worlds will collide Go back to the Grid in TRON: Ares trailer It's more of a standalone reboot rather than a direct sequel to 2010's TRON: Legacy. Jennifer Ouellette Apr 5, 2025 2:24 pm | 30 Let there be light cycles. Credit: Walt Disney Studios Let there be light cycles. Credit: Walt Disney Studios Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn more An AI program enters the real world in TRON: Ares. It's difficult to underestimate the massive influence that Disney's 1982 cult science fiction film, TRON, had on both the film industrythanks to combining live action with what were then groundbreaking visual effects, rife with computer-generated imageryand on nerd culture at large. Over the ensuing decades there has been one sequel, an animated TV series, a comic book miniseries, video games, and theme park attractions, all modeled on director Steve Lisberg's original fictional world.Now we're getting a third installment in the film franchise: TRON: Ares, directed by Joachim Rnning (Pirates of the Caribbean: Dead Men Tell No Tales, Maleficent: Mistress of Evil), that serves as a standalone sequel to 2010's TRON: Legacy. Disney just released the first trailer and poster art, and while the footage is short on plot, it's got the show-stopping visuals we've come to expect from all things TRON.(Spoilers for ending of TRON: Legacy below.)TRON: Legacy ended with Sam Flynn (Garrett Hedlund), son of Kevin Flynn (Jeff Bridges) from the original film, preventing the digital world from bleeding into the real world, as planned by the Grid's malevolent ruling program, Clu. He brought with him Quorra (Olivia Wilde), a naturally occurring isomorphic algorithm targeted for extinction by Clu. Jared Leto stars as the human form of Ares, a sophisticated computer program sent into the real world. Walt Disney Studios Jared Leto stars as the human form of Ares, a sophisticated computer program sent into the real world. Walt Disney Studios Eve Kim (Greta Lee) runs as digital and real worlds collide Walt Disney Studios Eve Kim (Greta Lee) runs as digital and real worlds collide Walt Disney Studios Jared Leto stars as the human form of Ares, a sophisticated computer program sent into the real world. Walt Disney Studios Eve Kim (Greta Lee) runs as digital and real worlds collide Walt Disney Studios Disney greenlit a third film in the franchise in October 2010, intended to pick up where Legacy left off and follow the adventures of Sam and Quorra as Sam took full control of his father's company, ENCOM. But by 2015, the studio had cancelled the project, allegedly the result of the dismal box office showing of Tomorrowland. By 2020, the project had been revived and reimagined as a standalone reboot rather than a Legacy sequel, although the main AI, Ares, did appear in earlier (pre-reboot) versions of the script. One pandemic and a couple of Hollywood strikes later, and the finished film is finally set to hit theaters this fall.The official premise is quite succinct: "TRON: Ares follows a highly sophisticated Program, Ares, who is sent from the digital world into the real world on a dangerous mission, marking humankinds first encounter with A.I. beings." Jared Leto stars as Ares, with Evan Peters and Greta Lee playing Julian Dillinger and Eve Kim, respectively. The cast also includes Jodie Turner-Smith, Cameron Monaghan, Sarah Desjardins, Hasan Minhaj, Arturo Castro, and Gillian Anderson in as-yet-undisclosed roles. Bridges is returning as Kevin Flynn. Nine Inch Nails composed the soundtrack.TRON: Ares hits theaters on October 10, 2025. Credit: Walt Disney Studios Jennifer OuelletteSenior WriterJennifer OuelletteSenior Writer Jennifer is a senior writer at Ars Technica with a particular focus on where science meets culture, covering everything from physics and related interdisciplinary topics to her favorite films and TV series. Jennifer lives in Baltimore with her spouse, physicist Sean M. Carroll, and their two cats, Ariel and Caliban. 30 Comments

Anti-extinction vs. de-extinction Editorial: Mammoth de-extinction is bad conservation Ecosystems are inconveniently complex, and elephants won't make good surrogates. Nitin Sekar Apr 5, 2025 7:05 am | 12 Are we ready for mammoths when we can't handle existing human-pachyderm conflicts? Credit: chuchart duangdaw Are we ready for mammoths when we can't handle existing human-pachyderm conflicts? Credit: chuchart duangdaw Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreThe start-up Colossal Biosciences aims to use gene-editing technology to bring back the woolly mammoth and other extinct species. Recently, the company achieved major milestones: last year, they generated stem cells for the Asian elephant, the mammoths closest living relative, and this month they published photos of genetically modified mice with long, mammoth-like coats. According to the companys founders, including Harvard and MIT professor George Church, these advances take Colossal a big step closer to their goal of using mammoths to combat climate change by restoring Arctic grassland ecosystems. Church also claims that Colossals woolly mammoth program will help protect endangered species like the Asian elephant, saying were injecting money into conservation efforts.In other words, the scientific advances Colossal makes in their lab will result in positive changes from the tropics to the Arctic, from the soil to the atmosphere.Colossals Jurassic Park-like ambitions have captured the imagination of the public and investors, bringing its latest valuation to $10 billion. And the companys research does seem to be resulting in some technical advances. But Id argue that the broader effort to de-extinct the mammoth isas far as conservation efforts goincredibly misguided. Ultimately, Colossals efforts wont end up being about helping wild elephants or saving the climate. Theyll be about creating creatures for human spectacle, with insufficient attention to the costs and opportunity costs to human and animal life.Shaky evidenceThe Colossal website explains how they believe resurrected mammoths could help fight climate change: cold-tolerant elephant mammoth hybrids grazing the grasslands [will] scrape away layers of snow, so that the cold air can reach the soil. This will reportedly help prevent permafrost from melting, blocking the release of greenhouse gasses currently trapped in the soil. Furthermore, by knocking down trees and maintaining grasslands, Colossal says, mammoths will help slow snowmelt, ensuring Arctic ecosystems absorb less sunlight.Conservationists often claim that the reason to save charismatic species is that they are necessary for the sound functioning of the ecosystems that support humankind. Perhaps the most well-known