Let's Know Things

This week we talk about Project Glasswing, Anthropic, and Q Day.
We also discuss exploit markets, vulnerabilities, and zero days.
Recommended Book: The Culture Map by Erin Meyer
Transcript
In the world of computer security, a zero-day vulnerability is an issue that exists within a system at launch—hence, zero-day, it’s there at day zero of the system being available—that is also unknown to those who developed said system.
Thus, if Microsoft released a new version of Windows that had a security hole that they didn’t know about, but someone else, a hacking group maybe, discovered before it was released, they might use that vulnerability in Windows or Word or whatever else to hack the end-users of that software.
While large companies like Microsoft do a pretty good job, considering the scope and scale of their product library, of identifying and fixing the worst of the security holes that might leave their customers prone to such attacks, that same scope and scale also means it’s nearly impossible to fill every single possible gap: a truism within the cybersecurity world is that defenders need to get it right every single time, and attackers only need to get it right once, and the same is true here. There’s never been a perfect piece of software, and as these things expand in capability and complexity, the opportunity to miss something also increases, and thus, so does the range of possible errors and exploitable imperfections.
Because of how damaging zero-days can be for both users of software and the companies that make that software, there are thriving marketplaces, similar to those that deal in other illicit goods, where those who discover such vulnerabilities can sell them, usually for cryptocurrencies or funds derived from stolen credit cards.
Software companies have countered the increasing sophistication of these exploit black markets with white and grey market efforts, the former being direct payouts to hackers, basically saying hey, thanks for finding this bug, here’s a lump-sum of money, a bug bounty, rather than punishing all hacking of their systems, which is how they would have previously responded, which had the knock-on effect of sending all hackers, even those who weren’t looking to cause trouble, either underground, or actively hunting for bugs for the black market.
The grey market is more complicated and diverse, and also the largest of marketplaces for those shopping around for these types of exploits. And it’s populated by the same sorts of neverdowells who might frequent the exploit black markets, but also includes all sorts of governments and intelligence agencies, who scoop up these sorts of vulnerabilities to use against their opponents, or to deny them to others who might use them instead, against them.
All sorts of governments, from the US to Russia to North Korea to Iran are regular shoppers on these computer system exploit grey markets, and that has created a complicated, entangled system of incentives, as is some cases, it’s better for the US government, or Iranian government, or whomever, if the company making these systems doesn’t know about a bug or other vulnerability, because they just spent several million dollars to buy a map to said bug or gap, which could, at some point in the future, allow them to tunnel into an enemy’s computers and cause damage or steal information.
What I’d like to talk about today is a new AI system that is apparently very, very good at identifying these sorts of exploits, and why this is being seen as a milestone moment for some people operating in the zero day, and overall computer security space.
—
On April 7, 2026, US-based AI company Anthropic announced Project Glasswing—a new initiative that is currently only available to 11 companies that’s meant to help those companies shore-up their cyber defenses before more AI systems like the one that underpins Project Glasswing, which is called Mythos Preview, hit the market.
So these companies, Amazon Web Services, Anthropic, Apple, Broadcom, Cisco, CrowdStrike, Google, JPMorganChase, the Linux Foundation, Microsoft, NVIDIA, and Palo Alto Networks, make a lot of stuff, and in particular make and maintain a lot of vital online and device-based software infrastructure, like operating systems and all the stuff that keeps things in our apps and on the web secure.
Mythos Preview is a new model created by Anthropic, similar to their existing Claude models, but apparently vastly more powerful. There are tests that AI companies use to compare the potency of their models at a variety of task types, but those are generally considered to be flawed or game-able in all sorts of ways, so the main thing to know here is that Mythos did way better at most of those tests, especially the coding, the programming-related ones, than the other, currently most capable models, the ones that professional programmers, most of them anyway, are using these days. It was also able to do impressive and worrying things like break out of the sandbox that contained it, accessing the internet when it wasn’t supposed to be able to do so.
And because of that leap forward in programming capability, Mythos Preview was tasked by Anthropic with finding vulnerabilities in all sorts of software systems, including operating systems—Windows, macOS, iOS—and browsers, like Chrome and Firefox.
Most AI systems, and most human coders, if they focus enough and look really hard for long enough, will tend to find some kind of vulnerability in just about anything, because this software is just that big and complex. But within a relatively short period of time, Mythos Preview found thousands of vulnerabilities in these systems, indicating that it’s a lot better at this kind of task than the other AI available these days, and so Anthropic created this project, Project Glasswing, to give these entities a head-start, helping them fill these gaps and bolster their defenses, before everyone else on the planet, including foreign governments, hacker and terrorist groups, but also just everyday people, suddenly have the ability to identify and possibly exploit these vulnerabilities, on scale.
This news hasn’t been super widely reported in the non-tech press quite yet, but within the tech world, it landed like a hand grenade in a crowded room.
And there are already quite a few perspectives on what this all means, including a fair bit of skepticism.
On the skeptic side, many analysts have noted that it’s a common tactic amongst AI companies to doomsay, to basically suggest that their models might end the world, might kill all of humanity, might dramatically change everything, put everyone out of work, maybe, not necessarily because the founders and employees at those companies believe that would be the case, but because the implication is that if these products are that powerful, well, investors should probably give them gobs of money, because a tool that could end the world or cause that much disruption might be the last tool available, or might become the next electricity or internet or whatever else. Claiming philosophical, humanistic concern for the super-weapon you just built, in other words, is one way for AI company leaders to say their product is superior to every other product ever while also seeming to suggest that they are the thoughtful, careful leaders that we need holding the reins of that sort of capacity.
