Let's Know Things

This week we talk about gene-editing, CRISPR/Cas9, and ammonia.We also discuss the germ line, mad scientists, and science research funding.Recommended Book: The Siren’s Call by Chris HayesTranscriptBack in November of 2018, a Chinese scientist named He Jiankui achieved global notoriety by announcing that he had used a relatively new gene-editing technique on human embryos, which led to the birth of the world’s first gene-edited babies.His ambition was to help people with HIV-related fertility problems, one of which is that if a parent is HIV positive, there’s a chance they could transmit HIV to their child.This genetic modification was meant to confer immunity to HIV to the children so that wouldn’t be an issue. And in order to accomplish that immunity, He used a technology called CRISPR/Cas9 to modify the embryos’ DNA to remove their CCR5 gene, which is related to immune system function, but relevant to this undertaking, also serves as a common pathway for the HIV-1 virus, allowing it to infect a new host.CRISPR is an acronym that stands for clustered regularly interspaced short palindromic repeats, and that refers to a type of DNA sequence found in all sorts of genomes, including about half of all sequenced bacterial genomes and just shy of 90% of all sequenced archaea genomes.Cas9 stands for CRISPR-associated protein 9, which is an enzyme that uses CRISPR sequences, those repeating, common sequences in DNA strands, to open up targeted DNA strands—and when paired with specific CRISPR sequences, this duo can search for selected patterns in DNA and then edit those patterns.This tool, then, allows researchers who know the DNA pattern representing a particular genetic trait—a trait that moderates an immune system protein that also happens to serve as a convenient pathway for HIV, for instance—to alter or eliminate that trait. A shorthand and incomplete way of thinking about this tool is as a sort of find and replace tool like you have in a text document on your computer, and in this instance, the gene sequence being replaced is a DNA strand that causes a trait that in turn leads to HIV susceptibility.So that’s what He targeted in those embryos, and the children those embryos eventually became, who are usually referred to as Lulu and Nana, which are pseudonyms, for their privacy, they were the first gene-edited babies; though because of the gene-editing state of the art at the time, while He intended to render these babies’ CCR5 gene entirely nonfunctional, which would replicate a natural mutation that’s been noted in some non-gene-edited people, including the so-called Berlin Patient, who was a patient in Germany in the late-90s who was functionally cured of HIV—the first known person to be thus cured—while that’s what He intended to do, instead these two babies actually carry both a functional and a mutant copy of CCR5, not just the mutant one, which in theory means they’re not immune to HIV, as intended.Regardless of that outcome, which may be less impactful than He and other proponents of this technology may have hoped, He achieved superstardom, briefly, even being named one of the most influential people in the world by Time magazine in 2019. But he was also crushed by controversy, stripped of his license to conduct medical research by the Chinese government, sent to prison for three years and fined 3 million yuan, which is more than $400,000, and generally outcast from the global scientific community for ethical violations, mostly because the type of gene-editing he did wasn’t a one-off sort of thing, it was what’s called germ-line editing, which means those changes won’t just impact Lulu and Nana, they’ll be passed on to their children, as well, and their children’s children, and so on.And the ethical implications of germ-line editing are so much more substantial because while a one-off error would be devastating to the person who suffers it, such an error that is passed on to potentially endless future generations could, conceivably, end humanity.The error doesn’t even have to be a botched job, it could be an edit that makes the edited child taller or more intelligent by some measure, or more resistant to a disease, like HIV—but because this is fringy science and we don’t fully understand how changing one thing might change other things, the implications for such edits are massive.Giving someone an immunity to HIV would theoretically be a good thing, then, but if that edit then went on the market and became common, we might see a generation of humans that are immune to HIV, but potentially more susceptible to something else, or maybe who live shorter lives, or maybe who create a subsequent generation who themselves are prone to all sorts of issues we couldn’t possibly have foreseen, because we made these edits without first mapping all possible implications of making that genetic tweak, and we did so in such a way that those edits would persist throughout the generations.What I’d like to talk about today is another example of a similar technology, but one that’s distinct enough, and which carries substantially less long-term risk, that it’s being greeted primarily with celebration rather than concern.—In early August of 2024, a gene-editing researcher at the University of Pennsylvania, Dr. Kiran Musunuru, was asked if there was anything he could do to help a baby that was being treated at the Children’s Hospital of Philadelphia for CPS1 deficiency, which manifests as an inability to get rid of the ammonia that builds up in one’s body as a byproduct of protein metabolism.We all generate a small amount of ammonia just as a function of living, and this deficiency kept the baby from processing and discarding that ammonia in the usual fashion. As a result, ammonia was building up in its blood and crossing into its brain.The usual method of dealing with this deficiency is severely restricting the suffer’s protein intake so that less ammonia is generated, but being a baby, that meant it wasn’t able to grow; he was getting just enough protein to survive and was in the 7th percentile for body weight.So a doctor at the Children’s Hospital wanted to see if there was anything this gene-editing researcher could do to help this baby, who was at risk of severe brain damage or death because of this condition he was born with.Gene-editing is still a very new technology, and CRISPR and associated technologies are even newer, still often resulting in inaccurate edits, many of which eventually go away, but that also means the intended edit sometimes goes away over time, too—the body’s processes eventually replacing the edited code with the original.That said, these researchers, working with other researchers at institutions around the world, though mostly in the US, were able to rush a usually very cumbersome and time-consuming process that would typically take nearly a decade, and came up with and tested a gene-editing approach to target the specific mutation that was causing this baby’s problems, and they did it in record time: the original email asking if Dr Musunuru might be able to help arrived in August of 2024, and in late-February of 2025, the baby received his first infusion of the substance that would make the proper edits to his genes; they divided the full, intended treatment into three doses, the first being very small, because they didn’t know how the baby would respond to it, and they wanted to be very, very cautious.There were positive signs within the first few weeks, so 22 days later, they administered the second dose, and the third followed after that.Now the research and medical worlds are waiting to see if the treatment sticks; the baby is already up to the 40th percentile in terms of weight for his age, is able to eat a lot more protein and is taking far less medication to help him deal with ammonia buildup, but there’s a chance that he may still need a liver transplant, that there might be unforeseen consequences due to that intended edit, or other, accidental edits made by the treatment, or, again, that the edits won’t stick, as has been the case in some previous trials.Already this is being heralded as a big success, though, as the treatment seems to be at least partially successful, hasn’t triggered any serious, negative consequences, and has stuck around for a while—so even if further treatments are needed to keep the gene edited, there’s a chance this could lead to better and better gene-editing treatments in the future, or that such treatments could replace some medications, or be used for conditions that don’t have reliable medications in the first place.This is also the first known case of a human of any age being given a custom gene-editing treatment (made especially for them, rather than being made to broadly serve any patient with a given ailment or condition), and in some circles that’s considered to be the future of this field, as individually tailored gene-treatments could help folks deal with chronic illnesses and genetic conditions (like cystic fibrosis, Huntington’s disease, muscular dystrophy, and sickle cell), but also possibly help fight cancers and similar issues.More immediately, if this treatment is shown to be long-term efficacious for this first, baby patient, it could be applied to other patients who suffer the same deficiency, which afflicts an estimated 1 in 1.3 million people, globally. It’s not common then—both parents have to have a mutant copy of a specific gene for their child to have this condition—but that’s another reason this type of treatment is considered to be promising: many conditions aren’t widespread enough to justify investment in pharmaceuticals or other medical interventions that would help them, so custom-tailored gene-editing could be used, instead, on a case-by-case basis.This is especially true if the speed at which a customized treatment can be developed is sped-up even further, though there are concerns about the future of this field and researchers’ ability to up its efficiency as, at least in the US, the current administration’s gutting of federal research bodies and funding looks likely to hit this space hard, and previous, similar victories that involved dramatically truncating otherwise ponderous developmental processes—like the historically rapid development of early COVID-19 vaccines—are not looked at favorably by a larger portion of the US electorate, which could mean those in charge of allocating resources and clearing the way for such research might instead pull even more funding and put more roadblocks in place, hobbling those future efforts, rather than the opposite.There are plenty of other researchers and institutions working on similar things around the world, of course, but this particular wing of that larger field may have higher hurdles to leap to get anything done in the coming years, if current trends continue.Again, though, however that larger context evolves, we’re still in the early days of this, and there’s a chance that this approach will turn out to be non-ideal for all sorts of reasons.The concept of tailored gene-editing therapies is an appealing one, though, as it could replace many existing pharmaceutical, surgical, and similar approaches to dealing with chronic, inherited conditions in particular, and because it could in theory at least allow us to address such issues rapidly, and without needing to mess around with the germ-line, because mutations could be assessed and addressed on a person-by-person basis, those edits staying within their bodies and not being passed on to their offspring, rather than attempting to make genetic customizations for future generations based on the imperfect knowledge and know-how of today’s science, and the biased standards and priorities of today’s cultural context.Show Noteshttps://www.nejm.org/doi/full/10.1056/NEJMoa2504747https://www.nih.gov/news-events/news-releases/infant-rare-incurable-disease-first-successfully-receive-personalized-gene-therapy-treatmenthttps://www.wired.com/story/a-baby-received-a-custom-crispr-treatment-in-record-time/https://www.wsj.com/tech/biotech/crispr-gene-editing-therapy-philadelphia-infant-8fc3a2c5https://www.washingtonpost.com/science/2025/05/15/crispr-gene-editing-breakthrough/https://www.npr.org/sections/shots-health-news/2025/05/15/nx-s1-5389620/gene-editing-treatment-crispr-inheritedhttps://interestingengineering.com/health/first-personalized-crispr-gene-therapyhttps://www.nature.com/articles/d41586-025-01496-zhttps://www.nytimes.com/2025/05/15/health/gene-editing-personalized-rare-disorders.htmlhttps://www.nytimes.com/2025/05/31/world/asia/us-science-cuts.htmlhttps://www.livescience.com/health/genetics/us-baby-receives-first-ever-customized-crispr-treatment-for-genetic-diseasehttps://en.wikipedia.org/wiki/He_Jiankui_affairhttps://en.wikipedia.org/wiki/CCR5https://en.wikipedia.org/wiki/Berlin_Patienthttps://en.wikipedia.org/wiki/CRISPR_gene_editinghttps://en.wikipedia.org/wiki/CRISPRhttps://pmc.ncbi.nlm.nih.gov/articles/PMC6813942/ 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 greenhouse gases, renewable energy capacity, and China’s economy.We also discuss coal power plants, natural gas, and gigatons.Recommended Book: What If We Get It Right? by Ayana Elizabeth JohnsonTranscriptIn 2024, global CO2 emissions hit a new all-time high of 37.8 gigatons, that figure including emissions from industrial processes, oil well flaring, and the combustion of fuel, like petroleum in a vehicle.And for context, a gigaton is one billion metric tons, which is about 2.2 trillion pounds. A single gigaton is about the weight of 10,000 fully equipped aircraft carriers, it’s about twice the mass of all human beings on the planet, and it’s approximately the same mass of all non-human land-mammals on earth.That’s one gigaton, and global CO2 emissions last year hit 37.8 gigatons; so quite a lot of carbon dioxide headed into the atmosphere, every year, these days.That’s up about .8% from 2023 levels, and it resulted in an atmospheric CO2 concentration of about 422.5 parts per million, which is around 3 ppm higher than 2023, and 50% higher than pre-industrial levels. And again, for context, if we don’t want to experience global average temperature increases, more extreme weather, and disrupted water cycles, the general consensus is that we want to keep atmospheric CO2 levels at or below 350 ppm, and we’re currently at 422.5 ppm.That said, while emissions grew last year, mostly because fuel combustion increased by around 1%, which overshadowed the decrease in industrial process emissions, which was down 2.3% for the year, emissions growth in 2024 was less than GDP growth; and that’s important because for a long time it was assumed that in order to grow global wealth, according to that metric for wealth, at least, more fossil fuels would need to be burned, because that was the pattern for a long time, industrial revolution, onward.Beginning in the early 2000s, though, GDP growth and emissions growth diverged, and that decoupling has become more prominent as many wealthy nations, including the US, have upped the efficiency of many previously energy-hogging aspects of their economies—things like appliances and the aforementioned industrial processes—while also shifting a lot of energy generation away from massively polluting fuels like coal and oil, over to less-polluting fuels like gas, and non-polluting sources like solar and wind, and in some cases nuclear, as well.This relationship varies significantly from country to country, and the benefits are mostly being seen in so-called advanced economies right now, as many poorer nations are still seeing increased emissions from more polluting power sources, generating electricity, and the growth in wealth leading to more people buying cars and scooters, many of which run on dirty fuels.In the US, though, GDP has doubled since 1990, but CO2 emissions have dropped back down to around 1990 levels.Which to be clear is still a whole lot, as Americans consume a lot of stuff and use a lot of energy, and there are a lot of people living in the US using all that energy and buying all that stuff. But it serves as a good example of this divergence, which we’re also seeing across the EU, where European economies, on average, are 66% larger than in 1990, and CO2 emissions are about a third lower than levels from that same year.What I’d like to talk about today, though, is how this dynamic is playing out in China, a place with a staggeringly high population, a rapidly enlarging middle class, and a whole lot of energy needs.