of these stories is about the ecological changes wolves drove when they were reintroduced to Yellowstone National Park. Through some 25 peer-reviewed papers, two ecologists claimed to demonstrate that the reappearance of wolves in Yellowstone changed the behavior of elk, causing them to spend less time browsing the saplings of trees near rivers. This led to a chain of cause and effect (a trophic cascade) that affected beavers, birds, and even the flow of the river. A YouTube video on the phenomenon called How Wolves Change Rivers has been viewed more than 45 million times.But other scientists were unable to replicate these findingsthey discovered that the original statistics were flawed, and that human hunters likely contributed to elk population declines in Yellowstone.Ultimately, a 2019 review of the evidence by a team of researchers concluded that the most robust science suggests trophic cascades are not evident in Yellowstone. Similar ecological claims about tigers and sharks as apex predators also fail to withstand scientific scrutiny.Elephantswidely described as keystone speciesare also stars of a host of similar ecological stories. Many are featured on the Colossal website, including one of the most common claims about the role elephants play in seed dispersal. Across all environments, reads the website, elephant dung filled with seeds serve to spread plants [] boosting the overall health of the ecosystem. But would the disappearance of elephants really result in major changes in plant life? After all, some of the worlds grandest forests (like the Amazon) have survived for millennia after the disappearance of mammoth-sized megafauna.For my PhD research in northeast India, I tried to systematically measure how important Asian elephants were for seed dispersal compared to other animals in the ecosystem; our teams work, published in five peer-reviewed ecological journals (reviewed here), does find that elephants are uniquely good at dispersing the seeds of a few large-fruited species. But we also found that domestic cattle and macaques disperse some species seeds quite well, and that 80 percent of seeds dispersed in elephant dung end up eaten by ants. After several years of study, I cannot say with confidence that the forests where I worked would be drastically different in the absence of elephants.The evidence for how living elephants affect carbon sequestration is also quite mixed. On the one hand, one paper finds that African forest elephants knock down softwood trees, making way for hardwood trees that sequester more carbon. But on the other hand, many more researchers looking at African savannas have found that elephants knock down lots of trees, converting forests into savannas and reducing carbon sequestration.Colossals website offers links to peer-reviewed research that support their suppositions on the ecological role of woolly mammoths. A key study offers intriguing evidence that keeping large herbivoresreindeer, Yakutian horses, moose, musk ox, European bison, yaks, and cold-adapted sheepat artificially high levels in a tussock grassland helped achieve colder ground temperatures, ostensibly protecting permafrost. But the study raises lots of questions: is it possible to boost these herbivores populations across the whole northern latitudes? If so, why do we need mammoths at allwhy not just use species that already exist, which would surely be cheaper?Plus, as ecologist Michelle Mack noted, as the winters warm due to climate change, too much trampling or sweeping away of snow could have the opposite effect, helping warm the soils underneath more quicklyif so, mammoths could be worse for the climate, not better.All this is to say that ecosystems are diverse and messy, and those of us working in functional ecology dont always discover consistent patterns. Researchers in the field often struggle to find robust evidence for how a living species affects modern-day ecosystemssurely it is far harder to understand how a creature extinct for around 10,000 years shaped its environment? And harder still to predict how it would shape tomorrows ecosystems? In effect, Colossals ecological narrative relies on that difficulty. But just because claims about the distant past are harder to fact-check doesnt mean they are more likely to be true.Ethical blind spotsColossals website spells out 10 steps for mammoth resurrection. Steps nine and 10 are: implant the early embryo into the healthy Asian or African elephant surrogates, and care for the surrogates in a world-class conservation facility for the duration of the gestation and afterward.Colossals cavalier plans to use captive elephants as surrogates for mammoth calves illustrate an old problem in modern wildlife conservation: indifference towards individual animal suffering. Leading international conservation NGOs lack animal welfare policies that would push conservationists to ask whether the costs of interventions in terms of animal welfare outweigh the biodiversity benefits. Over the years, that absence has resulted in a range of questionable decisions.Colossals efforts take this apathy towards individual animals into hyperdrive. Despite societys thousands of years of experience with Asian elephants, conservationists struggle to breed them in captivity. Asian elephants in modern zoo facilities suffer from infertility and lose their calves to stillbirth and infanticides almost twice as often as elephants in semi-wild conditions. Such problems will almost certainly be compounded when scientists try to have elephants deliver babies created in the lab, with a hodge podge of features from Asian elephants and mammoths. Credit: Paul Gilham Even in the best-case scenario, there would likely be many, many failed efforts to produce a viable organism before Colossal gets to a herd that can survive. This necessarily trial-and-error process could lead to incredible suffering for both elephant mothers and mammoth calves along the way. Elephants in the wild have been observed experiencing heartbreaking grief when their calves die, sometimes carrying their babies corpses for daysa grief the mother elephants might very well be subjected to as they are separated from their calves or find themselves unable to keep their chimeric offspring alive.For the calves that do survive, their edited genomes could lead to chronic conditions, and the ancient mammoth gut microbiome might be impossible to resurrect, leading to digestive dysfunction. Then there will likely be social problems. Research finds that Asian elephants in Western zoos dont live as long as wild elephants, and elephant researchers often bemoan the limited space, stimulation, and companionship available to elephants in captivity. These