Other skeptics have said that while this might be a step-up in terms of the speed at which such vulnerabilities can be identified in these sorts of systems, other AI systems, existing ones, even open source, free ones, have been able to do the same for a while now. So while Mythos Preview might be even better at it, and might be capable of running constantly, finding more and more of these things for a government that wants to save money they might otherwise spend on the grey market, scooping these things up for use against their enemies, or for defensive purposes, sharing some of them with their homegrown tech companies, perhaps, smaller, less-moneyed groups can already do the same, if they’re smart about how they apply existing, even free, lower-end AI systems.
Others have responded to this announcement similarly to how some have responded to the concept of Q Day, short for Quantum Day, which refers to the hypothetical moment at which quantum computers finally become powerful enough to break the encryption that allows the internet, and banking, and government privacy systems to function. If these encryption keys can be broken—and quantum computers should theoretically be able to do this a lot better than conventional computers, because of their very nature—if and when that happens, if these systems aren’t suitably prepared with new encryption that’s hardened against quantum systems, the entire banking sector could collapse, everything hackable, all the money stealable, none of it trustworthy anymore. The same with the whole of the web, with apps, with government systems that keep things hidden away and classified, with energy grids. It could be chaos.
The theory here, then, is that this type of AI, maybe Mythos Preview, maybe the other systems that it portends—because this whole industry seems to leapfrog itself every three or four months at this point, someone coming out with a big, cool, most powerful new thing, then their competitors coming out with something even more powerful within weeks or months—maybe these vulnerability-identifying and exploiting AI will result in something similar, all the world’s software and encryption a lot more vulnerable, all at once, essentially tomorrow.
It’s more of what we’ve already seen with AI, basically, these tools providing anyone who uses them more leverage to do all sorts of things. Not necessarily creating anything new—exploits and vulnerabilities have always existed—but giving a skilled hacker the ability to find and exploit thousands of them in the same time it would have previously taken them to find and exploit just one. And it could also give unskilled, non-hackery people and entities similar capabilities.
That creates a dramatically new cybersecurity landscape essentially overnight, and that’s why, at least according to their press releases on the matter, Anthropic is not releasing Mythos Preview to the public, and instead is taking the Project Glasswing approach: they don’t think other AI companies, like OpenAI or xAI, can be trusted not to just lob that grenade into the crowded room, so since they got there first, they’re going to try to help everyone protect themselves from that grenade when it inevitably lands.
This could, then, be quite the PR coup, giving Anthropic the opportunity to tout their superior products, while also allowing them to portray themselves as sort of the white knight in the AI world, helping everyone protect themselves, even though they probably could have made far more money by either selling the exploits and creating their own new market for them, or by somehow leveraging those exploits themselves.
At the same time, it could be that they are overselling the capabilities of this new model, painting a rosy picture with them as the heroes, while in turn makes their products seem more powerful than they are in order to bolster their public perception and future economic potential.
It could also be a bit of both; even those who are skeptical about this specific announcement and the implications of it do tend to agree it’s likely we’ll see more disruption from these sorts of models soon. Even if Mythos Preview isn’t the grenade everyone’s worried about, in other words, it’s likely we’ll face such a threat in the near-future, and even if Project Glasswing isn’t the defense we need against such a threat, it’s probably prudent that we be thinking about whatever it is we do need, and ideally building it, too, so it’s ready to go, already in place, when that new threat lands.
Show Notes
https://www.nytimes.com/2026/04/10/briefing/claude-mythos-preview.html
https://www.nytimes.com/2026/04/07/technology/anthropic-claims-its-new-ai-model-mythos-is-a-cybersecurity-reckoning.html
https://en.wikipedia.org/wiki/Claude_(language_model)#Claude_Mythos_Preview
https://www.newyorker.com/magazine/2026/04/13/sam-altman-may-control-our-future-can-he-be-trusted
https://www.anthropic.com/glasswing
https://www.wired.com/story/anthropic-mythos-preview-project-glasswing/
https://stratechery.com/2026/myth-and-mythos/
https://en.wikipedia.org/wiki/Zero-day_vulnerability
https://en.wikipedia.org/wiki/Market_for_zero-day_exploits


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 modems, WiFi, and kinda sorta bribes.
We also discuss Huawei, government subsidies, and the FCC.
Recommended Book: Replaceable You by Mary Roach
Transcript
Many homes, those with WiFi connections to the internet, have two different devices they use to make that connectivity happen.
The first is a modem, which is what connects directly to your internet service provider, often via an ethernet jack in the wall that connects to a series of cables webbed throughout your city.
The second is a router, which plugs into the modem and then spreads that signal, derived from that network of city-wide cables around your home, either by splitting that single ethernet jack into multiple ethernet jacks, allowing multiple devices to plug into that network, or by creating a wireless signal, WiFi, that multiple devices can connect to wirelessly in the same way. Many routers will have both options, though in most homes and for most modern devices, WiFi tends to be the more common access point because of its convenience, these days.
That WiFi signal, and the connection provided via those additional ethernet ports on the router, create what’s called a Local Area Network of devices, or LAN. This local area network allow these devices—your phone and your laptop, for instance—to connect to each other directly, but its primary role for most people is using that connection to the modem to grant these devices access the wider internet.
In addition to providing that internet access and creating the Local Area Network, connecting devices on that network to each other, routers also usually provide a layer of security to those devices. This can be done via firewalls and with encryption, which is important as unprotected networks can leave the devices plugged into them vulnerable to outside attack. That means if the router is breached or in some other way exploited, a whole company’s worth of computers, or all your local devices at home, could be made part of a botnet, could be held hostage by ransomware, or could be keylogged until you provide login information for your banks or other seemingly secure accounts to whomever broke into that insufficiently protected LAN.
What I’d like to talk about today is a recently announced ban on some types of routers in the US, the reasoning behind this ban, and what might happen next.
—
On March 23, 2026, the US Federal Communications Commission announced a ban on the import of all new consumer-grade routers not made in the United States.