—China is a renewable energy powerhouse.It’s an energy of all kinds powerhouse, truthfully, but its development of renewable energy technologies, and its deployment of those technologies, has been truly remarkable, especially over the past decade or so.China has more renewable energy capacity—mostly solar and wind—than the next 13 countries combined. The US comes in second place, but China has four-times as much renewable energy capacity than the US.Despite that, though, because of China’s huge population and its remarkable wealth-spreading success story, having brought something like 800 million people out of poverty over the past 40 years, a lot of people in the country need a lot more energy, every year. Because as people make more money, they tend to use more electricity and heat, and they tend to buy more things, need bigger homes, and so on. All of which requires more energy.So even though they’ve been building solar panels and wind turbines at a blistering rate, spreading these things all over the place, massively increasing their capacity for clean electricity, they’ve also been building more fossil fuel-burning power plants, especially coal power plants, and that’s made it the world’s biggest emitter of greenhouse gases; a little less than half of the country’s total installed generation capacity burns fossil fuels, which is a huge drop from even a handful of years ago, but because so much of the remaining fossil fuel stock is coal-burning, those energy assets account for an outsized portion of global emissions.New, official data released by the Chinese government, so probably smart to take it with a grain of salt, though these sorts of numbers are usually more reliable than the economy-related numbers they put out, these days, but new data was crunched by Carbon Brief, and they found that March of 2024 was China’s most recent peak in terms of emissions, and since then, their emissions have dropped by 1%.The drop might be accelerating, too, as they also found that new installations in the first quarter of 2025 dropped emissions by 1.6% compared to the same quarter in 2024, so they may be scaling up their renewables deployment efforts, which could lead to even more of a drop.And remarkably, China’s power sector tallied an emissions drop of 5.8%, despite demand for power increasing by 2.5% that same period: which suggests that although China’s population is still wanting more electricity and stuff, the same energy, or rather, a bit more of it, now produces fewer emissions, which means the ratio of renewables to non-renewables in their grid is shifting further in renewables’ favor.Now, as with many other countries, China is beginning to replace coal in some of their power plants with natural gas, instead of swapping them out for solar, wind, and nuclear. Which is absolutely better than coal, but gas still emits CO2 when burned, and there are entirely different problems associated with gas infrastructure, including leaky pipes than allow methane to seep into the atmosphere, which stays up there for a shorter duration than CO2, but is a lot more potent, in terms of heat-capture—so gas in better in some ways, especially short-term ways, than coal, and less polluting for people on the ground, too, but definitely not as good for long-term outcomes as renewables.All that said, there’s some optimism here, as this is the first time this sort of peak and drop has been noted in China’s emission numbers in a context where that drop hasn’t been directly attributable to economic factors; the pandemic, for instance, where a lot less energy was needed, fewer people were driving, and thus there were far fewer emissions globally, for a while.There’s a chance, though, that this trend could be disrupted by the burgeoning trade-war between the Trump administration and essentially everyone, but China in particular. The China-facing component of Trump’s tariffs has been mellowed for a few months, but is still significant at around 30% as of the day I’m recording this. And that could lead to a rewiring of global supply chains, but also a shift in what China manufacturers are producing, how they’re getting those goods to their destination.Those shifting variables could lead to short-term or long-term changes in who’s producing what, how it’s being shipped, and thus, what sorts of energy expenditures we’ll see, and how that energy’s being produced, because of the peculiarities of those new, perhaps rapidly deployed, needs.Show Noteshttps://www.worldbank.org/en/news/press-release/2022/04/01/lifting-800-million-people-out-of-poverty-new-report-looks-at-lessons-from-china-s-experiencehttps://e360.yale.edu/features/china-renewable-energyhttps://arstechnica.com/science/2025/05/analysis-shows-that-chinas-emissions-are-dropping-due-to-renewables/https://www.carbonbrief.org/analysis-clean-energy-just-put-chinas-co2-emissions-into-reverse-for-first-time/https://mn350.org/understanding350https://www.iea.org/reports/global-energy-review-2025/co2-emissionshttps://www.iea.org/commentaries/the-relationship-between-growth-in-gdp-and-co2-has-loosened-it-needs-to-be-cut-completelyhttps://wmo.int/media/news/record-carbon-emissions-highlight-urgency-of-global-greenhouse-gas-watchhttps://www.carbonbrief.org/analysis-global-co2-emissions-will-reach-new-high-in-2024-despite-slower-growth/https://ourworldindata.org/co2-emissions 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 kidnappings, ransoms, and bitcoin.We also discuss crypto wealth, robberies, and memecoins.Recommended Book: The Status Game by Will StorrTranscriptIn 2008, a white paper published by someone writing under the pen name Satoshi Nakamoto proposed a method for making a decentralized asset class called a cryptocurrency that led to the creation of bitcoin, which was implemented and began trading in 2009.While there were other variations on this theme, and attempts to create something like a cryptocurrency previously, bitcoin is generally considered to be the first modern incarnation of this asset class, and its approach—using a peer-to-peer network to keep track of who owns what tokens on a publicly distributed ledger called a blockchain—led to the development of many copycats, and many next-generation cryptoassets based on similar principles, or principles that have been iterated in all sorts of directions, based on the preferences and beliefs of those assets’ founders.In its early days, bitcoin didn’t make much of a splash and was considered to be kind of an anomaly, mostly interesting to a very small number of people who speculated about alternative currencies and how they might be developed and implemented in the real world, but as of mid-May 2025, the global market cap for all cryptocurrencies is $3.39 trillion, bitcoin accounting for more than $2 trillion of that total.That said, there are tens of thousands of cryptocurrencies available, these days, though the majority of them have been formally discontinued or simply allowed to go fallow, becoming functionally inactive.That’s partly the consequence of a surge in interest during the height of the COVID-19 pandemic, the price of bitcoin popping from just over $5,000 at the start of the pandemic to around $68,000 in late-2021.Bitcoin and most of the other crypto-assets that sprung up during that tumultuous period then collapsed when the US Federal Reserve hiked interest rates, intending to temper inflation, which had the knock-on effect of reining in risky bets on things like seed-level startups and alternative assets classes, like crypto—bitcoin dropped to less than $17,000 in 2023, partly as a result of that move—but as inflation levels cooled and investors started to look for assets that might pay out big time again, there was another wave of crypto-asset launches, especially of the ‘meme coin’ variety, which basically means a crypto token that’s launched either as a joke, or to try to make some money off something that’s trending—the most famous meme coin is probably Dogecoin, which was originally released in 2013 as a joke, but then boomed in popularity and price during the pandemic.Through it all, and well before most people knew what bitcoin or cryptocurrencies were, Coinbase has served as a central pillar of the crypto-asset ecosystem.The company was founded in 2012 by a former Airbnb engineer as a crypto exchange: a place where you can swap crypto assets for other crypto assets, but importantly, where you can also sell those assets for real world money, or buy them for real world money.And that’s what I want to talk about today, and more specifically a recent hack of Coinbase, and the potential implications of that hack.—In mid-May of 2025, Coinbase reported, in a legally required Securities and Exchange Commission filing, that their company was hacked, and that the hack may end up costing Coinbase between $180 and $400 million, all told.According to that filing, Coinbase received an email from the hacker on May 11, saying that they’d obtained a bunch of information about Coinbase customers and their accounts, alongside other Coinbase documentation related to their account management systems and customer service practices. The hacker demanded $20 million from the company, which the company refused to pay.Coinbase officials have been keen to note that passwords and private keys were not compromised in the hack, so the hackers couldn’t just log into someone’s account and empty their crypto wallets or the real-deal money they might be keeping there, but they did access names, addresses, phone numbers, and email addresses, alongside the last four digits of some users’ social security numbers, their government ID images, like drivers licenses and passports, and their account balances.All of which isn’t as bad as passwords and private keys being stolen, but it’s not good, either. The hackers, or people working with them, have reportedly been launching phishing attacks against some of the higher net-worth individuals whose information was stolen, those attacks—which usually involve tricking victims into divulging other information, like passwords—made a million times easier because the folks doing the attacks had that stolen information.What’s more, and this isn’t necessarily obvious from reading the pieces published about this hack, but it’s important context surrounding all of this, people who have a lot of money in crypto-assets are increasingly likely to be targeted for other sorts of crimes, compared to people with a lot of wealth in conventional assets, like money, homes, and stocks.Case in point, in early May of this year, a trio of Florida teens kidnapped a man at gunpoint in Las Vegas, drove him to a remote desert area about an hour away, and then stole about $4 million in crypto and other digital assets, like NFTs.They apparently waited for him at his apartment complex and when he pulled up, they threatened him, and said they had his dad, and would kill him if he didn’t get in their car, and then they got his account passwords and other information from him, once he was away from any possible help.In Canada, back in early November of 2024, the CEO of a crypto company based in Toronto was kidnapped during rushhour, forced into the kidnappers’ vehicle and forced to pay a million dollars in ransom before he was released.According to a security firm that specializes in protecting wealthy people with crypto-assets, that CEO’s kidnapping was the 171st instance of criminals using physical violence, including kidnapping, but also other types of robbery, to steal crypto assets; they also said there tend to be more such cases when the price of these assets is high.Well, the price of a bitcoin is high right now, more than $103,000 per coin, as of the day I’m recording this, and France and other Western European nations are seeing a spate of kidnappings of high net-worth crypto-holders, some of which have resulted in mutilation, as was the case with a 60-year-old man who was kidnapped in broad daylight, at 10:30am in Paris—four men in ski-masks pushed him into the back of a delivery van, and his kidnappers demanded his crypto-millionaire son pay a ransom; they cut off the older man’s finger during the ordeal.The kidnappers demanded something like 5-7 million euros, which wasn’t paid, and they were eventually captured by police. But law enforcement is seeing a lot of this sort of thing all over the world right now, people who made fortunes in crypto being kidnapped, and in some cases their friends and family, or partners, also being kidnapped, or kidnapped instead. Whatever the specifics, the person with the crypto-wealth is then hit up for a ransom.Often, the people being targeted are known to be wealthy because their wealth, their gains in this particular asset market, is publicized.The big concern amongst many people in the crypto-world right now, then, is that although the Coinbase hack didn’t result in lost passwords or keys, the information that was stolen, including the balance of users’ accounts, could make these users targets, giving anyone with access to this stolen data a list of people they might steal from, and information about where to find them, how to contact them, and how much they can probably hit them up for.On top of that, they can see who has had large balances in the past, how much cash they sold their holdings for, and who maybe previously had large holdings on Coinbase, but then transferred those assets to a private wallet—so even if they don’t have large Coinbase balances, they possibly have large balances hidden on a harddrive somewhere.Crypto wealth is generally considered to be easier to steal, and to get away with said theft, because of its very nature; it’s largely disconnected from traditional banking systems and many traditional banking regulations, and it’s often simpler to launder crypto assets than real money, converting bitcoin into stable coins into other coins before then converting those assets into real money, for instance.So while Coinbase seems to be doing what they can to make their users whole, including paying back users whose information was lost in the breach, that information then used to phish them, successfully—so if you were conned out of money because the hackers got this information and then tricked you—Coinbase will pay you back what you lost.But it’s not really possible to undo other, non-immediate damage, like the new level of threat some of these hacking victims maybe face, as the global economy gets weird, job security is iffy for many people in many industries, at best, and there’s this list of people who seem to have plenty of money, that money held in more-stealable-than-usual assets, alongside what amounts to a map to where they can be found, and all sorts of information that paints an incomplete, but potentially leveragable, portrait of their lives.Show Noteshttps://www.sec.gov/ix?doc=/Archives/edgar/data/0001679788/000167978825000094/coin-20250514.htmhttps://www.cnbc.com/2025/05/15/coinbase-says-hackers-bribed-staff-to-steal-customer-data-and-are-demanding-20-million-ransom.htmlhttps://techcrunch.com/2025/05/15/coinbase-says-customers-personal-information-stolen-in-data-breach/https://www.nytimes.com/2025/04/24/magazine/crybercrime-crypto-minecraft.html?unlocked_article_code=1.Hk8.UV7K.VEEqHFsUu24g&smid=url-sharehttps://www.yahoo.com/news/florida-teens-kidnap-las-vegas-200918594.htmlhttps://www.cbc.ca/news/canada/toronto/kidnapping-toronto-businessman-cryptocurrency-1.7376679https://www.theguardian.com/world/2025/may/04/french-police-investigate-spate-of-cryptocurrency-millionaire-kidnappingshttps://www.advisor.ca/investments/market-insights/the-reasons-behind-bitcoins-surge/https://www.statista.com/statistics/863917/number-crypto-coins-tokens/https://www.forbes.com/digital-assets/crypto-prices/?sh=c86585d24785https://en.wikipedia.org/wiki/Coinbasehttps://en.wikipedia.org/wiki/Cryptocurrencyhttps://en.wikipedia.org/wiki/Bitcoinhttps://www.sciencedirect.com/science/article/abs/pii/S0378437122005696https://en.wikipedia.org/wiki/Meme_coin 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 NHTSA, CAFE standards, and energy efficiency.We also discuss incentive programs, waste heat, and the EPA.Recommended Book: Africa Is Not a Country by Dipo FaloyinTranscriptIn the United States, fuel-efficiency laws for vehicles sold on the US market are set by the National Highway Traffic Safety Administration, or NHTSA. They set the Corporate Average Fuel Economy, or CAFE standards by which vehicle-makers have to abide, and that, in turn, establishes the minimum standards for companies like Ford or Toyota making vehicles for this market.That CAFE standard is paired with another guideline set by the Environmental Protection Agency that sets standards related to tailpipe emissions. The former says how many miles a vehicle should be able to travel on a gallon of fuel, while the latter says how much CO2, methane, and other pollutants can be legally emitted as that fuel is burnt and those miles are traversed.These two standards address different angles of this issue, but work together to, over time, reduce the amount of fuel consumed to do the same work, and pollution created as that work is accomplished; as a result, if you’re traveling 50 miles today and driving a modern car in the US, you’ll consume a lot less fuel than you would have traveling the same distance in a period-appropriate car twenty years ago.Back in the final year of the Biden administration, the president was criticized for not pushing for more stringent fuel-efficiency standards for US-sold and driven vehicles. The fuel economy requirements were increased by 2% per year for model years 2027 to 2031 for passenger cars, and the same 2% per year requirement will be applied to SUVs and other light trucks for model years 2029 to 2031.This is significantly lower than a previously proposed efficiency requirement, which would have seen new vehicles averaging about 43.5 mpg by model year 2032—an efficiency gain of 18%. And the explanation at the time was that Biden really wanted to incentivize carmakers to shift to EVs, and if they weren’t spending their time and resources on fuel-efficiency tech deployment for their gas-guzzlers, which Biden hoped to start phasing out, they could spend more on refining their EV offerings, which were already falling far behind China’s EV models.Biden wanted half of all new vehicles sold in the US by 2030 to be electric, so the theory was that fuel-efficiency standards were the previous war, and he wanted to fight the next one.Even those watered-down standards were estimated to keep almost 70 billion gallons of gasoline from being consumed through the year 2050, which in turn would reduce US driver emissions by more than 710 million metric tons of CO2 by that same year. They were also expected to save US drivers something like $600 in gas costs over the lifetime of each vehicle they own.Since current president Trump returned to office, however, all of these rules and standards have come into question. Just as when he was president the first time around, rolling back a bunch of Obama-era fuel-efficiency standards—which if implemented as planned would have ensured US-sold vehicles averaged 46.7 mpg by 2026, so better than we were expected to get by 2032 under Biden’s revised minimum—just as he did back then, Trump is targeting these new, Biden standards, while also doing away with a lot of the incentives introduced by the Biden administration meant to make EVs cheaper and more appealing to consumers, and easier to make and sell for car companies.What I’d like to talk about today is another standard, this one far less politicized and widely popular within the US and beyond, that is also being targeted by the second Trump administration, and what might happen if it goes away.