problems will surely also plague surviving animals.Introduction to the wild will probably result in even more suffering: elephant experts recommend against introducing captive animals that have had no natural foraging experience at all to the wild as they are likely to experience significant hardship. Modern elephants survive not just through instinct, but through culturematriarch-led herds teach calves what to eat and how to survive, providing a nurturing environment. We have good reason to believe mammoths also needed cultural instruction to survive. How many elephant/mammoth chimeras will suffer false starts and tragic deaths in the punishing Arctic without the social conditions that allowed them to thrive millennia ago?Opportunity costsIf Colossal (or Colossals investors) really wish to foster Asian elephant conservation or combat climate change, they have many better options. The opportunity costs are especially striking for Asian elephant conservation: while over a trillion dollars is spent combatting climate change annually, the funds available to address the myriad of problems facing wild Asian elephants are far smaller. Take the example of India, the country with the largest population of wild Asian elephants in the world (estimated at 27,000) in a sea of 1.4 billion human beings.Indians generally revere elephants and tolerate a great deal of hardship to enable coexistenceabout 500 humans are killed due to human-elephant conflict annually there. But as a middle-income country continuing to struggle with widespread poverty, the federal government typically budgets less than $4M for Project Elephant, its flagship elephant conservation program. Thats less than $200 per wild elephant and 1/2000th as much as Colossal has raised so far. Indias conservation NGOs generally have even smaller budgets for their elephant work. The result is that conservationists are a decade behindwhere they expected to be in mapping where elephants range.With Colossals budget, Asian elephant conservation NGOs could tackle the real threats to the survival of elephants: human-elephant conflict, loss of habitat and connectivity, poaching, and the spread of invasive plants unpalatable to elephants. Some conservationists are exploring creative schemes to help keep people and elephants safe from each other. There are also community-based efforts toremove invasive species like Lantana camara and restore native vegetation. Funds could enable development of an AI-powered system that allows the automated identification and monitoring of individual elephants. There is also a need for improved compensation schemes to ensure those who lose crops or property to wild elephants are made whole again.As a US-based synthetic biology company, Colossal could also use its employees skills much more effectively to fight climate change. Perhaps they could genetically engineer trees and shrubs to sequester more carbon. Or Colossal could help us learn to produce meat from modified microbes or cultivated lines of cow, pig, and chicken cells, developing alternative proteins that could more efficiently feed the planet, protecting wildlife habitat and reducing greenhouse gas emissions.The question is whether Colossals leaders and supporters are willing to pivot from a project that grabs news headlines to ones that would likely make positive differences. By tempting us with the resurrection of a long-dead creature, Colossal forces us to ask: do we want conservation to be primarily about feeding an unreflective imagination? Or do we want evidence, logic, and ethics to be central to our relationships with other species? For anyone who really cares about the climate, elephants, or animals in general, de-extincting the mammoth represents a huge waste and a colossal mistake.Nitin Sekar served as the national lead for elephant conservation at WWF India for five years and is now a member of the Asian Elephant Specialist Group of the International Union for the Conservation of Natures Species Survival Commission The views presented here are his own. 12 Comments

Supply and demand and zombies and viruses Dustland Delivery plays like a funny, tough, post-apocalyptic Oregon Trail Buy low, sell high, fight raiders, and fix tires in this quirky RPG. Kevin Purdy Apr 5, 2025 7:30 am | 0 Dustland Delivery cover art Credit: Lilith Games Dustland Delivery cover art Credit: Lilith Games Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreRoad trips with just two people always have their awkward silences. InDustland Delivery, my character, a sharpshooter, has tried to break the ice with the blacksmith he hired a few towns back, with only intermittent success.Remember that bodyguard, the one I unsuccessfully tried to flirt with at that bar? The blacksmith was uninterested. What about that wily junk dealer, or the creepy cemetery? Silence. She only wanted to discuss "Abandoned train" and "Abandoned factory," even though, in this post-apocalypse, abandonment was not that rare. But I made a note to look out for any rusted remains; stress and mood are far trickier to fix than hunger and thirst.Dustland Delivery release trailer. Dustland Delivery, available through Steam for Windows (and Proton/Steam Deck), puts you in the role typically taken up by NPCs in other post-apocalyptic RPGs. You're a trader, buying cheap goods in one place to sell at a profit elsewhere, and working the costs of fuel, maintenance, and raider attacks into your margins. You're in charge of everything on your trip: how fast you drive, when to rest and set up camp, whether to approach that caravan of pickups or give them a wide berth.Some of you, the types whose favorite part ofThe Oregon Trail was the trading posts, might already be sold. For the others, let me suggest that the game is stuffed full of little bits of weird humor and emergent storytelling, and a wild amount of replayability for what is currently a $5 game. There are three quest-driven scenarios, plus a tutorial, in the base game. A new DLC out this week, Sheol, adds underground cities, ruins expeditions, more terrains, and a final story quest for four more dollars. I thought it was a huge decision whether to drive over or around that river bed, but you couldn't care less! Lilith Games I thought it was a huge decision whether to drive over or around that river bed, but you couldn't care less! Lilith Games You can do a lot of things with a lot of people in each settlement's bar, but you won't necessarily succeed. Lilith Games You can do a lot of things with a lot of people in each settlement's bar, but you won't necessarily succeed. Lilith Games Get into the game a bit more, and you can go wild in the character creation screen. Lilith Games Get into