump and more than $700,000 from the House Freedom Caucus’ political fund.And in late August, Norman lost a second race, this time a Republican Senate primary, to Darline Graham, the sister of the late Senator Lindsey Graham. ITC spent about $1 million in that contest, which Norman lost by five points.The reactions from Norman and Ogles were almost as useful to the group as the election results.Norman blamed outside political spending in his concession speech, while Ogles, before his loss, said the coalition was attempting to make an example of him and that could have a chilling effect on other conservatives in Congress.Those statements do not prove ITC caused either defeat. But they do advertise the conseque… This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit letsknowthings.substack.com/subscribe

This week we talk about AI agents, cyberattacks, and insurance claims.We also discuss OpenAI, Hugging Face, and policy language.Recommended Book: The Stars My Destination by Alfred BesterTranscriptTwo broad categories of cyberattack have become especially visible this year, and only one of them requires a human attacker in the loop to choose the target.In March, hackers linked by the US government to Iranian intelligence broke into the medical-device manufacturer Stryker and remotely wiped tens of thousands of employee devices. The attack disrupted the company for days, affected its first-quarter earnings, and represented a shift from somewhat more subtle espionage toward more overt and deliberate destruction.Elsewhere, the market-research company Klue sat at the center of a breach affecting close to 200 customers. Attackers used an old credential to gain access to keys for customers’ cloud services.These incidents had people with recognizable motives behind them, and that sort of hack has become increasingly rare over the past decade or so: Black Kite counted 7,551 publicly disclosed ransomware victims over its latest twelve-month reporting period, alone, about 25 percent more than the year before.Then, over the course of about five weeks this summer, OpenAI, Anthropic, and Meta each disclosed incidents in which their AI models gained access to real computer systems they were not supposed to touch.The incidents were not identical, and describing all of them as escapes wouldn’t capture the important nuance of the situation.Anthropic and Meta said their models reached the internet because a third-party evaluation environment had been misconfigured. The models had been told they were operating inside a simulation, and their internal dialogues indicate they interpreted the real systems they encountered when they left that environment as simulated components of that exercise.OpenAI’s incident, however, was a bit different.Its models were being tested using ExploitGym, a benchmark intended to measure their ability to find and exploit software vulnerabilities. Internet access was supposed to be blocked, but the models discovered and exploited vulnerabilities in a package-registry service inside OpenAI’s research infrastructure, found a route to the open internet, and, as I related in a previous episode, eventually compromised the production systems of the AI platform Hugging Face.Since that episode, OpenAI has release a more complete postmortem, which they published in August, and which says the activity was primarily driven by an internal research model comparable in scale to GPT-5.6 Sol, though Sol participated in parts of the incident.The agents created an improvised message board inside the package manager, exchanged discoveries, divided up work, and sometimes described themselves as a swarm.At Hugging Face they executed code on dozens of servers, obtained broad access to several clusters, and exposed credentials associated with databases, cloud services, code repositories, and internal communications.Hugging Face reconstructed around 17,600 actions over a campaign lasting about four and a half days, including roughly two and a half days inside its own infrastructure.Notably, these AI agents were not trying to steal money or damage Hugging Face. They inferred that the company might possess the solutions to the benchmark, to the problem they were tasked with solving, and they went looking for them. They were, in effect, cheating on a test, looking for solutions to the problems rather than taking the test, as intended.That doesn’t mean the models had motives in the human sense. People created the evaluation, reduced the normal safeguards, and assigned an offensive task. But no operator selected Hugging Face or directed the individual steps; that was a path the AI decided to take all by itself.Anthropic’s review subsequently found three incidents among more than 141,000 evaluation runs for its own models. In one, a model unable to reach its fictional target scanned around 9,000 real ones, compromised an internet-facing application using an exposed password and SQL injection, then stopped after recognizing that the host was real.Another model published a malicious package to the real Python Package Index because it believed the registry was simulated. Fifteen outside systems downloaded and ran that malicious package before it was removed.Meta then disclosed that one of its models had reached the internet through a misconfiguration at the same evaluation vendor and exploited a vulnerability at an unnamed third party.No significant financial damages have been publicly reported from these events, by attacker or victim. But if there had been damages, who would have paid for them?What I’d like to talk about today is how autonomous AI systems complicate cyber insurance, how insurers have handled equally unfamiliar risks in the past, and why insurance contracts may soon become one of the more important forms of AI governance.