—In 1992, the US Environmental Protection Agency, under the endorsement of then-president George HW Bush, launched the Energy Star program: a voluntary labeling program that allowed manufacturers of various types of products to affix a little blue label that says Energy Star on their product, boxes, and/or advertising if their product met the efficiency standards set by this program.So it’s a bit like if those aforementioned fuel-efficiency standards set for vehicles weren’t required, and instead, if your car met the minimum standards, you could slap a little sticker on the car that said it was more energy efficient than cars without said sticker.A low bar to leap, and one that wasn’t considered to be that big a deal, either in terms of being cumbersome for product-makers, or in terms of accomplishing much of anything.Energy Star standards were initially developed for the then-burgeoning field of personal computers and accessories, but in 1995 things really took off, when the program was expanded to include heating and cooling infrastructure, alongside other components for housing and other buildings.From there, new product categories were added on a semi-regular basis, and the government agency folks running the program continued to deploy more technical support and testing tools, making it easier and easier for companies wanting to adhere to these standards to do so, relatively easily and inexpensively.And to provide a sense of what was required to meet Energy Star standards in the days when they were really beginning to take off and become popular, in the early 2000s, refrigerators needed to be about 20% more efficient, in terms of electricity consumption, than the minimum legal standard for such things, while dishwashers needed to be 41% more efficient. Computers around that time, more specifically in 2008, were required to have an 85% efficiency at half load and something close to that at 20% and 100% power load—which basically means it they needed to use most of the energy they drew, and release less of it as waste-heat, which was a big issue for desktop computers at the time.Energy Star TVs had to use 30% less energy than average, with more modern versions of the standard requiring they draw 3 watts or less while in standby mode, and a slew of 90s and early 2000s-era technologies, like VCRs and cordless home phones were required to use something like 90% less electricity than the average at the time.This standard helped push the development of more energy efficient everything, as it was a selling point for companies making things for real estate developers, in particular. Energy-hogs like light fixtures, which cost a fortune to power if you’re thinking in terms of skyscrapers or just building a bunch of houses, became far more energy efficient after the folks in charge of buying the lighting for these projects were able to eyeball options and use the Energy Star label as a shorthand indication that the cost of operation for those goods would be far less, over time, than their competition; it was kind of pointless to buy anything else in many cases, because why would you want to spend all that extra money over time buying less-efficient fluorescent lights for your office buildings, especially now that it was so easy to see, at a glance, which ones were best in this regard?And the same general consensus arrived on the consumer market not long after, as qualified lighting was something like 75% more efficient than non-qualified, legal-minimum-meeting lighting, and Energy Star verified homes were something like 20% cheaper to own.It was estimated that US homeowners living in Energy Star certified homes saved around $360 million on their energy bills in 2016, alone, and another estimate suggests that US citizens, overall, have saved about half a trillion dollars over the past 33 years as a result of the program and the efficiency standards it encourages.So this is a relatively lightweight program that’s optional, and which basically just rewards companies willing to put more efficient products on the market. They can use the little label if they live up to these standards, and that tells customers that this stuff will use less energy than other, comparable products, which in turn saves those customers money over time, and puts less strain on the US electrical grid.This program, consequently, has been very popular, for customers, for the companies making these products—because by jumping through a few hoops, they can get some of their products certified, and that gives them a competitive advantage over companies that don’t do the same, and especially over companies selling cheaper goods from overseas, which tend to be a lot less efficient because of that cheapness—and it’s been popular for politicians across the political spectrum, because people who buy things and pay energy bills vote those politicians into office, and companies that make such goods hire lobbyists to influence their decisions.All of which brings us to today, mid-May of 2025, a point at which the second Trump administration seems to be considering possibly getting rid of the Energy Star certification program.Initial reports on the matter are seemingly well-sourced, but anonymous, as is the case with a lot of White House briefs right now, so some of this should be taken with a grain of salt, because of how it’s being reported and because this administration has flip-flopped a whole lot already, and on things much bigger and more prominent than this, since returning to office, so this could just go away after being reported upon, even if they actually intended to do it before that pushback.But what seems to have happened is this:In January of 2025, after returning to the White House, Trump’s administration put a big Trump supporter and Republican politician, Lee Zelden, in charge of the Environmental Protection Agency.Zelden publicly holds a lot of standard Republican talking points, including what’s often called skepticism about climate science and vehement support of oil drilling, including fracking. He did say that climate change is a real issue that needs to be addressed during his EPA head confirmation hearing, however.Under Trump’s second administration, many government agencies have been either completely done away with, or wiped out, in terms of funding and staff, so that they’re basically just zombie agencies at this point, and the EPA is an agency that Trump has historically not been a big fan of, and which he seems to be trying to rewire toward deregulation: so regulations like fuel efficiency standards are not good according to some strains of usually more conservative politics, and for some business owners, because these are additional rules they have to legally abide by, which costs them money.And back in March of 2025 Zelden announced that the EPA would be pulling back on regulations related to power plants, would incentivize rather than disincentivize the production of oil and gas, would do away with a bunch of pollution-related standards, especially those related to coal power plants and how much pollution they can emit, and many other similar things, which—to shorthand all this—may be somewhat popular if you think climate change concerns are overblown and that it’s more important to keep coal mines operational than to keep streams and rivers clean, but which will generally look really, really bad if you’re any kind of environmentalist and/or are concerned about climate change.The government also recently cut the EPA’s budget by 54.5%, dropping said budget back to where it was when Ronald Reagan was president. This cut, along with cuts to other agencies responsible for tracking dangerous weather, saving sea turtles, and keeping US National Parks clean and functional, will, according to the government, save US taxpayers $163 billion.According to reports from a recent all-hands meeting of the EPA’s Office of Atmospheric Protection, Trump administration officials announced that that office would be dissolved, and that the Energy Star program would be eliminated.Now, there’s a chance that this is just the result of the administration’s at times seemingly blind cutting of budgets, backtracking only when there’s sufficient pushback, and there’s a chance this is a continuation of a political moment a few years back when the Biden administration was considering doing away with Energy Star certification for gas ranges, the idea being that if it uses gas instead of electricity, it’s part of the problem, even if it’s more efficient than other ranges.Republican politicians responded to lobbying efforts from the US gas industry and stirred that up into a big frenzy, to the point that people were vehemently defending their right to own a gas stove, which was never under threat, but that’s how these sorts of astroturfed moral panics work, and it could be that they’re looking to replicate some of that magic now, taking down a standard that they hope to frame as an example of liberal overreach, telling people that these things take away their right to choose what they want to buy, and how much energy or fuel to burn, even when that’s not actually true.There’s also a chance, as I mentioned earlier, though, that this is just a trial balloon, and that once they realize there’s a decent amount of bipartisan support for this program, they’ll step back from this cut, and maybe even claim it for themselves, using it as an example of American exceptionalism: look how great American-made goods are, we’re more efficient than anybody else—not bad messaging at a time in which that kind of competitive language is popular with those in charge, though that competition might not be the real point of all this, at least for some of the people making some of these decisions, right now.Show Noteshttps://www.usatoday.com/story/news/nation/2025/05/09/trump-budget-cuts-environmental-programs/83441472007/https://www.epa.gov/newsreleases/epa-launches-biggest-deregulatory-action-us-historyhttps://en.wikipedia.org/wiki/Lee_Zeldinhttps://web.archive.org/web/20201214180957/https://www.energystar.gov/about/origins_mission/energy_star_overview/about_energy_star_residential_sectorhttps://web.archive.org/web/20161202012204/https://www.energystar.gov/index.cfm?c=about.ab_milestoneshttps://web.archive.org/web/20170622184250/http://www.dailytech.com/New+Energy+Star+50+Specs+for+Computers+Become+Effective+Today/article15559.htmhttps://insideclimatenews.org/news/08052025/energy-star-program-could-be-eliminated-by-trump-administration/https://cleantechnica.com/2025/05/10/energy-star-program-gets-the-kiss-of-death/https://www.theverge.com/news/664670/water-energy-efficiency-standards-trump-dishwasher-washing-machine-showerhead-toilethttps://en.wikipedia.org/wiki/Energy_Starhttps://edition.cnn.com/2025/05/06/climate/energy-star-trumphttps://www.nytimes.com/2025/05/06/climate/epa-energy-star-eliminated.htmlhttps://www.washingtonpost.com/climate-environment/2025/05/06/energy-star-program-epa-trump/https://www.energy.gov/eere/buildings/energy-starhttps://www.npr.org/2025/05/07/g-s1-64905/energy-star-program-cutshttps://apnews.com/article/trump-appliances-consumers-energy-efficiency-3b6100e001a2629dfea9be231f467841https://www.reuters.com/article/business/environment/trump-finalizes-rollback-of-obama-era-vehicle-fuel-efficiency-standards-idUSKBN21I25R/https://apnews.com/article/climate-trump-mpg-fuel-economy-standards-automakers-0ef9147a0c3874a50a194e439f604261https://apnews.com/article/vehicle-fuel-economy-requirement-nhtsa-epa-85e4c3b7bbba9a9a9b7e5b117fe099bdhttps://apnews.com/article/epa-electric-vehicles-emissions-limits-climate-biden-e6d581324af51294048df24269b5d20ahttps://www.nhtsa.gov/laws-regulations/corporate-average-fuel-economy 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 Marshall Plan, standardization, and USB.We also discuss artificial intelligence, Anthropic, and protocols.Recommended Book: Fuzz by Mary RoachTranscriptIn the wake of WWII, the US government implemented the European Recovery Program, more commonly known as the Marshall Plan, to help Western Europe recover from a conflict that had devastated the afflicted countries’ populations, infrastructure, and economies.It kicked off in April of 1948, and though it was replaced by a successor program, the Mutual Security Act, just three years later in 1951—which was similar to the Marshall Plan, but which had a more militant, anti-communism bent, the idea being to keep the Soviets from expanding their influence across the continent and around the world—the general goal of both programs was similar: the US was in pretty good shape, post-war, and in fact by waiting to enter as long as it did, and by becoming the arsenal of the Allied side in the conflict, its economy was flourishing, its manufacturing base was all revved up and needed something to do with all the extra output capacity it had available, all the resources committed to producing hardware and food and so on, so by sharing these resources with allies, by basically just giving a bunch of money and assets and infrastructural necessities to these European governments, the US could get everybody on side, bulwarked against the Soviet Union’s counterinfluence, at a moment in which these governments were otherwise prone to that influence; because they were suffering and weaker than usual, and thus, if the Soviets came in with the right offer, or with enough guns, they could conceivably grab a lot of support and even territory. So it was considered to be in everyone’s best interest, those who wanted to keep the Soviet Union from expanding, at least, to get Europe back on its feet, posthaste.So this program, and its successor program, were highly influential during this period, and it’s generally considered to be one of the better things the US government has done for the world, as while there were clear anti-Soviet incentives at play, it was also a relatively hands-off, large-scale give-away that favorably compared with the Soviets’ more demanding and less generous version of the same.One interesting side effect of the Marshall Plan is that because US manufacturers were sending so much stuff to these foreign ports, their machines and screws and lumber used to rebuild entire cities across Europe, the types of machines and screws and lumber, which were the standard models of each in the US, but many of which were foreign to Europe at the time, became the de facto standard in some of these European cities, as well.Such standards aren’t always the best of all possible options, sometimes they stick around long past their period of ideal utility, and they don’t always stick, but the standards and protocols within an industry or technology do tend to shape that industry or technology’s trajectory for decades into the future, as has been the case with many Marshall Plan-era US standards that rapidly spread around the world as a result of these giveaways.And standards and protocols are what I’d like to talk about today. In particular a new protocol that seems primed to shape the path today’s AI tools are taking.