the game a bit more, and you can go wild in the character creation screen. Lilith Games You can do a lot of things with a lot of people in each settlement's bar, but you won't necessarily succeed. Lilith Games Get into the game a bit more, and you can go wild in the character creation screen. Lilith Games Towns typically have a store, a bar, and a point of interest, and also some street folks who say snarky things. Lilith Games Graciously, Dustland Delivery does not make you do the math when buying or selling. Lilith Games Driven to surviveI'm about five hours into this game, and so far, like Steam reviewers, my impressions are Very Positive. The interface is pixellated and a bit irregular, forcing you to learn where to click to get out of different screens. The tutorial mission is helpful to get started, but you'll definitely have to discover a lot of mechanics and failure states by trial and error. The basic gameplay loop, however, seems well-tested and considered.Dustland Delivery playslike a mix of FTL: Faster than Light and the overland travel mechanics of fantasy RPGs, mixed with the random encounters and gritty humanity of the 2D Fallout titles. You're constantly taking risks and running imprecise calculations. Should you push your truck's tire wear and engine coolant, and your team's hunger and fatigue, to make a big score on quail meat so that you can upgrade your hull armor? Or will that put you in a part of the map where fuel costs a lot and the terrain slows you down? How much virus exposure will you risk for a shortcut through the main quest?All that, too, is before you establish your own settlement, staffed by your truck's crew, later in the game.Dustland Deliverywill pull you in for some long sessions, but it easily fits into work breaks and "Just one more city" sprints, on just about any PC that can run Steam. Jump in the cab, haul some goods, and ask your cabmates later what they think of those mutants you fought 300 kilometers back.Kevin PurdySenior Technology ReporterKevin PurdySenior Technology Reporter Kevin is a senior technology reporter at Ars Technica, covering open-source software, PC gaming, home automation, repairability, e-bikes, and tech history. He has previously worked at Lifehacker, Wirecutter, iFixit, and Carbon Switch. 0 Comments

DOS=HIGH Microsoft turns 50 today, and it made me think about MS-DOS 5.0 A story about an obsolete PC, an old library book, and a one version of MS-DOS. Andrew Cunningham Apr 4, 2025 3:37 pm | 49 The first version of Microsoft's logo, used between 1975 and 1980 (though early versions split "Micro" and "Soft" by putting them on separate lines). Credit: Microsoft The first version of Microsoft's logo, used between 1975 and 1980 (though early versions split "Micro" and "Soft" by putting them on separate lines). Credit: Microsoft Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreOn this day in 1975, Bill Gates and Paul Allen founded a company called Micro-Soft in Albuquerque, New Mexico.The two men had worked together before, as members of the Lakeside Programming group in the early 70s and as co-founders of a road traffic analysis company called Traf-O-Data. But Micro-Soft, later renamed to drop the hyphen and relocated to its current headquarters in Redmond, Washington, would be the company that would transform personal computing over the next five decades.I'm not here to do a history of Microsoft, because Wikipedia already exists and because the company has already put together a gauzy 50th-anniversary retrospective site with some retro-themed wallpapers. But the anniversary did make me try to remember which Microsoft product I consciously used for the first time, the one that made me aware of the company and the work it was doing.To get the answer, just put a decimal point in the number "50"my first Microsoft product was MS-DOS 5.0.Riding with DOS in the Windows eraI remember this version of MS-DOS so vividly because it was the version that we ran on our first computer. I couldnt actually tell you what computer it was, though, not because I dont remember it but because it was a generic yellowed hand-me-down that was prodigiously out of date, given to us by well-meaning people from our church who didn't know enough to know how obsolete the system was.It was a clone of the original IBM PC 5150, initially released in 1981; I believe we took ownership of it sometime in 1995 or 1996. It had an Intel 8088, two 5.25-inch floppy drives, and 500-something KB of RAM (also, if memory serves, a sac of spider eggs). But it had no hard drive inside, meaning that anything I wanted to run on or save from this computer needed to use a pile of moldering black plastic diskettes, more than a few of which were already going bad.Bear in mind that my knowledge of and exposure to computers before this was minimal, and my exposure to PCs had been nearly nilI'd played Oregon Trail on an Apple IIe in elementary school, and the few computers that were in my classrooms or computer labs at the time were Macs. So all I really knew was that this computer was old but it was ours, and I set about learning what I could about it, blissfully unaware of its obsolescence and the fact that the computing world was well into the Windows 95 era.The operating system diskette for this PC just happened to be running MS-DOS 5.0, so that was where I started.MS-DOS 5.0 is actually a fairly significant release of the operating system. Released in June 1991, MS-DOS 5.0 was the first version to include Edit and Qbasic; it was also the last version to be jointly developed by Microsoft and IBM before the two companies formally parted ways. It was a welcome corrective after the ambitious-but-busted MS-DOS 4.0 release, with new features and higher RAM requirements but few of the same compatibility problems. It was also the first version offered as a shrink-wrapped retail product that PC owners could buy as a standalone upgrade, which might explain why that old PC was running it in the first place. I distinctly remember these 5.25-inch MS-DOS 5.0 upgrade disks, since a set of them came with our first-ever computer. Credit: Internet Archive/Microsoft Armed with nothing more than a version number, the next time we went to our local public library, I hit the boring non-fiction shelves to try and find more information. My memory is fuzzy, but I think the reference book I found was Carolyn Z. Gillay's DOS 5 Fundamentals, because I remember it having a sort of toothpaste-green colored cover and that it came with disks that had sample files on them. Whatever book it was, the fact that I found a tome that just happened to be about the exact version of DOS I was working with felt lucky (it was not a very up-to-date part