This ban does not impact routers that are already on the market and in homes, so if you have one already, you’re fine. And if you’re buying an existing model, that should be fine, too.
It will apply to new routers, though, and the rationale provided by the FCC with the announcement is that imported routers are a “severe cybersecurity risk that could be leveraged to immediately and severely disrupt US critical infrastructure.”
They also cited recent, major hacks like Salt Typhoon, saying that routers brought into the US provided a means of entry for some components of those attacks.
This stated concern is similar to the one that was at the center of the Trump administration’s 2019 ban of products made by Chinese tech company Huawei in the United States. Huawei made, and still makes, all kinds of products, including consumer-grade smartphones, and high-end 5G equipment sold to telecommunications companies around the world for use in their infrastructure.
The concern was that a company like Huawei might leverage its far better prices, which were partly possible because of backing from the Chinese government, to put foreign competitors out of business. From there, they could dominate these industries, while also getting their equipment deep in the telecommunications infrastructure of the US and US allies. Then, it would be relatively easy to insert spy equipment and eavesdrop on phone calls and data transmissions from phones, or to incorporate kill-switches into these grids, so if China ever needed to, for instance, distract the US and its allies while they invaded Taiwan, they could just push a button, kill the US telecommunications grid, and that would buy them some time and fog of war to do what they wanted to do without immediate repercussions; and undoing a successful invasion would be a million times more difficult than stepping in while it’s happening to prevent it.
As of 2024, Huawei still controlled about a third of the global 5G market. It controlled about 27.5% back in 2019, the year it was banned in the US and in many US allied nations, so while it’s possible they could have grown even bigger than that had the ban not been implemented, they still grew following its implementation.
Chinese companies currently control about 60% of the US router market, and it’s likely the local, US market will shift, reorienting toward US makers over the next decade or so. But it’s possible these Chinese companies will grow their global footprint even further, as previous US bans have pushed them into different, less exploited markets, and that’s resulted in a wider footprint for such companies, even if their profits may drop a little after leaving the spendier US market.
There’s also a pretty good chance we’ll see deals to move more manufacturing to the US, which could allow some of these companies to make relatively small changes to their operations in order to bypass the ban entirely.
This seems extremely likely, at least in the short term, as all major players in the US router market fall under the FCC’s definition of not being entirely US owned and operated, and making consumer-grade routers that are designed or manufactured outside the US. Even the ostensibly more US companies, based and founded here, make their stuff primarily in Southeast Asia; so even those companies would seem to fall afoul of this new rule.
The FCC has also given these companies the opportunity to apply for what’s called Conditional Approval from the Department of Defense and Department of Homeland Security, which would require they give a bunch of details about their company and products to these entities, along with plans to manufacture more stuff in the US, and these departments can then give them permission to keep selling in the meantime.
It’s worth mentioning here that this kind of set up has previously given foreign entities a chance to funnel money into President Trump’s properties and businesses, before then speaking with him or one of his representatives and coming to some kind of agreement, the President then instructing the relevant agencies or departments to let those companies through, the ban not applying to them or not applying in the same way.
There are concerns that such bans basically operate as requests for bribes, in other words, and those who don’t pay up see their customer base dwindle in the US market, while those who do get away with a slap on the wrist so long as they promise to make more stuff in the US at some point—though they’re not really held to that promise in any concrete way, and often that’s where their efforts stop, at the announcement of such changes.
Also worth mentioning is that it’s not clear why this applies only to consumer-grade routers, as it would seem like the industrial- and military-grade ones would be of even greater concern, at least based on the claims made by the FCC when announcing this ban.
We also don’t know why it’s being applied to new models, but not models currently being sold, and not those already in our homes; all of which would seem to be just as vulnerable as newer models that haven’t made it to the market yet.
There’s a chance those details will follow, and there’s also a chance, again, that this is more about the administration maybe accumulating promises from foreign companies to move manufacturing to the US, because that looks good in an election year, and it’s maybe another means of accumulating bribes from companies that would find it far cheaper to make contributions to organizations the President either controls or favors, than to build new manufacturing capacity in the US, or leave the market entirely.
Show Notes
https://www.theverge.com/tech/899906/fcc-router-ban-march-2026-explainer
https://archive.is/20260326232922/https://www.theverge.com/tech/899906/fcc-router-ban-march-2026-explainer
https://www.cnet.com/home/internet/i-review-routers-for-a-living-dont-buy-a-router-right-now/
https://en.wikipedia.org/wiki/Router_(computing)
https://www.fcc.gov/document/fcc-updates-covered-list-include-foreign-made-consumer-routers
https://www.reuters.com/sustainability/boards-policy-regulation/fcc-banning-imports-new-chinese-made-routers-citing-security-concerns-2026-03-23/
https://www.wired.com/story/us-government-foreign-made-router-ban-explained/
https://itif.org/publications/2025/10/27/backfire-export-controls-helped-huawei-and-hurt-us-firms/


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 cheap drones, energy resources, and Russia’s invasion of Ukraine.
We also discuss the Strait of Hormuz, the war in Iran, and economic asymmetry.
Recommended Book: The Age of Extraction by Tim Wu
Transcript
Russia’s invasion of Ukraine has been pretty universally bad for everyone involved, very much including Russia, which going into the fifth year of this conflict, which it started by massing troops on its neighbor’s border and invading, unprovoked, following years of funding asymmetric military incursions in Ukraine’s southeast. Following their full invasion though, Russia has reportedly suffered around 1.25 million casualties, with more than 400,000 of those casualties suffered in 2025, alone. It’s estimated that Russia has also suffered at least 325,000 deaths, and Ukrainian officials reported confirmed kills of more than 30,000 Russian soldiers just in January 2026.