—Today’s artificial intelligence, or AI, which is an ill-defined type of software that generally refers to applications capable of doing vaguely human-like things, like producing text and images, but also somewhat superhuman things, like working with large data-sets and bringing meaning to them, are developing rapidly, becoming more potent and capable seemingly every day.This period of AI development has been in the works for decades, and the technologies required to make the current batch of generative AI tools—the type that makes stuff based on libraries of training data, deriving patterns from that data and then coming up with new stuff based on the prompting of human users—were originally developed in the 1970s, but the transformer, which was a fresh approach to what’s called deep learning architectures, was first proposed in 2017 by a researcher at Google, and that led to the development of the generative pre-trained transformer, or GPT, in 2018.The average non-tech-world person probably started to hear about this generation of AI tools a few years later, maybe when the first transformer-based voice and image tools started popping up around the internet, mostly as novelties, or even more likely in late-2022 when OpenAI released the first version of ChatGPT, a generative AI system attached to a chatbot interface, which made these sorts of tools more accessible to the average person.Since then, there’s been a wave of investment and interest in AI tools, and we’ve reached a point where the seemingly obvious next-step is removing humans from the loop in more AI-related processes.What that means in practice is that while today these tools require human prompting for most of what they do—you have to ask an AI for a specific image, then ask it to refine that image in order to customize it for your intended use-case, for instance—it’s possible to have AI do more things on their own, working from broader instructions to refine their creations themselves over multiple steps and longer periods of time.So rather than chatting with an AI to come up with a marketing plan for your business, prompting it dozens or hundreds of times to refine the sales copy, the logo, the images for the website, the code for the website, and so on, you might tell an AI tool that you’re building a business that does X and ask it to spin up all the assets that you need. From there, the AI might research what a new business in that industry requires, make all the assets you need for it, go back and tweak all those assets based on feedback from other AI tools, and then deploy those assets for you on web hosting services, social media accounts, and the like.It’s possible that at some point these tools could become so capable in this regard that humans won’t need to be involved at all, even for the initial ideation. You could ask an AI what sorts of businesses make sense at the moment, and tell it to build you a dozen minimum viable products for those businesses, and then ask it to run those businesses for you—completely hands off, except for the expressing your wishes part, almost like you’re working with a digital genie.At the moment, components of that potential future are possible, but one of the main things standing in the way is that AI systems largely aren’t agentic enough, which in this context means they need a lot of hand-holding for things that a human being would be capable of doing, but which they largely, with rare exceptions, aren’t yet, and they often don’t have the permission or ability to interact with other tools required to do that kind of building—and that includes things like the ability to create a business account on Shopify, but also the ability to access and handle money, which would be required to set up business and bank accounts, to receive money from customers, and so on.This is changing at a rapid pace, and more companies are making their offerings accessible to specific AI tools; Shopify has deployed its own cluster of internal AI systems, for instance, meant to manage various aspects of a business its customers perch on its platform.What’s missing right now, though, is a unifying scaffolding that allows these services and assets and systems to all play nice with each other.And that’s the issue the Model Context Protocol is meant to address.The Model Context Protocol, or MCP, is a standard developed by AI company Anthropic, and it’s open and designed to be universal. The company intends for it to be the mycelium that connects large language model-based AI to all sorts of data and tools and other systems, a bit like the Hypertext Transfer Protocol, or HTTP, allows data on the web to be used and shared and processed, universally, in a standardized way, and to dip back into the world of physical objects, how standardized shipping containers make global trade a lot more efficient because everyone’s working with the same sized boxes, cargo vessels, and so on.The Universal Serial Bus standard, usually shorthanded as USB, is also a good comparison here, as the USB was introduced to replaced a bunch of other standards in the early days of personal computing, which varied by computer maker, and which made it difficult for those makers, plus those who developed accessories, to make their products accessible and inexpensive for end-users, as you might buy a mouse that doesn’t work with your specific computer hardware, or you might have a cable that fits in the hole on your computer, but doesn’t send the right amount of data, or provide the power you need.USB standards ensured that all devices had the same holes, and that a certain basic level of data and power transmission would be available. And while this standard has since fractured a bit, a period of many different types of USB leading to a lot of confusion, and the deployment of the USB C standard simplying things somewhat, but still being a bit confounding at times, as the same shaped plug may carry different amounts of data and power, despite all that, it has still made things a lot easier for both consumers and producers of electronic goods, as there are fewer plugs and charger types to purchase, and thus less waste, confusion, and so on. We’ve moved on from the wild west era of computer hardware connectivity into something less varied and thus, more predictable and interoperable.The MCP, if it’s successful, could go on to be something like the USB standard in that it would serve as a universal connector between various AI systems and all the things you might want those AI systems to access and use.That might mean you want one of Anthropic’s AI systems to build you a business, without you having to do much or anything at all, and it may be capable of doing so, asking you questions along the way if it requires more clarity or additional permissiosn—to open a bank account in your name, for instance—but otherwise acting more agentically, as intended, even to the point that it could run social media accounts, work with manufacturers of the goods you sell, and handle customer service inquiries on your behalf.What makes this standard a standout compared to other options, though—and there are many other proposed options, right now, as this space is still kind of a wild west—is that though it was developed by Anthropic, which originally made it to work with its Claude family of AI tools, it has since also been adopted by OpenAI, Google DeepMind, and several of the other largest players in the AI world.That means, although there are other options here, all with their own pros and cons, as was the case with USB compared to other connection options back in the day, MCP is usable with many of the biggest and most spendy and powerful entities in the AI world, right now, and that gives it a sort of credibility and gravity that the other standards don’t currently enjoy.This standard is also rapidly being adopted by companies like Block, Apollo, PayPal, CloudFlare, Asana, Plaid, and Sentry, among many, many others—including other connectors, like Zapier, which basically allows stuff to connect to other stuff, further broadening the capacity of AI tools that adopt this standard.While this isn’t a done deal, then, there’s a good chance that MCP will be the first big connective, near-universal standard in this space, which in turn means many of the next-step moves and tools in this space will need to work with it, in order to gain adoption and flourish, and that means, like the standards spread around the world by the Marshall Plan, it will go on to shape the look and feel and capabilities, including the limitations, of future AI tools and scaffoldings.Show Noteshttps://arstechnica.com/information-technology/2025/04/mcp-the-new-usb-c-for-ai-thats-bringing-fierce-rivals-together/https://blog.cloudflare.com/remote-model-context-protocol-servers-mcp/https://oldvcr.blogspot.com/2025/05/what-went-wrong-with-wireless-usb.htmlhttps://arxiv.org/html/2504.16736v2https://en.wikipedia.org/wiki/Model_Context_Protocol#cite_note-anthropic_mcp-1https://github.com/modelcontextprotocolhttps://www.anthropic.com/news/integrationshttps://www.theverge.com/2024/11/25/24305774/anthropic-model-context-protocol-data-sourceshttps://beebom.com/model-context-protocol-mcp-explained/https://techcrunch.com/2025/03/26/openai-adopts-rival-anthropics-standard-for-connecting-ai-models-to-data/https://techcrunch.com/2025/04/09/google-says-itll-embrace-anthropics-standard-for-connecting-ai-models-to-data/https://en.wikipedia.org/wiki/Generative_artificial_intelligencehttps://en.wikipedia.org/wiki/USBhttps://www.archives.gov/milestone-documents/marshall-planhttps://en.wikipedia.org/wiki/Marshall_Planhttps://www.congress.gov/crs-product/R45079https://www.ebsco.com/research-starters/history/marshall-planhttps://www.history.com/articles/marshall-plan 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 British India, Kashmir, and water treaties.We also discuss the global order, sovereignty, and tit-for-tat escalation.Recommended Book: Power Metal by Vince BeiserTranscriptWhen then British India was partitioned by the British in 1947, the country carved up by its colonialist rulers into two new countries, one Hindu majority, the Union of India, and one Muslim majority, the Dominion of Pakistan, the intention was to separate two religious groups that were increasingly at violent odds with each other, within a historical context in which Muslims were worried they would be elbowed out of power by the Hindu-majority, at a moment in which carving up countries into new nations was considered to be a solution to many such problems.The partition didn’t go terribly well by most measures, as the geographic divisions weren’t super well thought out, tens of millions of people had to scramble to upend their entire lives to move to their new, faith-designated homelands, and things like infrastructure and wealth were far from evenly distributed between the two new regions.Pakistan was also a nation literally divided by India, part of its landmass on the other side of what was now another country, and its smaller landmass eventually separated into yet another country following Bangladesh’s violent but successful secession from Pakistan in 1971.There was a lot more to that process, of course, and the reverberations of that decision are still being felt today, in politics, in the distribution of land and assets, and in regional and global conflict.But one affected region, Kashmir, has been more of a flashpoint for problems than most of the rest of formerly British India, in part because of where it’s located, and in part because of happenings not long after the partition.Formerly Jammu and Kashmir, the Kashmir region, today, is carved up between India, Pakistan, and China. India controls a little over half of its total area, which houses 70% of the region’s population, while Pakistan controls a little less than a third of its land mass, and China controls about 15%.What was then Jammu and Kashmir dragged its feet in deciding which side of the partition to join when the countries were being separated, the leader Hindu, though ruling over a Muslim state, but an invasion from the Pakistan side saw it cast its lot in with India. India’s counter-invasion led to the beginning of what became known as both the Indo-Pakistani war of 1947-1948, the first of four such wars, but is also sometimes called the first Kashmir war, the first of three, though there have been several other not-officially-a-war conflicts in and over the region, as well.Things only got more complicated over the next several decades; China seized the eastern part of the region in the 1950s, and while some Kashmiris have demanded independence, both India and Pakistan claim the region as totally their own, and point at historical markers that support their claim—some such markers based on fact, some on speculation or self-serving interpretations of history.What I’d like to talk about today is what looks to be a new, potentially serious buildup around Kashmir, following an attack at a popular tourist hotspot in the territory, and why some analysts are especially concerned about what India’s government will decide to do, next.—Early in the afternoon of April 22, 2025, a group of tourists sightseeing in a town in the southern part of Kashmir called Pahalgam were open-fired on by militants. 26 people were killed and another 17 were injured, marking one of the worst attacks on mostly Indian civilians in decades.In 2019, Kashmir’s semiautonomous governance was revoked by the Indian government, which in practice meant the Indian government took more complete control over the region, clamping down on certain freedoms and enabling more immigration of Indians into otherwise fairly Muslim-heavy Kashmir.It’s also become more of a tourist destination since then, as India has moved more soldiers in to patrol Indian Kashmir’s border with Pakistan Kashmir, and the nature of the landmass makes it a bit of a retreat from climate extremes; at times it’s 30 or 40 degrees cooler, in Fahrenheit, than in New Delhi, so spendy people from the city bring their money to Kashmir to cool off, while also enjoying the natural settings of this less-developed, less-industrialized area.Reports from survivors indicate that the attackers took their time and seemed very confident, and that no Indian security forces were anywhere nearby; they walked person to person, asking them if they were Muslim and executing those who were not. Around 7,000 people were visiting the area as tourists before the attack, but most of them have now left, and it’s unclear what kind of financial hit this will have on the region, but in the short-term it’s expected to be pretty bad.In the wake of this attack, the Indian government claimed that it has identified two of the three suspected militants as Pakistani, but Pakistan has denied any involvement, and has called for a neutral probe into the matter, saying that it’s willing to fully cooperate, seeks only peace and stability, and wants to see justice served.A previously unknown group calling itself the Kashmir Resistance has claimed responsibility for the attack, and Indian security forces have demolished the homes of at least five suspected militants in Kashmir in response, including one who they believe participated in this specific attack.The two governments have launched oppositional measures against each other, including Pakistan closing its airspace to Indian airlines and shutting down trade with its neighbor, and India shutting down a vital land crossing, revoking Pakistani visas, and suspending a 1960 treaty that regulates water-sharing along the Indus River and its tributaries—something that it’s threatened to do, previously, and which could devastate Pakistan’s agricultural sector and economy, as it basically regulates water that the country relies on for both human consumption and most of its crop irrigation; and for context, Pakistan’s agricultural sector accounts for about a forth of its economy.So if India blocks this water source, Pakistan would be in a very bad situation, and the Pakistani government has said that any blockage of water by India would be considered an act of war. Over the past week, a Pakistani official accused the Indian government of suddenly releasing a large volume of water from a dam into a vital river, which made flooding in parts of Pakistan-held Kashmir a real possibility, but as of the day I’m recording this they haven’t closed the taps, as Pakistan has worried.For its part, India wouldn’t really suffer from walking away from this treaty, as it mostly favors Pakistan. It serves to help keep the peace along an at times chaotic border, but beyond that, it does very little for India, directly.So historically, the main purpose of maintaining this treaty, for India, has been related to its reputation: if it walked away from it, it would probably suffer a reputational hit with the international community, as it would be a pretty flagrantly self-serving move that only really served to harm Pakistan, its weaker arch-nemesis.Right now, though, geopolitics are scrambled to such a degree that there are concerns India might not only be wanting to make such moves, whatever the consequences, but it may also be hankering for a larger conflict—looking to sort out long-term issues during a period in which such sorting, such conflict, may cause less reputational damage than might otherwise be the case.Consider that the US government has spoken openly about wanting to take, by whatever means, Greenland, from the Danish, a long-time ally, and that it’s maybe jokingly, but still alarmingly, said that Canada should join the US as the 51st state.These statements are almost certainly just braggadocio, but that the highest-rung people in the most powerful government on the planet would say such things publicly speaks volumes about the Wild West nature of today’s global order.Many leaders seem to be acting like this is a moment in which the prior paradigm, and the post-WWII rules that moderated global behavior within that paradigm, are fraying or disappearing, the global police force represented by the US and its allies pulling inward, not caring, and in some cases even becoming something like bandits, grabbing what they can.Under such circumstances, if you’re in a position of relative power that you couldn’t fully leverage previously, for fear of upsetting that global police force and tarnishing your reputation within that system they maintained, might you leverage it while you can, taking whatever you can grab and weakening your worst perceived enemy, at a moment in which it seems like the getting is good?It’s been argued that Russia’s violation of Ukraine’s sovereignty may have helped kick-off this new paradigm, but Israel’s behavior in Gaza, the West Bank, and increasingly Syria, as well, are arguably even better examples of this changing dynamic.While the Democrats and Joe Biden were in the White House, Israeli Prime Minister Netanyahu seemed to be mostly playing ball, at least superficially, even when he very clearly wasn’t—he did what he could to seem to be toeing rules-based-order lines, even when regularly stepping over them, especially in Gaza.But now, post-Trump’s return to office, that line-toeing has almost entirely disappeared, and the Israeli government seems to be grabbing whatever they can, including large chunks of southwestern Syria, which was exposed by the fall of the Assad regime. The Israeli military launched a full aerial campaign against the Syrian army’s infrastructure, declared a 1974 disengagement agreement with Syria to be void, and though it initially said it would hold the territory it has taken temporarily, it has more recently said it would hold it