of the library).That book taught me how to do all the stuff I couldn't figure out on this horrible old PCsearching through files and folders, writing snarky journal entries in the Edit app, copying and formatting disks (and, crucially, checking them for errors so I could weed out the bad ones), writing batch files, and creating additional system disks so we could have backups. I had the free time and the elastic, absorbent brain that only a kid can have, and I quickly soaked it all up.The information I was acquiring was already outdated by the standards of the day. But a working knowledge of DOS remained useful for decades afterward. A couple of years later when we got our first Windows PC (another hand-me-down, a similarly generic 486DX-based PC that did OK with Windows 95 but buckled under the load of Windows 98), MS-DOS was still there under the surface, and knowing how to use it was essential for the inevitable OS reinstallations. If anything, knowing about the MS-DOS undercurrents lurking beneath the surface of those Windows versions helped keep me curious about the underpinnings of other electronics I used, something that led me pretty much directly to my first IT jobs and then to writing detailed and overlong articles about technology.Microsoft is far from a perfect company. Strong-arm tactics and anticompetitive practices defined the MS-DOS and Windows eras, and I have all kinds of misgivings about the company's push into generative AI and the current state of Windows 11. But I owe a lot to that first computer and the software that it unhappily, begrudgingly ran. And that's why I found myself thinking about my strange, nostalgic attachment to a particular version of MS-DOS on Microsoft's 50th anniversary.Other stuff to do and readToday, you can relive the experience of using all kinds of MS-DOS versionsas well as PC-DOS, DR-DOS, and many retro Windows versionsin your web browser using the emulators hosted at PCjs machines.For further reading, here are a few other Microsoft and Microsoft-related retrospectives from the Ars archives:The complete history of the IBM PC, parts one and twoThe Windows Start menu saga, from 1993 up through the launch of Windows 10.Exploring modern retro-computing with the Book 8088 and Pocket 386 laptops.The rise and fall of Internet Explorer, which was finally laid to rest in 2022."Too much and too soon," a look back at Windows 8Microsoft open-sources MS-DOS 4.00An interview on the 30th anniversary of FreeDOS, an operating system that keeps the DOS software running on newer hardware Listing image: Samuel Axon Andrew CunninghamSenior Technology ReporterAndrew CunninghamSenior Technology Reporter Andrew is a Senior Technology Reporter at Ars Technica, with a focus on consumer tech including computer hardware and in-depth reviews of operating systems like Windows and macOS. Andrew lives in Philadelphia and co-hosts a weekly book podcast called Overdue. 49 Comments

Triopoly With new contracts, SpaceX will become the US militarys top launch provider The military's stable of certified rockets will include Falcon 9, Falcon Heavy, Vulcan, and New Glenn. Stephen Clark Apr 4, 2025 11:33 pm | 0 A SpaceX Falcon Heavy rocket lifts off on June 25, 2024, with a GOES weather satellite for NOAA. Credit: SpaceX A SpaceX Falcon Heavy rocket lifts off on June 25, 2024, with a GOES weather satellite for NOAA. Credit: SpaceX Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreThe US Space Force announced Friday it selected SpaceX, United Launch Alliance, and Blue Origin for $13.7 billion in contracts to deliver the Pentagon's most critical military to orbit into the early 2030s.These missions will launch the government's heaviest national security satellites, like the National Reconnaissance Office's large bus-sized spy platforms, and deploy them into bespoke orbits. These types of launches often demand heavy-lift rockets with long-duration upper stages that can cruise through space for six or more hours.The contracts awarded Friday are part of the next phase of the military's space launch program once dominated by United Launch Alliance, the 50-50 joint venture between legacy defense contractors Boeing and Lockheed Martin.After racking up a series of successful launches with its Falcon 9 rocket more than a decade ago, SpaceX sued the Air Force for the right to compete with ULA for the military's most lucrative launch contracts. The Air Force relented in 2015 and allowed SpaceX to bid. Since then, SpaceX has won more than 40 percent of missions the Pentagon has ordered through the National Security Space Launch (NSSL) program, creating a relatively stable duopoly for the military's launch needs.The Space Force took over the responsibility for launch procurement from the Air Force after its creation in 2019. The next year, the Space Force signed another set of contracts with ULA and SpaceX for missions the military would order from 2020 through 2024. ULA's new Vulcan rocket initially won 60 percent of these missionsknown as NSSL Phase 2but the Space Force reallocated a handful of launches to SpaceX after ULA encountered delays with Vulcan.ULA's Vulcan and SpaceX's Falcon 9 and Falcon Heavy rockets will launch the remaining 42 Phase 2 missions over the next several years, then move on to Phase 3, which the Space Force announced Friday.Spreading the wealthThis next round of Space Force launch contracts will flip the script, with SpaceX taking the lion's share of the missions. The breakdown of the military's new firm fixed-price launch agreements goes like this:SpaceX will get 28 missions worth approximately $5.9 billionULA will get 19 missions worth approximately $5.4 billionBlue Origin will get seven missions worth approximatelyThat equates to a 60-40 split between SpaceX and ULA for the bulk of the missions. Going into the competition, military officials set aside seven additional missions to launch with a third provider, allowing a new player to gain a foothold in the market. The Space Force reserves the right to reapportion missions between the three providers if one of them runs into trouble.The Pentagon confirmed an unnamed fourth company also submitted a proposal, but wasn't selected for Phase 3.Rounded to the nearest million, the contract with SpaceX averages out to $212 million per launch. For ULA, it's $282 million, and Blue Origin's price is $341 million per launch. But take these numbers with caution. The contracts include a lot of bells and whistles, pricing them higher than what a commercial customer might pay.According to the Pentagon, the contracts provide "launch services, mission unique services, mission acceleration, quick reaction/anomaly resolution, special studies, launch service