As of early 2026, Russian controlled about 20% of Ukraine, down from the height of its occupation, back in March of 2022, when it controlled 26% of the country.
And due to a combination of military spending, intense and expansive international sanctions, and damage inflicted by Ukraine, it’s estimated that Russia has incurred about $1 trillion in damages, about a fifth of that being direct operational expenses, and around a fourth the result of reduced growth and lost assets stemming from all those sanctions.
There’s a good chance that all of these numbers, aside from the land controlled, are undercounts, too, as some estimates rely on official figures, and those figures are generally assumed to be partially fabricated to allow Russia to keep face in what is already a pretty humiliating situation—a war they started and which they thought would be a walk in the park, lasting maybe a week, but which has instead gone on to reshape their entire country and present one of the biggest threats to Putin’s control over the Kremlin since he took office.
That in mind, a report from last week, at the tail-end of March, suggests that the Kremlin knows things aren’t looking great for them, and they asked Russian oligarchs to donate money to the cause, to help stabilize Russian finances. This report, which is unconfirmed, but has been reported by multiple Russian media entities, arrives at a moment in which the Russian government is also planning cuts to all sorts of spending, including military spending, but also a reported 10% across the board, to all “non-sensitive” matters in its 2026 budget.
Despite these fairly abysmal figures, though, there’s some optimism in Russia-supporting circles right now, in large part because the conflict in Iran, and Iran’s near shutting down of the Strait of Hormuz, which is an important channel for the flow of international energy assets, that’s goosed the price of oil and gas, which in turn has goosed Russian income substantially.
What I’d like to talk about today are the interconnections between the conflict in Ukraine and the conflict in Iran, and how Ukraine being invaded seems to have put them in a position of relative influence and authority in this new conflict in the Middle East.
—
From the moment Russia launched its full-scale invasion of Ukraine, the Ukrainian military, and its government, industrial base, and pretty much everyone else, scrambled to find an asymmetric means of keeping a far larger, wealthier, and ostensibly more experienced and better backed foe from just steam-rolling over them.
They found that by leveraging lower-cost deterrents, like cheap rockets and drones, they could pay something like $10,000 to take out a tank or other weapons platform that cost Russia a million or ten million dollars. That’s a pretty stellar trade-off, and if you can do that over and over again, eventually you make the cost of the conflict just ridiculously unbalanced, each trade of hardware costing you very little and them a whole lot, which with time can making waging war unsustainable for the side paying orders of magnitudes more.
Russia is of course making use of inexpensive drones and rockets, as well. That’s become a norm in modern conflicts, especially over the past five years or so, as cheap but capable and easy to produce models have started rolling of manufacturing lines in Iran and Turkey, allowing them to become popular sources of single-use but quite agile and deadly aerial weaponry.
Ukraine has gone further than most other entities, though, as they’re immensely incentivized to get this right, and to put their full support behind anything that gives them the upper-hand against what’s still a powerful and otherwise overwhelming invading force. And this patchwork of companies, independent and government supported, large-ish and tiny enough to operate under constant fire and in wartime conditions, has since scaled-up so that they’re expected to manufacture about 7 million drones of many different varieties in 2026.
This scaling has attracted a lot of outside investment, and Ukraine is now considered to be not just a bulwark against current Russian aggression in Europe, taking the brunt of the damage so that Russia isn’t able to turn its attention to the Baltic states and other potential, future targets. It’s also considered to be a vital resource for future protection against Russia, as the US has become a less reliable ally, and NATO, which until recently has been mostly funded and armed by the US, is still getting its legs under it, more members contributing both money and other resources, but possibly not fast enough.
If Russia were to either win in Ukraine and then turn its full-tilt military machine further west, toward other parts of Europe, or if it were to come to some kind of stalemate or peace agreement in Ukraine and then do the same, many leaders throughout Europe believe that Ukrainians, grizzled and scar from this current invasion, will be the ones to train up comparably inexperienced NATO and European Union forces, and to provide the best new, asymmetry-focused military hardware, like drones of all shapes and sizes, as well.
They’ll be not just the arsenal of NATO and the EU, they’ll also probably be the training officers and commanders.
We already see evidence of this probable future demand for Ukrainian goods and services in Gulf states that were attacked by Iran shortly after Israel and the US launched their own attacks that killed Iran’s leader and caused a great deal of damage throughout the country.
Five Iranian neighbors have reportedly made deals with Ukraine to help them defend against future attacks from Iran, especially drone and missile attacks against their energy and water infrastructure.
This help comes in the form of Ukrainian technology, which has been forged by their war, defending against Russia’s incursion, but also training by Ukrainian experts, who are a lot more informed by those war-time realities, and know how to keep infrastructure safe while at the same time taking out the enemy’s capacity to attack in the future.
Ukraine’s hardware is also super cheap compared to comparable alternatives. Ukraine can produce a long-range strike drone for about $200,000, compared to similar drones made by companies in other western countries that cost between $5-10 million. Ukrainian companies also produce far cheaper anti-personnel drones, and interceptor drones and rockets that can flip the cost considerations in some types of conflict.
Often the attacker will launch a bunch of multi-million dollar rockets, alongside a bunch of $10,000 decoys. If your anti-rocket interceptors hit the decoys, and your interceptors cost more than those decoys, maybe a few million dollars apiece, you very quickly end up spending more than your attacker. Reducing the cost of those defensive materials, then, can give the defender the cost advantage, which makes holding out over the long-haul, but also producing enough interceptors to prevent infrastructure damage and save lives, more financially feasible.
There’s a strange interconnectedness between these two conflicts, then, as Russia’s invasion of Ukraine has turned Ukraine into a military product and services powerhouse that’s only just now beginning to scale up, but already in high-demand, while at the same time, Iran’s actions in the Strait of Hormuz, cutting off energy product flow through this vital channel, is boosting Russia’s income dramatically at a moment in which it desperately needs that income to keep invading Ukraine.