indefinitely—possibly permanently expanding its country’s land mass at the expense of its neighbor, another sovereign nation, at a moment in which it felt it could get away with doing so.It’s not clear that India has any ambitions on Pakistani territory, beyond what it holds in Kashmir, at least, but there’s a chance it sees this moment the same way the Israeli government does: as a perhaps finite moment during which the previous state of things, the global rules-based-order, no longer applies, or doesn’t apply as much, which suggests it could do some serious damage to its long-time rival and not suffer the consequences it would have, reputationally or otherwise, even half a year ago.And India’s leader, Narendra Modi, is in some ways even better positioned than Israel’s Netanyahu to launch such a campaign, in part because India is in such a favorable geopolitical position right now. As the US changes stance, largely away from Europe and opposing Russia and its allies, toward more fully sidling up to China in the Pacific, India represents a potential counterweight against Chinese influence in the region, where it has successfully made many of its neighbors reliant on its trade, markets, and other resources.Modi has reliably struck stances midway between US and Chinese spheres of influences, allowing it to do business with Russia, buying up a lot of cheap fuel that many other nations won’t touch for fear of violating sanctions, while also doing business with the US, benefitting from a slew of manufacturers who are leaving China to try to avoid increasingly hefty US tariffs.If India were to spark a more concentrated conflict with Pakistan, then, perhaps aiming to hobble its economy, its military, and its capacity to sponsor proxies along its border with India, which periodically launch attacks, including in Kashmir—that might be something that’s not just tolerated, but maybe even celebrated by entities like China and the US, because both want to continue doing their own destabilizing of their own perceived rivals, but also because both would prefer to have India on their side in future great power disagreements, and in any potential future large-scale future conflict.India is richer and more powerful than Pakistan in pretty much every way, but in addition to Pakistan’s decently well-developed military apparatus, like India, it has nukes. So while there’s a chance this could become a more conventional tit-for-tat, leading to limited scuffles and some artillery strikes on mostly military installations across their respective borders, there’s always the potential for misunderstandings, missteps, and tit-for-tat escalations that could push the region into a nuclear conflict, which would be absolutely devastating in terms of human life, as this is one of the most densely populated parts of the world, but could also pull in neighbors and allies, while also making the use of nuclear weapons thinkable by others once more, after a long period of that fortunately not being the case.Show Noteshttps://www.france24.com/en/asia-pacific/20250427-indian-pakistani-troops-exchange-fire-for-third-night-in-disputed-kashmirhttps://www.france24.com/en/live-news/20250422-at-least-24-killed-in-kashmir-attack-on-tourists-indian-police-sourcehttps://www.france24.com/en/asia-pacific/20250424-india-will-identify-track-and-punish-kashmir-attack-perpetrators-modi-sayshttps://www.nytimes.com/2025/04/27/world/asia/india-pakistan-kashmir.htmlhttps://archive.is/20250426143222/https://www.reuters.com/world/india/india-pakistan-exchange-gunfire-2nd-day-ties-plummet-after-attack-2025-04-26/https://www.nytimes.com/2025/04/24/world/asia/india-pakistan-indus-waters-treaty.htmlhttps://www.nytimes.com/2025/04/23/world/asia/kashmir-pahalgam-attack-victims.htmlhttps://apnews.com/article/india-pakistan-kashmir-attack-829911d3eae7cfe6738eda5c0c84d6aehttps://www.bbc.com/news/world-south-asia-11693674https://en.wikipedia.org/wiki/Partition_of_Indiahttps://en.wikipedia.org/wiki/Kashmirhttps://en.wikipedia.org/wiki/Kashmir_conflicthttps://en.wikipedia.org/wiki/Indo-Pakistani_war_of_1947%E2%80%931948 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 chatbots, virtual avatars, and romance novels.We also discuss Inkitt, Galatea, and LLM grooming.Recommended Book: New Cold Wars by David E. SangerTranscriptThere’s evidence that the US Trump administration used AI tools, possibly ChatGPT, possibly another, similar model or models, to generate the numbers they used to justify a recent wave of new tariffs on the country’s allies and enemies.It was also recently reported that Democratic mayoral candidate Andrew Cuomo used AI-generated text and citations in a plan he released called Addressing New York’s Housing Crisis. And this case is a bit more of a slam dunk, as whomever put the plan together for him seems to have just copy-pasted snippets from the ChatGPT interface without changing or checking them—which is increasingly common for all of us, as such interfaces are beginning to replace even search engine results, like those provided by Google.But it’s also a practice that’s generally frowned upon, as—and this is noted even in the copy provided alongside many such tools and their results—these systems provide a whole lot of flawed, false, incomplete, or otherwise not-advisable-to-use data, in some cases flubbing numbers or introducing bizarre grammatical inaccuracies, but in other cases making up research or scientific papers that don’t exist, but presenting them the same as they would a real-deal paper or study. And there’s no way to know without actually going and checking what these things serve up, which can, for many people at least, take a long while; so a lot of people don’t do this, including many politicians and their administrations, and that results in publishing made-up, baseless, numbers, and in some cases wholesale fabricated claims.This isn’t great for many reasons, including that it can reinforce our existing biases. If you want to slap a bunch of tariffs on a bunch of trading partners, you can ask an AI to generated some numbers that justify those high tariffs, and it will do what it can to help; it’s the ultimate yes-man, depending on how you word your queries. And it will do this even if your ask is not great or truthful or ideal.These tools can also help users spiral down conspiracy rabbit holes, can cherry-pick real studies to make it seem as if something that isn’t true is true, and it can help folks who are writing books or producing podcasts come up with just-so stories that seem to support a particular, preferred narrative, but which actually don’t—and which maybe aren’t even real or accurate, as presented.What’s more, there’s also evidence that some nation states, including Russia, are engaging in what’s called LLM grooming, which basically means seeding false information to sources they know these models are trained on so that said models will spit out inaccurate information that serves their intended ends.This is similar to flooding social networks with misinformation and bots that seem to be people from the US, or from another country whose elections they hope to influence, that bot apparently a person who supports a particular cause, but in reality that bot is run by someone in Macedonia or within Russia’s own borders. Or maybe changing the Wikipedia entry and hoping no one changes it back.Instead of polluting social networks or Wikis with such misinfo, though, LLM grooming might mean churning out websites with high SEO, search engine optimization rankings, which then pushes them to the top of search results, which in turn makes it more likely they’ll be scraped and rated highly by AI systems that gather some of their data and understanding of the world, if you want to call it that, from these sources.Over time, this can lead to more AI bots parroting Russia’s preferred interpretation, their propaganda, about things like their invasion of Ukraine, and that, in turn, can slowly nudge the public’s perception on such matters; maybe someone who asks ChatGPT about Russia’s invasion of Ukraine, after hearing someone who supports Russia claiming that it was all Ukraine’s fault, and they’re told, by ChatGPT, which would seem to be an objective source of such information, being an AI bot, that Ukraine in fact brought it upon themselves, or is in some way actually the aggressor, which would serve Russia’s geopolitical purposes. None of which is true, but it starts to seem more true to some people because of that poisoning of the informational well.So there are some issues of large, geopolitical consequence roiling in the AI space right now. But some of the most impactful issues related to this collection of technologies are somewhat smaller in scale, today, at least, but still have the potential to disrupt entire industries as they scale up.And that’s what I’d like to talk about today, focusing especially on a few recent stories related to AI and its growing influence in creative spaces.—There’s a popular meme that’s been shuffling around social media for a year or two, and a version of it, shared by an author named Joanna Maciejewska (machie-YEF-ski) in a post on X, goes like this: “You know what the biggest problem with pushing all-things-AI is? Wrong direction. I want AI to do my laundry and dishes so that I can do art and writing, not for AI to do my art and writing so that I can do my laundry and dishes.”It could be argued, of course, that we already have technologies that do our laundry and dishes, and that AI has the capacity to make both of those machines more efficient and effective, especially in term of helping manage and moderate increasingly renewables-heavy electrical grids, but the general concept here resonates with a lot of people, I think: why are some of the biggest AI companies seemingly dead-set on replacing creatives, who are already often suffering from financial precarity, but who generally enjoy their work, or at least find it satisfying, instead of automating away the drudgery many of us suffer in the work that pays our bills, in our maintenance of our homes, and in how we get around, work on our health, and so on.Why not automate the tedious and painful stuff rather than the pleasurable stuff, basically?I think, looking at the industry more broadly, you can actually see AI creeping up on all these spaces, painful and pleasurable, but generative AI tools, like ChatGPT and its peers, seem to be especially good at generating text and images and such, in part because it’s optimized for communication, being a chatbot interface over a collection of more complex tools, and most of our entertainments operate in similar spaces; using words, using images, these are all things that overlap with the attributes that make for a useful and convincing chatbot.The AI tools that produce music from scratch, writing the lyrics and producing the melodies and incorporating different instruments, working in different genres, the whole, soup to nuts, are based on similar principles to AI systems that work with large sets of linguistic training data to produce purely language based, written outputs.Feed an AI system gobs of music, and it can learn to produce music at the prompting of a user, then, and the same seems to be true of other types of content, as well, from images to movies to video games.This newfound capacity to spit out works that, for all their flaws, would have previously requires a whole lot of time and effort to produce, is leading to jubilation in some spaces, but concern and even outright terror in others.I did an episode not long ago on so-called ‘vibe coding,’ about people who in some cases can’t code at all, but who are producing entire websites and apps and other products just by learning how to interact with these AI tools appropriately. And these vibe coders are having a field day with these tools.The same is increasingly true of people without any music chops who want to make their own songs. Folks with musical backgrounds often get more out of these tools, same as coders tend to get more from vibe coding, in part because they know what to ask for, and in part because they can edit what they get on the other end, making it better and tweaking the output to make it their own.But people without movie-making skills can also type what they want into a box and have these tools spit out a serviceable movie on the other end, and that’s leading to a change similar to what happened when less-fiddly guns were introduced to the battlefield: you no longer needed to have super well-trained soldiers to defeat your enemies, you could just hand them a gun and teach them to shoot and reload it, and you’d do pretty well; you could even defeat some of your contemporaries who had much better trained and more experienced soldiers, but who hadn’t yet made the jump to gunpowder weapons.There are many aspects to this story, and many gray areas that are not as black and white as, for instance, a non-coder suddenly being able to out-code someone who’s worked really hard to become a decent coder, or someone who knows nothing about making music creating bops, with the aide of these tools, that rival those of actual musicians and singers who have worked their whole life to be able to the same.There have been stories about actors selling their likenesses to studios and companies that work with studios, for instance, those likenesses then being used by clients of those companies, often without the actors’ permission.For some, this might be a pretty good deal, as that actor is still free to pursue the work they want to do, and their likeness can be used in the background for a fee, some of that fee going to the actor, no additional work necessary. Their likeness becomes an asset that they wouldn’t have otherwise had—not to be used and rented out in that capacity, at least—and thus, for some, this might be a welcome development.This has, in some cases though, resulted in situations in which said actor discovers that their likeness is being used to hawk products they would never be involved with, like online scams and bogus health cures. They still receive a payment for that use of their image, but they realize that they have little or no control over how and when and for what purposes it’s used.And because of the aforementioned financial precarity that many creatives in particular experience as a result of how their industries work, a lot of people, actors and otherwise, would probably jump at the chance to make some money, even if the terms are abusive and, long-term, not in their best interest.Similar tools, and similar financial arrangements, are being used and made in the publishing world.An author named Manjari Sharma wrote her first book, an enemies-to-lovers style romance, in a series of installments she published on the free fanfic platform Wattpad during the height of the Covid pandemic. She added it to another, similar platform, Inkitt, once it was finished, and it garnered a lot of attention and praise on both.As a result of all that attention, the folks behind Inkitt suggested she move it from their free platform to their premium offering, Galatea, which would allow Sharma to earn a portion of the money gleaned from her work.The platform told her they wanted to turn the book into a series in early 2024, but that she would only have a few weeks to complete the next book, if she accepted their terms. She was busy with work, so she accepted their offer to hire a ghostwriter to produce the sequel, as they told her she’d still receive a cut of the profits, and the fan response to that sequel was…muted. They didn’t like it. Said it had a different vibe, wasn’t well-written, just wasn’t very good. Lacked the magic of the original, basically.She was earning extra money from the sequel, then, but no one really enjoyed it, and she didn’t feel great about that. Galatea then told Sharma that they would make a video series based on the books for their new video app, 49 episodes, each a few minutes long, and again, they’d handle everything, she’d just collect royalties.The royalty money she was earning was a lot less than what traditional publishers offer, but it was enough that she was earning more from those royalties than from her actual bank job, and the company, due to the original deal she made when she posted the book to their service, had the right to do basically anything they wanted with it, so she was kind of stuck, either way.So she knew she had to go along with whatever they wanted to do, and was mostly just trying to benefit from that imbalance where possible. What she didn’t realize, though, was that the company was using AI tools to, according to the company’s CEO, “iterate on the stories,” which basically means using AI to produce sequels and video content for successful, human-written books. As a result of this approach, they have just one head of editorial and five “story intelligence analysts” on staff, alongside some freelancers, handling books and supplementary content written by about 400 authors.As a business model, it’s hard to compete with this approach.As a customer, at the moment, at least, with today’s tools and our approach to using them, it’s often less than ideal. Some AI chatbots are helpful, but many of them just gatekeep so a company can hire fewer customer service humans, saving the business money at the customer’s expense. That seems to be the case with this book’s sequel, too, and many of the people paying to read these things assumed they were written by humans, only to find, after the fact, that they were very mediocre AI-generated knock-offs.There’s a lot of money flooding into this space predicated in part on the promise of being able to replace currently quite expensive people, like those who have to be hired and those who own