support, fleet surveillance, and early integration studies/mission analysis."Essentially, the Space Force is paying a premium to all three launch providers for schedule priority, tailored solutions, and access to data from every flight of each company's rocket, among other things. New Glenn lifts off on its debut flight. Credit: Blue Origin "Winning 60% percent of the missions may sound generous, but the reality is that all SpaceX competitors combined cannot currently deliver the other 40%!," Elon Musk, SpaceX's founder and CEO, posted on X. "I hope they succeed, but they arent there yet."This is true if you look at each company's flight rate. SpaceX has launched Falcon 9 and Falcon Heavy rockets 140 times over the last 365 days. These are the flight-proven rockets SpaceX will use for its share of Space Force missions.ULA has logged four missions in the same period, but just one with the Vulcan rocket it will use for future Space Force launches. And Blue Origin, Jeff Bezos's space company, launched the heavy-lift New Glenn rocket on its first test flight in January."We are proud that we have launched 100 national security space missions and honored to continue serving the nation with our new Vulcan rocket," said Tory Bruno, ULAs president and CEO, in a statement.ULA used the Delta IV and Atlas V rockets for most of the missions it has launched for the Pentagon. The Delta IV rocket family is now retired, and ULA will end production of the Atlas V rocket later this year. Now, ULA's Vulcan rocket will take over as the company's sole launch vehicle to serve the Pentagon. ULA aims to eventually ramp up the Vulcan launch cadence to fly up to 25 times per year.After two successful test flights, the Space Force formally certified the Vulcan rocket last week, clearing the way for ULA to start using it for military missions in the coming months. While SpaceX has a clear advantage in number of launches, schedule assurance, and pricingand reliability comparable to ULABruno has recently touted the Vulcan rocket's ability to maneuver over long periods in space as a differentiator."This award constitutes the most complex missions required for national security space," Bruno said in a ULA press release. "Vulcan continues to use the worlds highest energy upper stage: the Centaur V. Centaur V's unmatched flexibility and extreme endurance enables the most complex orbital insertions continuing to advance our nations capabilities in space."Blue Origin's New Glenn must fly at least one more successful mission before the Space Force will certify it for Lane 2 missions. The selection of Blue Origin on Friday suggests military officials believe New Glenn is on track for certification by late 2026."Honored to serve additional national security missions in the coming years and contribute to our nations assured access to space," Dave Limp, Blue Origin's CEO, wrote on X. "This is a great endorsement of New Glenn's capabilities, and we are committed to meeting the heavy lift needs of our US DoD and intelligence agency customers."Navigating NSSLThere's something you must understand about the way the military buys launch services. For this round of competition, the Space Force divided the NSSL program into two lanes.Friday's announcement covers Lane 2 for traditional military satellites that operate thousands of miles above the Earth. This bucket includes things like GPS navigation satellites, NRO surveillance and eavesdropping platforms, and strategic communications satellites built to survive a nuclear war. The Space Force has a low tolerance for failure with these missions. Therefore, the military requires rockets be certified before they can launch big-ticket satellites, each of which often cost hundreds of millions, and sometimes billions, of dollars.The Space Force required all Lane 2 bidders to show their rockets could reach nine "reference orbits" with payloads of a specified mass. Some of the orbits are difficult to reach, requiring technology that only SpaceX and ULA have demonstrated in the United States. Blue Origin plans to do so on a future flight. This image shows what the Space Force's fleet of missile warning and missile tracking satellites might look like in 2030, with a mix of platforms in geosynchronous orbit, medium-Earth orbit, and low-Earth orbit. The higher orbits will require launches by "Lane 2" providers. Credit: Space Systems Command The military projects to order 54 launches in Lane 2 from this year through 2029, with announcements each October of exactly which missions will go to each launch provider. This year, it will be just SpaceX and ULA. The Space Force said Blue Origin won't be eligible for firm orders until next year. The missions would launch between 2027 and 2032."America leads the world in space launch, and through these NSSL Phase 3 Lane 2 contracts, we will ensure continued access to this vital domain," said Maj. Gen. Stephen Purdy, Acting Assistant Secretary of the Air Force for Space Acquisition and Integration. "These awards bolster our ability to launch critical defense satellites while strengthening our industrial base and enhancing operational readiness."Lane 1 is primarily for missions to low-Earth orbit. These payloads include tech demos, experimental missions, and the military's mega-constellation of missile tracking and data relay satellites managed by the Space Development Agency. For Lane 1 missions, the Space Force won't levy the burdensome certification and oversight requirements it has long employed for national security launches. The Pentagon is willing to accept more risk with Lane 1, encompassing at least 30 missions through the end of the 2020s, in an effort to broaden the military's portfolio of launch providers and boost competition.Last June, Space Systems Command chose SpaceX, ULA, and Blue Origin for eligibility to compete for Lane 1 missions. SpaceX won all nine of the first batch of Lane 1 missions put up for bids. The military recently added Rocket Lab's Neutron rocket and Stoke Space's Nova rocket to the Lane 1 mix. Neither of those rockets have flown, and they will need at least one successful launch before approval to fly military payloads.The Space Force has separate contract mechanisms for the military's smallest satellites, which typically launch on SpaceX rideshare missions or dedicated launches with companies like Rocket Lab and Firefly Aerospace.Military leaders like having all these options, and would like even more. If one launch provider or launch site is unavailable due to a technical problemor, as some military officials now worry, an enemy attackcommanders want multiple backups in their toolkit. Market forces dictate that more competition should also lower prices."A robust and resilient space launch architecture is the foundation of both our economic prosperity and our national security," said US Space Force Chief of Space Operations Gen. Chance Saltzman. "National Security Space Launch isn't just a program; it's a strategic necessity that delivers the critical space capabilities our warfighters depend on to fight and win."Stephen ClarkSpace ReporterStephen ClarkSpace Reporter Stephen Clark is a space reporter at Ars Technica, covering private space companies and the worlds space agencies. Stephen writes about the nexus of technology, science, policy, and business on and off the planet. 0 Comments