—A typical cyber-insurance policy covers a broad portfolio of costs.These can include ransom payments, forensic investigations, legal expenses, restoring systems and data, notifying customers, and compensating victims and possibly a victims’ customers for the revenue lost while a company’s operations are interrupted.Business interruption is often one of the largest portions of a claim, and policies can respond to malicious attacks as well as non-malicious failures.This market grew by more than 30% a year between 2017 and 2022, as ransomware, a type of attack that became a lot more common during that period, in part because of increased automation and a franchising model that became really popular and increased the reach of the most powerful ransomware tools, almost broke this industry.In 2021, attacks on Colonial Pipeline, the insurer CNA, and meat processor JBS produced multimillion-dollar ransom payments and costly disruptions. Insurance prices surged, sometimes by more than 100%, while some companies found they could not obtain coverage because insurers just couldn’t make the numbers work for them.Insurers responded to this more complex hacking environment by raising prices, but they also made coverage conditional on specific defenses. Companies increasingly had to demonstrate that they used multifactor authentication, endpoint monitoring, restricted administrator access, and backups that attackers could not alter, as a baseline.Loss ratios then fell, more insurance capital entered the market, and prices eventually came down again, stabilizing after that frantic and uncertain period.According to Marsh, global cyber-insurance rates fell 4% in the second quarter of 2026, the twelfth consecutive quarterly decline. Primary pricing is now about 42% below its 2022 peak.The market is not necessarily becoming safer, though. US cyber premiums reached about $7.5 billion in 2025, while the share of premiums consumed by claims rose to 53%—the first time it ticked above 50% since the pandemic-era ransomware surge.Globally, Munich Re estimates the market was worth nearly $15 billion last year and could approach $28 billion by 2030.During this period, insurance applications have also become a consequential part of a company’s security system.In one particularly clear example, Travelers rescinded a million-dollar policy after a ransomware claim revealed that the customer’s multifactor authentication protected only its firewall, despite application answers saying the control was used much more broadly.Companies that don’t live up to cyber insurance expectations can thus be left in the lurch, so in a very real way, insurers have helped make multifactor authentication a standard business practice by attaching a price to its absence. This industry could move faster than regulators because they didn’t have to ban insecure behavior and pass legislation to make that happen; they just had to decline to insure anyone who didn’t live up to their basic security s… This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit letsknowthings.substack.com/subscribe

This week we talk about money policies, yield curves, and government bonds.We also discuss the Fed, the Treasury Department, and a WWII accord between them.Recommended Book: Paved Paradise by Henry GrabarTranscriptIn April of 1942, a few months after the United States entered World War 2, the US Treasury Department asked the Federal Reserve to help it borrow a truly staggering amount of money, and as cheaply as possible. The Fed agreed, committing itself to holding short-term Treasury bill rates at three-eighths of 1%, while also capping the yield on long-term government bonds at 2.5%.This was a type of yield curve control. Rather than allowing the market to decide how much interest the government would pay, the Fed decided that price and promised to enforce it.That helped finance the war, because the Treasury knew its borrowing costs wouldn’t spiral out of control at a moment when it needed to spend unprecedented sums on ships, planes, weapons, soldiers, and all the other machinery of an ongoing global conflict.The downside was that the Fed lost control of an important monetary policy lever.Bond prices and yields move in opposite directions, so keeping yields below a certain level meant the Fed had to stand ready to buy bonds whenever their prices dropped. It couldn’t decide in advance how many it would buy, or how much money it would create in the process. The market would thus forth decide that, instead.Consequently, the Fed became, in some ways, an extension of the Treasury’s debt-management operation, its inflation-related responsibilities made secondary to the government’s need for cheap financing.That arrangement persisted after the war ended, despite the return of inflation, and President Harry Truman’s administration pushed to maintain it during the Korean War, as well.Fed officials resisted, though, with inflation running at more than 8%, and after a very public, very contentious standoff, on March 4, 1951, the Treasury and the Fed announced that they had reached what became known as the Treasury-Fed Accord.That agreement did not make the Fed independent all at once, but it established the principle underlying the modern relationship between these institutions: the Treasury manages government borrowing, while the Fed sets monetary policy based on inflation and employment, not on how much that policy costs the government.The market, in other words, would once again be allowed to decide the price of long-term US debt.What I’d like to talk about today is what happens when that price goes up, what’s pushing long-term US borrowing costs toward levels we haven’t seen in decades, and why two people appointed by the same president are pulling in opposite directions on this issue.