intellectual property, like the rights to books and the ideas and characters they contain, with near-free versions of the same, the AI doing similar-enough work alongside a human skeleton crew, and that model promises crazy profits by earning the same level of revenue but with dramatically reduced expenses.The degree to which this will actually pan out is still an open question, as, even putting aside the moral and economic quandary of what all these replaced creatives will do, and the legal argument that these AI companies are making right now, that they can just vacuum up all existing content and spit it back out in different arrangements without that being a copyright violation, even setting all of that aside, the quality differential is pretty real, in some spaces right now, and while AI tools do seem to have a lot of promise for all sorts of things, there’s also a chance that the eventual costs of operating them and building out the necessary infrastructure will fail to afford those promised financial benefits, at least in the short term.Show Noteshttps://www.theverge.com/news/648036/intouch-ai-phone-calls-parentshttps://arstechnica.com/ai/2025/04/regrets-actors-who-sold-ai-avatars-stuck-in-black-mirror-esque-dystopia/https://archive.ph/gzfVChttps://archive.ph/91bJbhttps://www.cnn.com/2025/03/08/tech/hollywood-celebrity-deepfakes-congress-law/index.htmlhttps://www.npr.org/2024/12/21/nx-s1-5220301/deepfakes-memes-artificial-intelligence-electionshttps://techcrunch.com/2025/04/13/jack-dorsey-and-elon-musk-would-like-to-delete-all-ip-law/https://www.404media.co/this-college-protester-isnt-real-its-an-ai-powered-undercover-bot-for-cops/https://hellgatenyc.com/andrew-cuomo-chatgpt-housing-plan/https://www.theverge.com/news/642620/trump-tariffs-formula-ai-chatgpt-gemini-claude-grokhttps://www.wsj.com/articles/ai-cant-predict-the-impact-of-tariffsbut-it-will-try-e387e40chttps://www.washingtonpost.com/technology/2025/04/17/llm-poisoning-grooming-chatbots-russia/ 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 smishing, Huione, and scams.We also discuss money laundering, the Cambodian government, and Tether.Recommended Book: The Longevity Imperative by Andrew J. ScottTranscriptThe portmanteau ‘smishing’ combines SMS and phishing to refer to the practice of using text messages to trick the recipients of said messages into revealing information that allows scammers to access their victim’s accounts on various platforms.One common variation of smishing, which I’ve seen a lot recently, personally, are messages purportedly from toll road operators that tell the recipient they’ve got an unpaid toll, and they need to follow a link that’s provided in order to pay it. If the person receiving that message follows the instructions, they’ll tend to land on a webpage that’s convincing enough, which looks like the sort of site you might go to if you’re paying that kind of toll, online, and you enter your payment information and are then either immediately charged for this fake toll, or that information is used in some more cohesive manner—maybe the card is stolen, maybe it’s added to a larger collection of data they have on you which is then leveraged for a larger payout.This type of scam has become more common in recent years because of innovations deployed by what security researchers have called the Smishing Triad, which is a trio of mobile phishing groups operating out of China that seem to have refined their infrastructure and techniques so that messages they send via iMessage to iPhone users and RCS to Android users can bypass mobile phone networks and enjoy a nearly 100% delivery rate—which makes the name a little ironic, since these groups don’t use SMS to deliver these scam texts anymore, as those other methods of delivery are more reliable for such messages, these days.The big innovation introduced by these groups, though, beyond that deliverability, is the productization of mobile phishing, which basically means they’ve packaged up applications that allow their customers, which are usually smaller-time phishing groups and individuals, to share links to convincing-looking copies of Paypal, Mastercard, Stripe, and CitiGroup payment sites, among others, including individual banks, and that makes knee-jerk payments from the victims receiving these texts more likely, and less likely to set of alarm bells in the minds those receiving them, because they look like just normal payment sites.These pre-packaged scam assets also include regularly rotated web domains, which makes them less likely to trigger the recipient’s anti-scam software—their browser will be less likely to flag them as problematic, basically. And the Triad has hundreds of actual humans working desk jobs, worldwide, supporting their customer base, which again is a bunch of scammers that use this package of tools to try to steal money from their marks.All of this is enabled, in part, by clever emulation software that allows Triad customers to leverage legit and legit-seeming phone numbers from a computer or phone, those devices then sending out around 100 messages per second, per device, to phone numbers in the targeted region. They’re able to do this on a budget because of the efficiency of the software acquired from the Smishing Triad, and the Triad stays just ahead of regulators and law enforcement by rapidly iterating their offerings, which in turn does the same for all of their customers—which grants the benefits of a larger institution to all these individual and smaller scam groups.What I’d like to talk about today is another alleged backend for scammers, this one this more overt and public facing, and perhaps even more impactful because of its size and because of the nature of its offerings.—The Huione (hu-WAY-wahn) Group is a financial conglomerate primarily based in Cambodia, though it also has satellite offices in other countries, mostly in Southeast Asia.Folks use the entity’s QR codes to pay for stuff all around Cambodia, from restaurant tabs to hotel bills to supermarket tallies, and it offers normal banking stuff like checking and savings accounts, alongside things like escrow services and a cryptocurrency exchange.This is a company that buys billboards along major highways throughout the country and which has well-connected people in charge, including one of the Cambodian prime minister’s cousins, who is the director of a Huione company.In addition to its many legitimate offerings, though, Huione has also been accused to providing a range of gray and blackmarket products and services to folks who are doing skeevy but partially legal things, alongside wholly criminal enterprises, like a human trafficking outfit in Myanmar and folks running large smishing schemes in other parts of Southeast Asia.Huione’s primary offering for the criminal underworld though, is allegedly serving as a money laundering go-between.If you run a smishing scammer network, or a group that kidnaps people and sell them into various types of modern slavery in Myanmar, you may have trouble using the money you earn for these efforts because they’re off-book, blackmarket sorts of income. You need to clean, to launder that money to make it seem legitimate, so that you can put it in banks or otherwise use it to pay for things like you would with normal, non-illegally earned money.Money laundering matchmaker services maintain networks of what are called money mules, and these mules are sometimes individuals, and they’re sometimes shell companies with bank accounts or their own cryptocurrency wallets.If you’re scamming people out of their money, you might use this type of service to connect you with a money mule, and you provide that mule’s bank or crypto account information to your victim—so when you receive a scammy text message and follow it to completing, the bank your money is sent to will probably be that of a mule, not the person or group doing the scamming.So the victim transfers their money to that mule’s account, and the mule then moves said money from one account to another to another to another to another, eventually converting it into an asset like a cryptocurrency, once the path has been suitably muddled. They take their cut, which is often something like 15%, somewhere along the way, and you, their customer, the scammer, are handed neutralized, clean resources in the form of that cryptocurrency—which you can then convert into real money at some point—on the other end.An entity like Huione makes money by connecting scammers and other criminals with mules, but also by serving as a guarantor on these transactions.So this entity allegedly, via a network of Telegram channels it maintains, telegram being an anonymizing chat app similar to WhatsApp, it allows matchmakers to advertise on these channels, using thinly veiled language to promote their services, and Huione is able to make money selling ads to mules and other matchmakers who want to promote via these highly trafficked channels, one of which has more than 400,000 users—and they have many of these things, and that alone apparently brings in a fair bit of revenue, serving as a sort of hard-to-track Craigslist for this component of the scam economy.The guarantor component of this digital bazaar means that Huione holds the transactions between scammer and mules in escrow, just like any other escrow service: they take the money and hold it until the service has been completed, at which point they release it, taking a small cut for the service of ensuring that no one gets ripped off—except for the original victim of the scam, of course.The majority of these transactions are completed using Tether, which is a stablecoin that tries to peg its value to the US dollar, each token worth exactly one USD, rather than fluctuating like speculative crypto assets, like Bitcoin, and this allows everyone involved to maintain a veil of both feigned ignorance and anonymity, making it difficult to track who does what, how much money changes hands, and who gets paid and does the paying.This setup allows Huione to claim ignorance any time someone accuses them of doing illegal stuff: after all, they can’t possibly be responsible for what all the entities using their services are up to, right? All everything is just muddled and anonymized enough to grant seeming truthfulness to that claim of ignorance.Because of how all this is set up, most of what we know about this is the result of whistleblowing from insiders and leaked documents, alongside divulgences from security researchers who know how to get into these sorts of networks and who at times hack those involved in various ways.And it seems, based on those divulgences and other gleaned knowledge, that Huione’s money laundering services, alone, have been linked to nearly $27 billion in cryptocurrency transactions since 2021—though that could be a significant undercount because of the blurry nature of this industry and the entities involved with it.Thus far, Huione has never been targeted for sanctions by any government.Tether took action to freeze some of its accounts after law enforcement officials flagged them for criminal behavior, and Telegram has closed some of those illicit, matchmaking channels, but it’s easy enough to set up new versions of both, while the escrow subsidiary of Huione, previously called Huione Guarantee, denies any connection to these activities and even changed its name to Haowang Guarantee in October of 2024, though that denial seems to be public-facing only: the escrow-providing company continues to claim that the larger Huione Group is one of its strategic partners and shareholders.Huione also has its own matchmatching service, called Huione International Pay, which operates as a real-deal bank, but also does what all the other matchmakers do—it helps criminal enterprises shuffle their money around, taking a fee to provide them with clean money, usually in the shape of Tether crypto tokens, on the other end.Though notably, Huione also recently launched their own stablecoin called USDH, alongside an in-house communication service called ChatMe and an array of mini-games that seem optimized for automation, which is another means of laundering money via what seems like gambling apps, allowing their clients to cut out the casinos that are sometimes used as part of the laundering process. All of which seems primed to internalize more of this process, slowly doing away with the need for Telegram and Tether and those casinos, which would seem to remove some of the risk associated with those external, uncontrolled-by-Huione, platforms.Despite all this, this enterprise has been allowed to flourish and grow like it has, according to a threat analyst with the UN, at least, because of lax enforcement in Cambodia, and the conglomerate’s connections with the government and ability to say, basically, we’re legit, look, we’re just a bank, we can’t control what other people might do with our services. Their whole setup is obscure enough, too, that anyone who takes a close look at their entangled business structure quickly gets lost in its complexity and many tangles and dead-ends.Some governments, including the Chinese government, have been cracking down on entities like Huione operating within their borders, but many such crackdowns are hobbled when they’re aimed at operations based in different countries, especially those with lax enforcement, like Cambodia.Also worth noting is that if someone’s going to get caught, it’ll most likely be the mules, not the matchmakers or scammers, and that’s by design. It’s a bit like street-level drug dealers being more likely to be picked up by police than the folks running the larger drug enterprise of which they’re a part. Huione and other entities like it are largely insulated from major consequences, even if the mules who use their services periodically get caught in dragnets cast by law enforcement.That said, the National Bank of Cambodia recently announced that it hasn’t renewed Huione’s license to operate its payment service in the country, the one that runs all those QR codes, because it didn’t meet renewal requirements. That happened in late-March of 2025, so pretty recently, though the company has already said that it will register its business in Japan and Canada, so it seems to be looking for a suitable plot of land on which to rebuild this component of its setup.Many security researchers and law enforcement officials have warned that the time to crack down on Huione and similar conglomerates is now, because they’re currently reliant on partially exposed third-parties like Telegram and Tether. Once they successfully move those activities inward, they’ll be a lot more difficult to track, but also nearly impossible to shutter, unless there’s a significant change in the government and enforcement climate in the countries in which they’re based, which at this point at least, looks unlikely.Show Noteshttps://www.nytimes.com/2025/03/23/world/asia/cambodia-money-laundering-huione.htmlhttps://www.wired.com/story/the-largest-illicit-online-marketplace-ever-is-growing-at-an-alarming-rate/https://www.wired.com/story/pig-butchering-scam-crypto-huione-guarantee/https://www.wired.com/story/interpol-pig-butchering-scams-rename/https://www.propublica.org/article/casinos-cambodia-myanmar-laos-southeast-asia-fraud-cybercrimehttps://krebsonsecurity.com/2025/04/china-based-sms-phishing-triad-pivots-to-banks/#more-70793https://en.wikipedia.org/wiki/Mobile_phone_spam 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 taxes, reciprocity, and recession.We also discuss falling indices, stagflation, and theories of operation.Recommended Book: The Serviceberry by Robin Wall KimmererTranscriptStagflation, which is a portmanteau of stagnation and inflation, is exactly what it sounds like: a combination of those two elements, usually with high levels of unemployment, as well, that can cause a prolonged period of economic sluggishness and strain that slows growth and can even lead to a recession.The term was coined in the UK in the 1960s to describe issues they were facing at the time, but it was globally popularized by the oil shocks of the 1970s, which sparked a period of high prices and slow growth in many countries, including in the US, where inflation boomed, productivity floundered, and economic growth plateaud, leading to a stock market crash in 1973 and 1974.Inflation, unto itself, can be troubling, as it means prices are going up faster than incomes, so the money people earn and have saved is worth less and less each day. That leads to a bunch of negative knock-on effects, which is a big part of why the US Fed has kept interest rates so high, aiming to trim inflation rates back to their preferred level of about 2% as quickly as possible in the wake of inflation surges following the height of the Covid pandemic.Stagnant economic growth is also troubling, as it means lowered GDP, reduced future outlook for an economy, and that also tends to mean less investment in said economy, reduced employment levels—and likely even lower employment levels in the future—and an overall sense of malaise that can become a self-fulfilling prophecy, no one feeling particularly upbeat about where their country is going; and that’s not great economically, but it can also lead to all sorts of social issues, as people with nothing to look forward to but worse and worse outcomes are more likely to commit crimes or stoke revolutions than their upbeat, optimistic, comfortable kin.The combination of these two elements is more dastardly than just the sum of their two values implies, though, as measures that government agencies might take to curb inflation, like raising interest rates and overall tightening monetary policy, reduces business investment which can lead to unemployment. On the flip-side, though, things a government might do to reduce unemployment, like injecting more money into the economy, tends to spike inflation.It’s a lose-lose situation, basically, and that’s why government agencies tasked with keeping things moving along steadily go far out of their way to avoid stagflation; it’s not easily addressed, and it only really goes away with time, and sometimes a very long time.There are two primary variables that have historically led to stagflation: supply shocks and government policies that reduce output and increase the money supply too rapidly.The stagflation many countries experienced in the 1970s was the result of Middle Eastern oil producing nations cutting off the flow of oil to countries that supported Israel during the 1973 Yom Kippur War, though a sharp increase in money supply and the end of the Bretton Woods money management system, which caused exchange rate issues between global currencies, also contributed, and perhaps even more so than the oil shock.What I’d like to talk about today is another major variable, the implementation of a huge package of new tariffs on pretty much everyone by the US, that many economists are saying could lead to a new period of stagflation, alongside other, more immediate consequences.