Passing on the costs Not just Switch 2: ESA warns Trumps tariffs will hurt the entire game industry "[It's] hard to imagine a world where tariffs like these dont impact pricing." Kyle Orland Apr 4, 2025 4:36 pm | 29 Credit: Aurich Lawson | Getty Images Credit: Aurich Lawson | Getty Images Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreThis morning's announcement that Nintendo is delaying US preorders for the Switch 2 immediately increased the salience of President Trump's proposed wide-reaching import tariffs for millions of American Nintendo fans. Additionally, the Entertainment Software Associationa lobbying group that represents the game industry's interests in Washingtonis warning that the effects of Trump's tariffs on the gaming world won't stop with Nintendo."There are so many devices we play video games on," ESA senior vice president Aubrey Quinn said in an interview with IGN just as Nintendo's preorder delay news broke. "There are other consoles... VR headsets, our smartphones, people who love PC games; if we think it's just the Switch, then we aren't taking it seriously."This is company-agnostic, this is an entire industry," she continued. "There's going to be an impact on the entire industry."While Trump's tariff proposal includes a 10 percent tax on imports from pretty much every country, it also includes a 46 percent tariff on Vietnam and a 54 percent total tariff on China, the two countries where most console hardware is produced. Quinn told IGN that it's "hard to imagine a world where tariffs like these dont impact pricing" for those consoles.More than that, though, Quinn warns that massive tariffs would tamp down overall consumer spending, which would have knock-on effects for game industry revenues, employment, and research and development investment."Video game consoles are sold under tight margins in order to reduce the barrier to entry for consumers," the ESA notes in its issue page on tariffs. "Tariffs mean that the additional costs would be passed along to consumers, resulting in a ripple effect of harm for the industry and the jobs it generates and supports.Not just a foreign problemThe negative impacts wouldn't be limited to foreign companies like Nintendo, Quinn warned, because "even American-based companies, they're getting products that need to cross into American borders to make those consoles, to make those games. And so there's going to be a real impact regardless of company."Some might argue that video game companies (and others) should simply bring more of their production costs within the US to avoid paying Trump's tariffs. But in an interview with Game File on Wednesday, Quinn noted that "supply chains are complicated and, certainly, supply chains dont change overnight. Everything that is considered or decided cant be a quick turnaround and cant be a knee-jerk reaction to any particular announcement."Last month, Circana analyst Mat Piscatella warned that proposed (and later largely delayed) 25 percent tariffs on imports from Mexicowhere the vast majority of physical video games are producedcould lead to "a sharp downtick in the number of disc-based games that get released physically in the US." Replacing that Mexican disc production capacity with domestic alternatives would take "significant investment" in a market segment that is "now half what it was in 2021 and declining rapidly," Piscatella said.Making your voice heardLate last month, the ESA joined in a multi-industry letter to US Trade Representative Jamieson Greer saying that "U.S. tariffs on imports of critical technology inputs and products would harm the very U.S. businesses the President seeks to boost and would risk undercutting long-term U.S. technology leadership." The letter urged Greer to "promote the global engagement with our trading partners" and to "use existing trade tools that strengthen trade relations with key markets for U.S. products and services" instead of tariffs.In 2020, the ESA partnered with the Consumer Technology Association to successfully argue for an exemption to tariffs then being imposed on China. Trump has recently signaled that similar industry-specific exemptions may be possible this time around as well.While Quinn told IGN that the ESA is meeting with employees at the White House and US Trade Representative's office, she said those talks are more likely to make an impact if "more members of government, elected officials, and their staff ... hear that their constituents are concerned."Kyle OrlandSenior Gaming EditorKyle OrlandSenior Gaming Editor Kyle Orland has been the Senior Gaming Editor at Ars Technica since 2012, writing primarily about the business, tech, and culture behind video games. He has journalism and computer science degrees from University of Maryland. He once wrote a whole book about Minesweeper. 29 Comments

Image Synthesis Midjourney introduces first new image generation model in over a year The new model is now in public alpha and has personalization enabled by default. Samuel Axon Apr 4, 2025 6:34 pm | 12 Midjourney V7 claims to be much more consistent at generating things like hands that don't look strange. Credit: Xeophon Midjourney V7 claims to be much more consistent at generating things like hands that don't look strange. Credit: Xeophon Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreAI image generator Midjourney released its first new model in quite some time today; dubbed V7, it's a ground-up rework that is available in alpha to users now.There are two areas of improvement in V7: the first is better images, and the second is new tools and workflows.Starting with the image improvements, V7 promises much higher coherence and consistency for hands, fingers, body parts, and "objects of all kinds." It also offers much more detailed and realistic textures and materials, like skin wrinkles or the subtleties of a ceramic pot.Those details are often among the most obvious telltale signs that an image has been AI-generated. To be clear, Midjourney isn't claiming to have made advancements that make AI images unrecognizable to a trained eye; it's