—The Federal Reserve’s primary interest-rate lever is the federal funds rate, which is the overnight rate banks charge each other to borrow money. The Fed currently targets a range of 3.5 to 3.75 percent for that rate, and while it has other tools, this is the number people are usually talking about when they say the Fed raised, cut, or held rates.The Fed does not directly set the yield on 10- or 30-year Treasuries, though.Those securities are sold at auction and then traded in a huge secondary market, and their yields reflect a combination of what investors expect inflation to look like, where they think short-term rates will go over the life of the bond, and what’s called the term premium.The term premium is basically extra compensation for uncertainty. If you lock up your money for 30 years instead of rolling over short-term debt, you accept the risk that inflation, growth, government policy, and other variables will change in ways that make your bond less valuable over that thirty year period. The more uncertain the future seems, the more compensation you’re likely to demand.And again, when demand for a bond falls, its price falls and its yield rises. When we say yields are rising, that means borrowers have to offer investors, the people and institutions giving them the money they want to borrow, more money, more interest, to convince them to buy those bonds.That doesn’t only affect the government. The 10-year Treasury serves as something like a reference rate for the entire economy, influencing mortgages, business loans, and the value of long-lived assets.As of September 3 of 2026, the average US 30-year fixed mortgage rate was 6.71%, up from 6.5% a year earlier. That increase is the result of yield increases in the bond market.Long-term Treasury yields have been climbing for much of 2026, and that climb accelerated over the summer.The 30-year yield reached about 5.31 percent on August 17, its highest level since 2007. A few days earlier, the Treasury sold 30-year bonds at a yield of 5.216%, the highest borrowing cost at one of those auctions since 2001.The 10-year yield briefly hit about 4.81% this past week, its highest level since early 2025, and ended Friday at about 4.78%. The two-year yield, which tends to track expectations about contemporary Fed policy more closely, ended at about 4.37%.There isn’t one clean cut reason for these yield bumps. Instead, there are a bunch of forces pushing in roughly the same direction.The first is government borrowing. The Congressional Budget Office now expects a roughly 2.1 trillion dollar federal deficit this fiscal year, which is 200 billion dollars more than it projected in February. Covering that gap means issuing more debt, and more supply generally means the Treasury has to offer a better return to attract enough buyers.The second is competition from corporations, especially technology companies borrowing to build AI infrastructure and data centers.The Dallas Fed estimates that AI-related investment-grade bond issuance—these companies borrowing money, in the form of bonds, to help build more data centers and other AI-enabling stuff—could total around $300 billion this year, creating long-duration debt equivalent to about an eighth of what the Treasury is expected to issue. Some of the companies selling this debt have extremely strong balance sheets and high credit ratings, so investors who want safe-ish, long-term bonds suddenly have a lot more options, and the US government has to compete with that for a finite pool of investor resources.Third, oil prices have surged following renewed strikes and attacks around the Strait of Hormuz, with US benchmark prices recently climbing above $90 a barrel. More expensive energy can goose inflation across the economy, which makes locking in a fixed return for 10 or 30 years less appealing, because those yields might not keep up with the practical devaluation of the dollar.Fourth, that aforementioned term premium has risen as investors ask to be paid more for uncertainty related to inflation, deficits, geopolitics, and future Treasury issuance.And fifth, the pool of buyers is changing. Foreign investors still own trillions of dollars in Treasuries, but private foreign demand for notes and bonds fell sharply in June, even as corporate bonds attracted more of that finite sum of money.A big shift we seem to be seeing here is that some investors seem to be judging Treasuries less as a bet on the next Fed meeting, and more as a long-term bet on whether the US political system can manage its finances. And that shift is showing up at an awkward moment for the two institutions involved in the 1951 Accord.Kevin Warsh, who became Fed chair in May, used his August 28 speech at Jackson Hole to say that although inflation expectations remain anchored, the Fed still has work to do if underlying inflation is not moving toward its target quickly enough.Markets read that as a warning that a rate hike could be coming, and the unexpectedly strong August jobs report reinforced that interpretation: employers added 162,000 jobs, far more than economists anticipated, while estimates for June and July were revised upward.The Treasury Department, meanwhile, is moving in the opposite direction.On August 19, … This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit letsknowthings.substack.com/subscribe