—A tariff is a type of tax that’s imposed on imported goods, usually targeting specific types of goods, or goods from a particular place.Way back in the day these were an important means of funding governments: the US government actually made most of its revenue, about 90% of it, from tariffs before 1863, because there just wasn’t a whole of lot other ways for the young country to make money at the time.Following the War of 1812, the US government attempted to double tariffs, but that depleted international trade, which led to less income, not more—gross imports dropped by 71%, and the government scrambled to implement direct and excise taxes, the former of which is the tax a person or business pays that isn’t based on transactions, while the latter is a duty that’s paid upon the manufacture of something, as opposed to when it’s sold.Tariffs resurfaced in the following decades, but accounted for less and less of the government’s income as the country’s manufacturing base increased, and excise and income taxes made up 63% of the US’s federal revenue by 1865.Tax sources have changes a lot over the years, and they vary somewhat from country to country.But the dominant move in the 20th century, especially post-WWII, has been toward free trade, which usually means no or low tariffs on goods being made in one place and sold in another, in part because this tends to lead to more wealth for everyone, on average, at least.This refocus toward globalized free trade resulted in a lot of positives, like being able to specialize and make things where they’re cheap and sell them where they’re precious, but also some negatives, like the offshoring of jobs—though even those negatives, which sucked for the people who lost their jobs, have been positive for some, as the companies who offshored the jobs did so because it saved them money, the folks who were hired were generally paid more than was possible in their region, previously, and the people consuming the resulting goods were able to get them cheaper than would otherwise be feasible.It’s been a mixed bag, then, but the general consensus among economists is that open trade is good because it incentivizes competition and productivity. Governments are less likely to implement protectionist policies to preserve badly performing local business entities from better performing foreign versions of the same, and that means less wasted effort and resources, more options for everyone, and more efficient overall economic operation, which contributes to global flourishing. And not for nothing, nations that trade with each other tend to be less likely to go to war with each other.Now that’s a massively simplified version of the argument, but again, that’s been the outline for how things are meant to work, and aside from some obvious exceptions—like China’s protection of its local tech sector from foreign competition, and the US’s protection of its aviation and car industries—it’s generally worked as intended, and the world has become massively wealthier during this period compared to before this state of affairs was broadly implemented, post-WWII; there’s simply no comparison, the difference is stark.There are renewed concerns about stagflation in the United States, however, because of a big announcement made by US President Trump on April 2, 2025, that slapped substantial and at times simply massive new tariffs on just about everyone, including the country’s longest-term allies and most valuable trading partners.On what the president called “Liberation Day,” he announced two new types of tariff: one is a universal 10% import duty on all goods brought into the US, and another that he called a reciprocal tariff on imports from scores of countries, including 15 that will be hit especially hard—a list that includes China, EU nations, Canada, and Japan, among others.The theory of these so-called reciprocal tariffs is that Trump thinks the US is being taken advantage of, as, to use one example that he cited, the US charges a 2.5% tariff on imported cars, while the EU charges a 10% tariff on American cars imported to their union.The primary criticism of this approach, which has been cited by most economists and entities like the World Trade Organization, is that the numbers the US administration apparently used to make this list don’t really add up, and seem to include some made-up measures of trade deficits, which some analysts suspect were calculated by AI tools like ChatGPT, as the same incorrect measures are spat out by commonly use chatbots like ChatGPT when they’re asked about how to balance these sorts of things. But the important takeaway, however they arrived at these numbers, is that the comparisons used aren’t really sensical when you look at the details.Some countries simply can’t afford American exports, for instance, while others have no use for them. The idea that a country that can’t afford American goods should have astoundingly large tariffs applied to their exports to the US is questionable from the get-go, but it also means the goods they produce, which might be valuable and important for Americans, be they raw materials like food or manufactured goods like car parts, will become more expensive for Americans, either because those Americans have to pay a higher price necessitated by the tax, or because the lower-price supplier is forced out of the market and replaced by a higher-price alternative.In short, the implied balance of these tariffs don’t line up with reality, according to essentially everyone except folks working within Trump’s administration, and the question then is what the actual motivation behind them might be.The Occam’s Razor answer is that Trump and/or people in his administration simply don’t understand tariffs and global economics well enough to understand that their theory on the matter is wrong. And many foreign leaders have said these tariffs are not in any way reciprocal, and that the calculation used to draw them up was, in the words of Germany’s economic minister, “nonsense.” That’s the general consensus of learned people, and the only folks who seem to be saying otherwise are the one’s responsible for drawing these tariffs up, and defending them in the press.Things have been pretty stellar for most of the global economy since free trade became the go-to setup for imports and exports, but this administration is acting as if the opposite is true. That might be a feigned misunderstanding, or it might be genuine; they might truly not understand the difference between how things have been post-WWII and how they were back in the 1800s when tariffs were the go-to method of earning government revenue.But in either case, Trump is promising that rewiring the global order, the nature of default international trade in this way, will be good for Americans because rather than serving as a linchpin for that global setup, keeping things orderly by serving as the biggest market in the world, the American economy will be a behemoth that gets what it’s owed, even if at the expense of others—a winner among losers who keep playing because they can’t afford not to, rather than a possibly slightly less winning winner amongst other winners.This theory seems to have stemmed from a 1980s understanding of things, which is a cultural and economic milieu from which a lot of Trump’s views and ideas seem to have originated, despite in many cases having long since been disproved or shown to be incomplete. But it’s also a premise that may be more appealing to very wealthy people, because a lot of the negative consequences from these tariffs will be experienced by people in lower economic classes and people from poorer nations, where the price hikes will be excruciating, and folks in the middle class, whose wealth is primarily kept in stocks. Folks in the higher economic echolons, including those making most of these decisions, tend to make and build their wealth via other means, which won’t be entirely unimpacted, but will certainly be less hurt by these moves than everyone else.It’s also possible, and this seems more likely to me, but it’s of course impossible to know the truth of the matter right now, that Trump is implementing a huge version of his go-to negotiating tactic of basically hurting the folks on the other end of a negotiation in order to establish leverage over them, and then starting that negotiation by asking what they’ll do for him if he limits or stops the pain.The US is expected to suffer greatly from these tariffs, but other countries, especially those that rely heavily on the US market as their consumer base, and in some cases for a huge chunk of their economy, their total GDP, will suffer even more.There’s a good chance many countries, in public or behind closed doors, will look at the numbers and decide that it makes more sense to give Trump and his administration something big, up front, in exchange for a lessening of these tariffs. That’s what seems to be happening with Vietnam, already, and Israel, and there’s a good chance other nations have already put out feelers to see what he might want in exchange for some preferential treatment in this regard—early reports suggest at least 50 governments have done exactly that since the announcement, though those reports are coming from within the White House, so it’s probably prudent to take them with a grain of salt, at this point. That said, this sort of messaging from the White House suggests that the administration might be hoping for a bunch of US-favoring deals and will therefore make a lot of noise about initial negotiations to signal that that’s what they want, and that the pain can go away if everyone just kowtows a little and gestures at some new trade policies that favor the US and make Trump look like a master negotiator who’s bringing the world to heel.There’s been pushback against this potentiality, however, led by China, which has led with its own, very large counter-tariffs rather than negotiating, and the EU looks like it might do the same. If enough governments do this, it could call Trump’s bluff while also making these other entities, perhaps especially China, which was first out the door with counter-tariffs and statements about not be cowed by the US’s bluster, seem like the natural successors to the US in terms of global economic leadership. It could result in the US giving away all that soft power, basically, and that in turn could realign global trade relationships and ultimately other sorts of relationships, too, in China’s favor.One other commonly cited possibility, and this is maybe the grimmest of the three, but it’s not impossible, is that Trump and other people in his administration recognize that the world is changing, that China is ascendent and the US is by some metrics not competing in the way it needs to in order to keep up and retain its dominance, and that’s true in terms of things like manufacturing and research, but also the potential implications of AI, changing battlefield tactics, and so on. And from that perspective, it maybe makes sense to just shake the game board, knocking over all the pieces rather than trying to win by adhering to what have become common conventions and normal rules of play.If everyone takes a hit, if there’s a global recession or depression and everything is knocked asunder because those variables that led to where we are today, with all their associated pros and cons, are suddenly gone, that might lead to a situation in which the US is hurt, but not as badly as everyone else, including entities like China. And because the US did the game board shaking, the US may thus be in a better position as everything settles back into a new state of affairs; a new state of affairs that Trump and his people want to be more favorable to the US, long-term.There’s some logic to this thinking, even if it’s a very grim, me-first, zero-sum kind of logic. The US economy is less reliant on global trade than the rest of the G20, the wealthiest countries in the world; only about 25% of its GDP is derived from trade, while that number is 37% for China, 63% for France, and a whopping 88% for Germany.Other nations are in a relatively more vulnerable position than the US in a less-open, more tariff-heavy world, then, and that means the US administration may have them over a barrel, making the aforementioned US-favoring negotiations more likely, but also, again, potentially just hurting everyone, but the US less so. And when I say hurting, I mean some countries losing a huge chunk of their economy overnight, triggering a lot more poverty, maybe stagflation and famines, and possibly even revolutions, as people worldwide experience a shocking and sudden decrease in both wealth and future economic outlook.Already, just days after Trump announced his tariffs, global markets are crashing, with US markets on track to record its second-worst three-day decline in history, after only the crash of 1987—so that’s worse than even the crashes that followed 9/11, the Covid-19 pandemic, the debt crisis, and many others.Foreign markets are doing even worse, though, with Hong Kong’s recently high-flying Hang Seng falling 13% in trading early this week, and Japan’s Nikkei dropping 8%.Other market markers are also dropping, the price of oil falling to a pandemic-era level of $60 per barrel, Bitcoin losing 10% in a day, and even the US dollar, which theoretically should rise in a tariff scenario, dropping 0.1%—which suggests investors are planning for a damaging recession, and the US market and currency as a whole might be toxic for a while; which could, in turn, lead to a boom for the rest of the world, the US missing out on that boom.There are also simpler theories, I should mention, that tariffs may be meant to generate more profits to help pay for Trump’s expanded tax cuts without requiring he touch the third-rails of Medicare or Social Security, or that they’re meant to address the US’s booming debt by causing investors to flee to Treasury bills, which has the knock-on effect of reducing the interest rates that have to be paid on government debt.That flight toward Treasuries is already happening, though it seems to be primarily because investors are fleeing the market as stocks collapse in value and everyone’s worrying about their future, about stagflation, and about mass layoffs and unemployment.It may be that all or most of these things are true, too, by the way, and that this jumble of events, pros and cons alike, are seen as a net-positive by this administration.For what it’s worth, too, the US Presidency doesn’t typically get to set things like tariffs—that’s congress’ responsibility and right. But because Congress is currently controlled by Republicans, they’ve yet to push back on these tariffs with a veto, and they may not. There are rumblings within the president’s party about this, and a lot of statements about how it’ll ultimately be good, but that maybe they would have done things differently, but there hasn’t been any real action yet, just hedging. And that could remain the case, but if things get bad enough, they could be forced by their constituents to take concrete action on the matter before Trump’s promised, theoretical positive outcomes have the chance to emerge, or not.Show