That influx of resources could help Russia maintain its invasion for longer than they could otherwise manage, and it could give them a leg up, an even bigger advantage than they already have, which in turn could force Ukraine to become even more skillful and experienced, even better at what they do, leading to even better weapons and tactics that they then share with clients and allies in the Middle East for use against Iran.
Show Notes
https://www.cfr.org/articles/securing-ukraines-future-in-europe-ukraines-defense-industrial-base-an-anchor-for-economic-renewal-and-european-security
https://finance.yahoo.com/news/why-ukraine-drone-defense-ecosystem-205253252.html
https://www.nytimes.com/2026/03/24/world/europe/ukraine-middle-east-oil-and-gas-drones.html
https://gssr.georgetown.edu/the-forum/regions/eurasia/a-first-point-view-examining-ukraines-drone-industry/
https://www.forbes.com/sites/vikrammittal/2026/02/01/ukraine-is-winning-the-economics-battle-against-russian-geran-drones/
https://www.nytimes.com/2026/03/11/world/europe/ukraine-drones-china.html
https://spectrum.ieee.org/drone-warfare-ukraine
https://www.reuters.com/business/aerospace-defense/ukraines-interceptor-drone-makers-look-exports-gulf-iran-war-flares-2026-03-07/
https://united24media.com/war-in-ukraine/what-is-ukraines-interceptor-one-of-the-worlds-most-in-demand-drones-17055
https://www.aa.com.tr/en/asia-pacific/4-years-of-war-counting-russia-s-costs/3838920
https://www.aljazeera.com/news/2026/2/23/the-ukraine-war-in-numbers-people-territory-money
https://www.csis.org/analysis/russias-grinding-war-ukraine
https://news.sky.com/story/putin-asks-oligarchs-to-donate-to-budget-as-cost-of-ukraine-war-soars-13524940
https://www.reuters.com/world/putin-asks-oligarchs-donate-russias-budget-cost-ukraine-war-soars-bell-media-2026-03-27/
https://apnews.com/article/turkish-oil-tanker-attacked-black-sea-2998c366a90ed280e9781a8b030a050c
https://apnews.com/article/russia-ukraine-war-offensive-drones-c9976319f077c743317edec8a20f57f3
https://apnews.com/article/war-russia-ukraine-drones-innovation-interceptor-shahed-e9de7db6437d3cbb428a6bacac326fb3
https://apnews.com/article/russia-ukraine-war-us-talks-iran-drones-40ad8f5481d954fe8207c3d576d540f7
https://www.independent.co.uk/bulletin/news/russia-blackmail-us-zelensky-ukraine-trump-b2945767.html
https://www.newsweek.com/donald-trump-nato-rebuke-iran-war-11738554
https://www.euronews.com/2026/03/26/pentagon-mulls-redirecting-ukraine-military-aid-to-middle-east-reports-claim
https://www.reuters.com/business/energy/ukraine-using-strikes-pressure-russia-after-oil-sanctions-eased-zelenskiy-says-2026-03-26/
https://www.aljazeera.com/features/2026/3/27/ukraine-fends-off-increased-attacks-strikes-russian-oil-revenue


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 the Cold War, the Cuban Missile Crisis, and decapitation attacks.
We also discuss Venezuela, Iran, and the Platt Amendment.
Recommended Book: The Will of the Many by James Islington
Transcript
Cuba is a large island nation, about the same size as the US state of Tennessee, which formally gained its independence from Spain in late 1898, following three wars of independence, the last of which brought the US, Puerto Rico, and the Philippines into play against the Spanish when the Spanish military sunk the USS Maine in Havana Harbor, triggering the Spanish-American War.
That conflict, which Spain lost, led to the US’s acquisition of Puerto Rico, Guam, and the Philippines, and it led to a piece of US legislation called the Platt Amendment, which redefined the relationship between the US and Cuba, following the war, making Cuba a protectorate of the United States, the US promising to leave, withdrawing its troops from Cuban soil, only if seven conditions were met, and an additional provision that Cuba sign a treaty indicating they would continue to adhere to these conditions moving forward—making them permanent.
Most of these conditions relate to Cuba’s ability to enter into relationships with other nations, but provision three also says the US can intervene if doing so will preserve Cuban independence, and that Cuba will sell or lease to the US the land it needs to base its naval vessels in the area, so that it can intervene, militarily if necessary, to keep Cuba independent.
The other provisions are largely related to ensuring Cuba stays financially solvent and clean, the former meant to help maintain that independence, so Cuba doesn’t make deals with other nations, perhaps US enemies, in order to bail itself out when financially in trouble, and the latter meant to help prevent the bubbling up of diseases in a not well-maintained Cuba, that might then spread to the US.
These concerns were concerns for the US government because Cuba is very, very close to the US. It’s just over 90 miles away from Key West, Florida, and that means in the mind of those tasked with defending the US against foreign incursion, Cuba has long represented an uncontrolled variable where enemies could conceivably base all sorts of military assets, including but not limited to nuclear weapons.
That makes Cuba, again, in the minds of defense strategists looking to help the US secure its borders, long-term, something like an aircraft carrier slash nuclear submarine the size of Tennessee, located so close to the US that it could take out all sorts of major assets in a flash, long before the US could respond, getting the same sorts of strike craft and missiles to the Soviet Union.
This framing of the situation, and this collection of concerns, is what led to the Cuban Missile Crisis back in 1962, when the US deployed nuclear weapons in the UK, Italy, and Turkey, all of which were closer to major Soviet hubs than the US, and that led to a tit-for-tat move by the Soviets to deploy nuclear missiles to Cuba, both to get their own weapons closer to the US, just as the US did to them with those new deployments, but also to deter a potential US invasion of Cuba, which was a staunch ally of the Soviet Union.
The crisis lasted 13 days, and though then US President Kennedy was advised to launch an air strike against Soviet missile supplies, and to then invade the Cuban mainland to prevent the basing of Soviet nuclear weapons there, he instead opted for a naval blockade of Cuba, hoping to keep more missile supplies from arriving, and to thus avoid a strike on a Soviet ally that could accidentally spark a shooting war.
After this nearly two-week standoff, the US and Soviet leaders agreed that the Soviets would dismantle the offensive weapons they were building in Cuba in exchange for a public declaration by the US to not invade Cuba. The US also secretly pledged to dismantle its own offensive weapons that it had recently deployed to Italy and Turkey, and the weapons they deployed to the UK were also disbanded the following year.
This sequence of events is generally seen as a minor victory for the US during an especially fraught portion of the Cold War, as that secret agreement between Kennedy and Soviet leader Khrushchev meant that the Soviet people and leadership perceived this agreement as an embarrassing loss, and an example of Soviet weakness on the international stage—they blinked and the US got what they wanted without giving much of anything, though of course, again, the US gave a fair bit too, just in secret.
What I’d like to talk about today is a recent escalation in the US’s posture toward Cuba, and what might happen next, as a result of that change.
—
In early January 2026, the US military, ostensibly as part of a larger effort aimed at disrupting a network of watercraft that carry drugs from mostly South and Central American drugmakers across the border, into US markets, called Operation Southern Spear, the United States implemented a new blockade aimed at sanctioned oil tankers carrying fuel from Venezuela to, among other destinations, Cuban ports.
Shortly before this blockade was declared, the US seized an oil tanker off the coast of Venezuela, then harassed, boarded, and intimidated other tankers, including one from Russia, that were also dealing in Venezuelan oil—something that US sanctions disallowed, and which the Trump administration had decided to focus on, ostensibly as part of that anti-drug effort, but also seemingly as part of a then-impending mission to kidnap Venezuelan President Maduro, who was then secreted away to the US to face trial, which is where he is, today.
These seizures hit Cuba especially hard because the country is highly reliant on all sorts of imports, much of its income derived from tourism, not manufacturing or raw materials, and fuel coming from Venezuela was especially vital—about 72% of Cuba’s electricity generation comes from oil-fueled power plants, and basically its entire transportation section is reliant on the same.
Venezuela under Maduro also provided oil to Cuba at a discount, subsidizing it because those US sanctions didn’t allow Venezuela to have many other reliable customers, and because the authoritarian governments of these two nations saw each other as fellow-travelers in the region, and thus wanted to keep each other propped up against constant pressure from the US and other democracies in the Americas.
As of March 2026, Cuba has gone without crude oil deliveries for three months, and this has led to waves of flight cancellations and a depletion of tourism, which again, is the country’s most vital income source. As of mid-March, Cuba’s energy grid has also collapsed, which has left about 10 million people without power most of the time, amplifying existing problems caused by the country’s antiquated energy generation and distribution systems.
All of which seems to be according to plan for the second US Trump administration, which announced, as far back as January of this year, that it was seeking regime change in Cuba, and these blackouts have triggered exceedingly rare violent protests against that regime by Cuban citizens; these protests haven’t led to any real change or consequences yet, but they could, with time.
For their part, the Cuban government has said they’ve entered diplomatic talks with the US, and they’ve already agreed to release 51 political prisoners, just as an up-front, good will gesture. But they’ve also said changes to the Cuban political system or government—which is an authoritarian regime with absolute power, and which, like most such regimes, is openly corrupt, those in charge enriching themselves at the expense of everyone else, while keeping control via state-sanctioned violence against its own citizens—they’ve said changing that is non-negotiable, also noting that if there is direct aggression against Cuba by the US, they’ll fight and offer up “impenetrable resistance.’
The change that the US government seeks is reportedly similar to what was accomplished in Venezuela: booting the current leader, but keeping the existing regime, the power behind the publicly visible throne, intact, and then the US government influencing that existing regime from afar.
This deviates from the assumed model, attempted by previous US and other governments throughout history, to boot the leaders of opposing government types and then replace them, and the local system, with something closer to their own. This new approach is possibly what the Trump administration is aiming for in Iran, as well, though it’s difficult to say how well the model will work even in Venezuela, where it’s still early days after the US’s seemingly successful decapitation attack, much less in places like Cuba, where there’s no single central power in the public-facing government, much of that power spread between Communist Party leaders, rather than hoarded by a single individual—a far cry from how things were under Castro during the Cold War.
As of the day I’m recording this, there’s a new wrinkle in this blockade: a Russian oil tanker has been tracked heading along a trajectory that would seem to lead to Cuba, which, if accurate, could put the US and Russia at odds over deliveries to the island once more—though in this case it would be oil instead of offensive nuclear weapons that are on board the incursionary vessel.
This ship may veer off that current course and head elsewhere, or it could be meant to test the US oil blockade, intentionally poking at what seems to be an impenetrable barrier, to see if it’s all just talk. Even if just that one tanker makes it through, it’s carrying enough oil to provide about a week’s worth of energy to the Cuban people, which could serve as a sort of release valve on the pressure-cooker stress that has led to the aforementioned protests against the government.
Most analysts expect this and future vessels will turn off when formally confronted, though, and this isn’t the first ship that’s attempted to break this new blockade of Cuba; and previous attempters have indeed pulled off before a shot was fired by the blockading fleet.
Trump has in recent weeks said that he believes he’ll be able to take Cuba, and/or do whatever he wants to the island and its people, and that could just be talk, or it could be that, like in Venezuela, and to some degree Iran, many of the locals would welcome that kind of change, despite the violence and suffering that would no doubt come with it.