Noteshttps://www.everycrsreport.com/files/20060925_RL33665_4a8c6781ce519caa3e6b82f95c269f73021c5fdf.pdfhttps://en.wikipedia.org/wiki/Tariffhttps://www.washingtonpost.com/business/2025/03/31/tariffs-affect-consumer-spending/https://www.wsj.com/tech/exempt-or-not-the-chip-industry-wont-escape-tariffs-a6c771dbhttps://www.wsj.com/economy/central-banking/goldman-sachs-lifts-u-s-recession-probability-to-35-ce285ebchttps://www.axios.com/newsletters/axios-am-9d85eb00-1184-11f0-8b11-0da1ebc288e3.htmlhttps://apnews.com/article/trump-tariffs-democrats-economy-protests-financial-markets-90afa4079acbde1deb223adf070c1e98https://www.wsj.com/economy/trade/trade-war-explodes-across-world-at-pace-not-seen-in-decades-0b6d6513https://www.mufgamericas.com/sites/default/files/document/2025-04/The-Long-Shadow-of-William-McKinley.pdfhttps://x.com/krishnanrohit/status/1907587352157106292https://www.nytimes.com/2025/04/04/business/trump-stocks-tariffs-trade.htmlhttps://www.nytimes.com/2025/04/05/opinion/trump-tariffs-theories.htmlhttps://www.nytimes.com/2025/04/06/world/asia/vietnam-trump-tariff-delay.htmlhttps://www.nytimes.com/2025/04/06/world/europe/trade-trump-tariffs-brexit.htmlhttps://marginalrevolution.com/marginalrevolution/2025/04/why-do-domestic-prices-rise-with-tarriffs.htmlhttps://www.foxnews.com/politics/how-we-got-liberation-day-look-trumps-past-comments-tariffshttps://www.pbs.org/wgbh/frontline/article/trumps-tariff-strategy-can-be-traced-back-to-the-1980s/https://www.nytimes.com/2024/12/12/us/politics/trump-tv-stock-market.htmlhttps://www.hudsonbaycapital.com/documents/FG/hudsonbay/research/638199_A_Users_Guide_to_Restructuring_the_Global_Trading_System.pdfhttps://economictimes.indiatimes.com/news/international/us/over-50-countries-push-for-tariff-revisions-will-donald-trump-compromise-heres-what-the-white-house-said/articleshow/120043664.cmshttps://www.nytimes.com/2025/04/06/business/stock-market-plunge-investment-bank-impact.htmlhttps://www.wsj.com/livecoverage/stock-market-trump-tariffs-trade-war-04-07-25https://www.wsj.com/world/china/china-trump-tariff-foreign-policy-6934e493https://www.wsj.com/economy/in-matter-of-days-outlook-shifts-from-solid-growth-to-recession-risk-027eb2b4https://asia.nikkei.com/Business/Markets/Asia-Pacific-stocks-sink-from-Trump-s-tariff-barrage-Hong-Kong-down-13https://www.reuters.com/markets/eu-seeks-unity-first-strike-back-trump-tariffs-2025-04-06/https://www.washingtonpost.com/politics/2025/04/07/trump-presidency-news-tariffs/https://www.nytimes.com/2025/04/07/world/asia/china-trade-war-tariffs.htmlhttps://www.bloomberg.com/news/newsletters/2025-04-07/global-rout-carries-whiff-of-panic-as-trump-holds-fast-on-tariffshttps://en.wikipedia.org/wiki/Stagflationhttps://finance.yahoo.com/news/economists-fed-recent-projections-signal-120900777.htmlhttps://en.wikipedia.org/wiki/1973_oil_crisishttps://en.wikipedia.org/wiki/Economic_stagnation 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 Studio Ghibli, Andrej Karpathy, and OpenAI.We also discuss code abstraction, economic repercussions, and DOGE.Recommended Book: How To Know a Person by David BrooksTranscriptIn late-November of 2022, OpenAI released a demo version of a product they didn’t think would have much potential, because it was kind of buggy and not very impressive compared to the other things they were working on at the time. This product was a chatbot interface for a generative AI model they had been refining, called ChatGPT.This was basically just a chatbot that users could interact with, as if they were texting another human being. And the results were good enough—both in the sense that the bot seemed kinda sorta human-like, but also in the sense that the bot could generate convincing-seeming text on all sorts of subjects—that people went absolutely gaga over it, and the company went full-bore on this category of products, dropping an enterprise version in August the following year, a search engine powered by the same general model in October of 2024, and by 2025, upgraded versions of their core models were widely available, alongside paid, enhanced tiers for those who wanted higher-level processing behind the scenes: that upgraded version basically tapping a model with more feedstock, a larger training library and more intensive and refined training, but also, in some cases, a model that thinks longer, than can reach out and use the internet to research stuff it doesn’t already know, and increasingly, to produce other media, like images and videos.During that time, this industry has absolutely exploded, and while OpenAI is generally considered to be one of the top dogs in this space, still, they’ve got enthusiastic and well-funded competition from pretty much everyone in the big tech world, like Google and Amazon and Meta, while also facing upstart competitors like Anthropic and Perplexity, alongside burgeoning Chinese competitors, like Deepseek, and established Chinese tech giants like Tencent and Baidu.It’s been somewhat boggling watching this space develop, as while there’s a chance some of the valuations of AI-oriented companies are overblown, potentially leading to a correction or the popping of a valuation bubble at some point in the next few years, the underlying tech and the output of that tech really has been iterating rapidly, the state of the art in generative AI in particular producing just staggeringly complex and convincing images, videos, audio, and text, but the lower-tier stuff, which is available to anyone who wants it, for free, is also valuable and useable for all sorts of purposes.Just recently, at the tail-end of March 2025, OpenAI announced new multimodal capabilities for its GPT-4o language model, which basically means this model, which could previously only generate text, can now produce images, as well.And the model has been lauded as a sort of sea change in the industry, allowing users to produce remarkable photorealistic images just by prompting the AI—telling it what you want, basically—with usually accurate, high-quality text, which has been a problem for most image models up till this point. It also boasts the capacity to adjust existing images in all sorts of ways.Case-in-point, it’s possible to use this feature to take a photo of your family on vacation and have it rendered in the style of a Studio Ghibli cartoon; Studio Ghibli being the Japanese animation studio behind legendary films like My Neighbor Totoro, Spirited Away, and Princess Mononoke, among others.This is partly the result of better capabilities by this model, compared to its precursors, but it’s also the result of OpenAI loosening its policies to allow folks to prompt these models in this way; previously they disallowed this sort of power, due to copyright concerns. And the implications here are interesting, as this suggests the company is now comfortable showing that their models have been trained on these films, which has all sorts of potential copyright implications, depending on how pending court cases turn out, but also that they’re no long being as precious with potential scandals related to how their models are used.It’s possible to apply all sorts of distinctive styles to existing images, then, including South Park and the Simpsons, but Studio Ghibli’s style has become a meme since this new capability was deployed, and users have applied it to images ranging from existing memes to their own self-portrait avatars, to things like the planes crashing into the Twin Towers on 9/11, JFK’s assassination, and famous mass-shootings and other murders.It’s also worth noting that the co-founder of Studio Ghibli, Hayao Miyazaki, has called AI-generated artwork “an insult to life itself.” That so many people are using this kind of AI-generated filter on these images is a jarring sort of celebration, then, as the person behind that style probably wouldn’t appreciate it; many people are using it because they love the style and the movies in which it was born so much, though. An odd moral quandary that’s emerged as a result of these new AI-provided powers.What I’d like to talk about today is another burgeoning controversy within the AI space that’s perhaps even larger in implications, and which is landing on an unprepared culture and economy just as rapidly as these new image capabilities and memes.—In February of 2025, the former AI head at Tesla, founding team member at OpenAI, and founder of an impending new, education-focused project called Eureka Labs named Andrej Karpathy coined the term ‘vibe coding’ to refer to a trend he’s noticed in himself and other developers, people who write code for a living, to develop new projects using code-assistant AI tools in a manner that essentially abstracts away the code, allowing the developer to rely more on vibes in order to get their project out the door, using plain English rather than code or even code-speak.So while a developer would typically need to invest a fair bit of time writing the underlying code for a new app or website or video game, someone who’s vibe coding might instead focus on a higher, more meta-level of the project, worrying less about the coding parts, and instead just telling their AI assistant what they want to do. The AI then figures out the nuts and bolts, writes a bunch of code in seconds, and then the vibe coder can tweak the code, or have the AI tweak it for them, as they refine the concept, fix bugs, and get deeper into the nitty-gritty of things, all, again, in plain-spoken English.There are now videos, posted in the usual places, all over YouTube and TikTok and such, where folks—some of whom are coders, some of whom are purely vibe coders, who wouldn’t be able to program their way out of a cardboard box—produce entire functioning video games in a matter of minutes.These games typically aren’t very good, but they work. And reaching even that level of functionality would previously have taken days or weeks for an experienced, highly trained developer; now it takes mere minutes or moments, and can be achieved by the average, non-trained person, who has a fundamental understanding of how to prompt AI to get what they want from these systems.Ethan Mollick, who writes a fair bit on this subject and who keeps tabs on these sorts of developments in his newsletter, One Useful Thing, documented his attempts to make meaning from a pile of data he had sitting around, and which he hadn’t made the time to dig through for meaning. Using plain English he was able to feed all that data to OpenAI’s Deep Research model, interact with its findings, and further home in on meaningful directions suggested by the data.He also built a simple game in which he drove a firetruck around a 3D city, trying to put out fires before a competing helicopter could do the same. He spent a total of about $13 in AI token fees to make the game, and he was able to do so despite not having any relevant coding expertise.A guy named Pieter Levels, who’s an experienced software engineer, was able to vibe-code a video game, which is a free-to-play, massively multiplayer online flying game, in just a month. Nearly all the code was written by Cursor and Grok 3, the first of which is a code-writing AI system, the latter of which is a ChatGPT-like generalist AI agent, and he’s been able to generate something like $100k per month in revenue from this game just 17 days, post-launch.Now an important caveat here is that, first, this game received a lot of publicity, because Levels is a well-known name in this space, and he made this game as part of a ‘Vibe Coding Game Jam,’ which is an event focused on exactly this type of AI-augmented programming, in which all of the entrants had to be at least 80% AI generated. But he’s also a very skilled programmer and game-maker, so this isn’t the sort of outcome the average person could expect from these sorts of tools.That said, it’s an interesting case study that suggests a few things about where this category of tools is taking us, even if it’s not representative for all programming spaces and would-be programmers.One prediction that’s been percolating in this space for years, even before ChatGPT was released, but especially after generative AI tools hit the mainstream, is that many jobs will become redundant, and as a result many people, especially those in positions that are easily and convincingly replicated using such tools, will be fired. Because why would you pay twenty people $100,000 a year to do basic coding work when you can have one person working part-time with AI tools vibe-coding their way to approximately the same outcome?It’s a fair question, and it’s one that pretty much every industry is asking itself right now. And we’ve seen some early waves of firings based on this premise, most of which haven’t gone great for the firing entity, as they’ve then had to backtrack and starting hiring to fill those positions again—the software they expected to fill the gaps not quite there yet, and their offerings suffering as a consequence of that gambit.Some are still convinced this is the way things are going, though, including people like Elon Musk, who, as part of his Department of Government Efficiency, or DOGE efforts in the US government, is basically stripping things down to the bare-minimum, in part to weaken agencies he doesn’t like, but also, ostensibly at least, to reduce bloat and redundancy, the premise being that a lot of this work can be done by fewer people, and in some cases can be automated entirely using AI-based systems.This was the premise of his mass-firings at Twitter, now X, when he took over, and while there have been a lot of hiccups and issues resulting from that decision, the company is managing to operate, even if less optimally than before, with about 20% the staff it had before he took over—something like 1,500 people compared to 7,500.Now, there are different ways of looking at that outcome, and Musk’s activities since that acquisition will probably color some of our perceptions of his ambitions and level of success with that job-culling, as well. But the underlying theory that a company can do even 90% as well as it did before with just a fifth of the workforce is a compelling argument to many people, and that includes folks running governments, but also those in charge of major companies with huge rosters of employees that make up the vast majority of their operating expenses.A major concern about all this, though, is that even if this theory works in broader practice, and all these companies and governments can function well enough with a dramatically reduced staff using AI tools to augment their capabilities and output, we may find ourselves in a situation in which the folks using said tools are more and more commodified—they’ll be less specialized and have less education and expertise in the relevant areas, so they can be paid less, basically, the tools doing more and the humans mostly being paid to prompt and manage them. And as a result we may find ourselves in a situation where these people don’t know enough to recognize when the AI are doing something wrong or weird, and we may even reach a point where the abstraction is so complete that very few humans even know how this code works, which leaves us increasingly reliant on these tools, but also more vulnerable to problems should they fail at a basic level, at which point there may not be any humans left who are capable of figuring out what went wrong, since all the jobs that would incentivize the acquisition of such knowledge and skill will have long since disappeared.As I mentioned in the intro, these tools are being applied to images, videos, music, and everything else, as well. Which means we could see vibe artists, vibe designers, vibe musicians and vibe filmmakers. All of which is arguably good in the sense that these mediums become more accessible to more people, allowing more voices to communicate in more ways than ever before.But it’s also arguably worrying in the sense that more communication might be filtered through the capabilities of these tools—which, by the way, are predicated on previous artists and writers and filmmakers’ work, arguably stealing their styles and ideas and regurgitating them, rather than doing anything truly original—and that could lead to less originality in these spaces, but also a similar situation in which people forget how to make their own films, their own art, their own writing; a capability drain that gets worse with each new generation of people who are incentivized to hand those responsibilities off to AI tools; we’ll all become AI prompters, rather than all the things we are, currently.This has been the case with many technologies over the years—how many blacksmiths do we have in 2025, after all? And how many people actually hand-code the 1s and 0s that all our coding languages eventually write, for us, after we work at a higher, more human-optimized level of abstraction?But because our existing economies are predicated on a certain type of labor and certain number of people being employed to do said labor, even if those concerns ultimately don’t end up being too big a deal, because the benefits are just that much more impactful than the downsides and other incentives to develop these or similar skills and understandings arise, it’s possible we could experience a moment, years or decades long, in which the whole of the employment market is disrupted, perhaps quite rapidly, leaving a lot of people without income and thus a lot fewer people who can afford the products and services that are generated more cheaply using these tools.A situation that’s ripe with potential for those in a position to take advantage of it, but also a situation that could be devastating to those reliant on the current state of employment and income—which is the vast, vast majority of human beings on the planet.Show Noteshttps://en.wikipedia.org/wiki/X_Corphttps://devclass.com/2025/03/26/the-paradox-of-vibe-coding-it-works-best-for-those-who-do-not-need-it/https://www.wired.com/story/doge-rebuild-social-security-administration-cobol-benefits/https://www.wired.com/story/anthropic-benevolent-artificial-intelligence/https://arstechnica.com/tech-policy/2025/03/what-could-possibly-go-wrong-doge-to-rapidly-rebuild-social-security-codebase/https://en.wikipedia.org/wiki/Vibe_codinghttps://www.newscientist.com/article/2473993-what-is-vibe-coding-should-you-be-doing-it-and-does-it-matter/https://nmn.gl/blog/dangers-vibe-codinghttps://x.com/karpathy/status/1886192184808149383https://simonwillison.net/2025/Mar/19/vibe-coding/https://arstechnica.com/ai/2025/03/is-vibe-coding-with-ai-gnarly-or-reckless-maybe-some-of-both/https://devclass.com/2025/03/26/the-paradox-of-vibe-coding-it-works-best-for-those-who-do-not-need-it/https://www.creativebloq.com/3d/video-game-design/what-is-vibe-coding-and-is-it-really-the-future-of-app-and-game-developmenthttps://arstechnica.com/ai/2025/03/openais-new-ai-image-generator-is-potent-and-bound-to-provoke/https://en.wikipedia.org/wiki/Studio_Ghibli 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