In the meantime, though, millions of Cubans are going without reliable energy, food, medical care, and other modern necessities, which could push them to take the risk of revolutionary action, but it could also turn them against the outside enemy, reinforcing support for the tyrannical Cuban government against the harmful and oppressive actions of the American military, rather than nudging them into government overthrow.
Show Notes
https://en.wikipedia.org/wiki/Platt_Amendment
https://en.wikipedia.org/wiki/United_States_embargo_against_Cuba
https://en.wikipedia.org/wiki/2026_Cuban_crisis
https://en.wikipedia.org/wiki/Cuban_Missile_Crisis
https://www.reuters.com/business/energy/cubas-national-electric-grid-collapses-says-grid-operator-2026-03-16
https://www.nytimes.com/2026/03/20/world/americas/cuba-fuel-blockade-aid-convoy.html
https://www.nytimes.com/2026/03/20/travel/cuba-flights-travel-advice-power-oil.html
https://www.reuters.com/world/americas/cuba-says-its-presidents-term-not-subject-negotiation-talks-with-us-2026-03-20/
https://www.dw.com/en/cuba-faces-economic-collapse-as-us-oil-blockade-hits-tourism/video-76398387


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 BYD, Tesla, and the Blade Battery 2.0.
We also discuss EVs, internal-combustion engines, and autonomous vehicles.
Recommended Book: Blank Space by W. David Marx
Transcript
Petroleum-powered vehicles, cars and trucks and SUVs of the kind that have become the standard since the mid-20th century, work by mixing fuel that you put in the tank when you fill up at the gas station with air, in the engine, and then creating a controlled explosion—in modern vehicles using what’s called a four-stroke combustion cycle of intake, compression, combustion, then exhaust—in order to move pistons which, in turn generate mechanical power by transferring that movement to the vehicle’s wheels.
An electric vehicle, in contrast, functions by using electricity from a battery pack to power an electric motor. So rather than needing fuel to combust, which then moves pistons which then moves the wheels, EVs are a straighter-shot with less conversion of energy necessary, electricity powering the motor which powers the wheels.
That simpler setup comes with many advantages, and that difference in the conversion of energy is a big one. Because most of the energy injected into the EV’s system is converted into mechanical movement for the wheels, this type of vehicle only loses about 11% of the energy you put into it to that conversion of electricity to mechanical energy process—around 31-35% is initially lost while charging, converting electricity to motion, and so on, but about 22% is recaptured by the vehicle’s brakes during operation, leading to that 11% average loss.
A gas-powered vehicle, in contrast, because of the inefficiencies inherent in converting fuel to combustion to movement, loses somewhere between 75-84% of the energy you put into it at the gas station, much of that loss in the form of heat that is emitted as a result of that conversion process; this is an inevitable consequence of the thermodynamics of burning fuel to create motion, and one that means operating a gas-powered vehicle is inherently lossier, in the sense that you can’t help but lose the majority of what you put into it as waste, compared to an electric vehicle, which is less lossy to begin with, but even more efficient when you include that in-operation energy recovery.
That baseline reality of energy usage means that modern electric vehicles will typically be cheaper to fuel, to power, because it requires less energy input to get the same amount of travel. This cost-benefit comparison shifts even further in favor of EVs when gas prices are high, though, and though currently the cost of EVs tend to be higher than gas-powered vehicles in most countries, EVs also offer substantially lower lifetime maintenance costs—an average of 40% lower than gas-powered vehicles, due largely to the dramatically reduced number of moving parts in EVs, and the lack of regular, recurring engine-related maintenance tasks, like oil changes and replacing spark plugs.
Not even considering the externalities-related savings of owning and operating an EV, then, like the environmental costs of fuel emissions, such vehicles can save owners tens of thousands of dollars in costs over the span of their ownership—though gas-powered vehicles are still more popular in most markets in part because they’re just more common on car lots, their infrastructure—gas stations versus charging stations—are also more common, and because there are numerous convenience issues, like it being quite a bit faster to pump a tank full of gas than to charge EVs, which is more efficient, but also a piece-of-mind sort of benefit.
What I’d like to talk about today are some recent innovations in the EV and especially EV-scale battery space, and what it might mean for this market in the coming years.
—
After a relatively boom-y period in which EV sales saw a significant uptick, that uptick the consequence of friendly policies and subsidies from successive federal administrations and the rapid-fire innovations arriving in each new generation of EV model being pumped out by US makers, especially Tesla, the US car industry has in recent years pulled back from electric vehicles substantially—the most recent evidence of this being Honda’s recent announcement that three EV models they were planning to manufacture in the US will no longer see production.
This was mostly a money decision, the raw and partially manufactured components necessary for US-based car companies to produce EVs are now burdened with new, Trump-era tariffs, that make producing finished products of this kind in the US all but impossible; simply too expensive to make.
This is also an acknowledgment, though, that Chinese EVs have just gotten so good and so inexpensive for what you get, that it’s simply not possible to compete, not within the current economic and regulatory climate, but also not in the immediate future, even lacking those tariffs, because of how much of a lead Chinese car companies have earned for themselves in this space.
New impositions by the second Trump administration, including those tariffs, but also the killing of EV incentives, and a recent decision to cease enforcing emissions and fuel economy standards, basically telling the industry to make vehicles that pollute more, if they like, have absolutely influenced this state of affairs.
But the quality of new Chinese EVs, the speed at which a large quantity of them can be produced, and the affordability of these vehicles is simply too much for even the world’s most otherwise competitive and industry-owning companies, the most renowned car brands, to match.
There are a few serious EV players in other parts of Asia, and some US companies, like Lucid Motors, are still trying to carve out a space for themselves, pivoting toward skateboard-style platforms that will allow them to use fewer scarce products, like expensive wiring, by using essentially the same base for all of their models, allowing them to ramp-up efficiencies of scale faster, and Rivian, which is trying to claim the outdoorsy, Jeep-esque facet of the US EV market; and Tesla of course continues to own a lot of mindshare in this industry, despite seeming to be pivoting toward AI, autonomous vehicles, and political concerns in recent years.
But this is increasingly China’s domain, and that dominance is the result of a multi-decade push to own basically all the infrastructure and technologies required to electrify their economy, from the ground-up.
As a consequence of that dominance, and all the renewables and battery-making facilities and investments in the relevant companies made by the government for the past few decades, we’re now seeing impressive technological feats coming out of China, like the recently successfully test-flown Sky Dragon electric vertical takeoff and landing aircraft, which looks something like a drone combined with a helicopter, and which can reportedly carry either 10 passengers or a ton of cargo up to 155 miles, which is about 250 kilometers, on a single charge, taking off and landing from helipads, so no runway necessary.
But the already on-the-market, everyday applications of this tech are arguably even more impressive, considering that car-markers in other countries cannot accomplish anywhere near the same, and maybe won’t be able to do so for years.
Chinese carmaker BYD is the top entrant in this space right now, in China and globally, by many metrics, and in early March of 2026 they announced a new battery, called the Blade Battery 2.0, which allows the vehicles it powers to be driven more than 621 miles on a single charge.
That’s compared to the around 400 mile range most large-tanked gas-powered cars can claim. Though even as batteries have gotten larger and more efficient, in terms of their energy storage and expenditure, charging them up has still taken quite a bit longer than filling a tank with gas, often requiring a wait of 30 minutes, though that’s usually just for a small top-up, and only if you have access to a fast-charger. A full-charge sometimes requires as much as 24 hours, if you’re using a small, non-fast public or a home charger.
This differs quite a lot depending where in the world you are, the nature of your EV, and the capacity of the charger you’re using. In general, Tesla superchargers can take a Tesla’s battery from 20% to 89% in around 15-30 minutes, which on average provides another 200 miles of travel; topping it up to 100% usually takes about an hour.
This new battery from BYD, though, which has that 621 mile capacity, can be charged from 10% to 80% in just 6 and a half minutes—and that’s not theory, that exact feat was shown in a public, onstage demonstration.
This isn’t a claim about a technology that will soon arrive, in other words, this is a technology that’s already here, for BYD vehicles, at least. And at six and a half minutes for around 300 miles of range, that brings EVs into the same convenience range as gas vehicles, just a minute or so longer than the average stop at a gas station.
This of course will require specialized charging stations, and those stations will take a while to roll out. The company has said they’ll have 15,000 of their so-called megawatt charging stations available across China by the end of 2026, building 4000 of them, themselves, and the rest through joint ventures. They’re also planning to have about 3000 of these chargers built across European by the end of the year.
All of which will likely further reinforce and lock-in BYD’s advantages over its local and foreign competition, at least for the next several years.
Now, it’s worth mentioning that China’s ’s EV industry is currently a bit tumultuous, the stock prices of companies like BYD tumbling due to wild competition on the Chinese market that until recently has been encouraged by the government, which favors a brutal sort of evolutionary business environment for its favored industries, most of the entrants eventually dying off and leaving fewer, but very strong and internationally competitive companies once the melee has died down.
It’s generally assumed that companies like BYD will cope with this crisis of too-low prices and vehicle overproduction—they and their Chinese competitors are making a lot more EVs than their existing markets can bear—they’ll cope by becoming more aggressive with their international expansion, dropping gobs of these incredibly competitive vehicles in more markets, hoping to offload all that stock, but also to suffocate inferior but more expensive local offerings and, consequently, create more lock-in with international customers through those superior products.
There’s a parallel push for autonomous EVs in many of these markets, which is several years behind the evolution of EV tech, but is also evolving rapidly within China, using that same ultra-brutal competition tactic. These companies are thus quite a bit further along than most of their global competitors, and it seems likely that the semi-autonomous tech built into these newly exported vehicles will help give Chinese companies a leg-up when it comes to exporting autonomous tech to the world, in the next few years.
All of which demonstrates the Chinese market’s major head-start in this and connected technologies, and which points at a serious concern, not just for the US, but for pretty much everyone, as most of these technologies, like better batteries, are relevant not just for the consumer car industry, but also basically every other field, including future military technologies, and tech related to the AI and broader semiconductor industries, which could lead to still-more, and more varied advantages in the near-future.
Show Notes
https://yaleclimateconnections.org/2024/01/electric-vehicles-use-half-the-energy-of-gas-powered-vehicles/
https://www.nrdc.org/stories/electric-vs-gas-cars-it-cheaper-drive-ev#lifetime-costs
https://afdc.energy.gov/vehicles/how-do-all-electric-cars-work
https://www.energy.gov/cmei/vehicles/articles/fotw-1360-sept-16-2024-typical-ev-87-91-efficient-compared-30-conventional
https://arstechnica.com/cars/2026/03/facing-heavy-losses-honda-cancels-its-three-us-made-electric-vehicles/
https://arstechnica.com/cars/2026/03/rivian-reveals-pricing-and-trim-details-for-its-r2-suv/
https://arstechnica.com/cars/2026/03/lucid-announces-midsize-ev-platform-says-profitability-lies-with-suvs/
https://www.livescience.com/technology/electric-vehicles/giant-10-person-flying-taxi-passes-first-flight-test-in-china
https://www.fastcompany.com/91503415/byd-ev-battery-competes-with-gas-engines
https://arstechnica.com/cars/2026/03/byds-latest-evs-can-get-close-to-full-charge-in-just-12-minutes/
https://www.nytimes.com/2026/02/19/business/china-electric-vehicle-troubles.html
https://www.kbb.com/car-advice/how-long-charge-tesla/


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