just saying that some of the messiness we're accustomed to has been cleaned up to a significant degree. V7 can reproduce materials and lighting situations that V6.1 usually couldn't. Credit: Xeophon On the features side, the star of the show is the new "Draft Mode." On its various communication channels with users (a blog, Discord, X, and so on), Midjourney says that "Draft mode is half the cost and renders images at 10 times the speed."However, the images are of lower quality than what you get in the other modes, so this is not intended to be the way you produce final images. Rather, it's meant to be a way to iterate and explore to find the desired result before switching modes to make something ready for public consumption.V7 comes with two modes: turbo and relax. Turbo generates final images quickly but is twice as expensive in terms of credit use, while relax mode takes its time but is half as expensive. There is currently no standard mode for V7, strangely; Midjourney says that's coming later, as it needs some more time to be refined.V7 works with most parameters from previous versions (--ar, --seed, etc), including users' existing --sref codes from 6.1, as well as the recently introduced personalization feature. In fact, V7 is the first Midjourney model that has personalization enabled by default, meaning users will have to train it by picking at least 200 images to build their aesthetic profile.Personalization presents you with a choice between two images hundreds of times so it can learn what you find "beautiful" and tailor its generations to those tastes. You can disable personalization if you want in V7 just like prior models, though.Midjourney was one of the first AI image generation tools that found widespread use. Initially, it was available on Discord and usable via a somewhat arcane syntax, but it has since launched a more modern web interface.A significant portion of the AI art shared on social media has been made with Discord. It's also a key part of the workflow for many AI video creators, who often make the initial image in Midjourney before using the image-to-video feature of applications like Runway.However, as popular as it is, Midjourney has been the subject of multiple lawsuits, and is part of the ongoing debate about whether training AI models on copyrighted works found on the web constitutes fair use. (Anyone who's used Midjourney knows it was trained on copyrighted work; it even sometimes generates watermarks and artist signatures in its outputs.)Recently, the company announced it plans to launch hardware in the future, but it remains unclear what that will look like.Samuel AxonSenior EditorSamuel AxonSenior Editor Samuel Axon is a senior editor at Ars Technica, where he is the editorial director for tech and gaming coverage. He covers AI, software development, gaming, entertainment, and mixed reality. He has been writing about gaming and technology for nearly two decades at Engadget, PC World, Mashable, Vice, Polygon, Wired, and others. He previously ran a marketing and PR agency in the gaming industry, led editorial for the TV network CBS, and worked on social media marketing strategy for Samsung Mobile at the creative agency SPCSHP. He also is an independent software and game developer for iOS, Windows, and other platforms, and heis a graduate of DePaul University, where he studied interactive media and software development. 12 Comments

Time is on The Times' side Judge calls out OpenAIs straw man argument in New York Times copyright suit OpenAI loses bid to dismiss NYT claim that ChatGPT contributes to users infringement. Ashley Belanger Apr 4, 2025 5:19 pm | 15 Credit: gmast3r | iStock / Getty Images Plus Credit: gmast3r | iStock / Getty Images Plus Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreAfter The New York Times sued OpenAI in December 2023alleging that ChatGPT outputs violate copyrights by regurgitating news articlesthe ChatGPT maker tried and failed to argue that the claims were time-barred.According to OpenAI, the NYT should have known that ChatGPT was being trained on its articles and raised its lawsuit in 2020, partly because of the newspaper's own reporting. To support this, OpenAI pointed to a single November 2020 article, where the NYT reported that OpenAI was analyzing a trillion words on the Internet. But on Friday, US district judge Sidney Stein disagreed, denying OpenAI's motion to dismiss the NYT's copyright claims partly based on one NYT journalist's reporting.In his opinion, Stein confirmed that it's OpenAI's burden to prove that the NYT knew that ChatGPT would potentially violate its copyrights two years prior to its release in November 2022. And so far, OpenAI has not met that burden."The fact that one of The Timess reporters discussed OpenAIs" AI training fails to make it clear that the newspaper knew that ChatGPT outputs years later could possibly regurgitate the NYT's reporting, Stein explained.And OpenAI's other argumentthat it was "common knowledge" that ChatGPT was trained on NYT articles in 2020 based on other reportingalso failed for similar reasons."OpenAI fails to explain why the articles, even if their existence had been known to plaintiffs at the time of their publishing, are sufficient to put plaintiffs on notice of the particular infringing conduct by defendants that provides the basis for plaintiffs claims," Stein wrote, which "involve the specific copying of plaintiffs works by OpenAI."Essentially, the judge agreed with the NYT that OpenAI has not yet provided any evidence that the newspaper knew how ChatGPT would perform until the product was out in the wild. Therefore, he denied OpenAI's motion to dismiss those claims as time-barred, while denouncing as a "straw man" an OpenAI argument that the NYT, "as a 'sophisticated publisher,' had a duty 'to take prompt action after being put on notice of what it now claims to be alleged infringement.'"OpenAI may still be able to prove through discovery that the NYT knew that ChatGPT would have infringing outputs in 2020, Stein said. But at this early stage, dismissal is not appropriate, the judge concluded. The same logic follows in a related case from The Daily News, Stein ruled.Davida Brook, co-lead counsel for the NYT, suggested in a statement to Ars that the NYT counts Friday's ruling as a win."We appreciate Judge Stein's careful consideration of these issues," Brook said. "As the opinion indicates, all of our copyright claims will continue against Microsoft and OpenAI for their widespread theft of millions of The Timess works, and we look forward to continuing to pursue them."OpenAI may contribute to ChatGPT users infringementThe New York Times is also arguing that OpenAI contributes to ChatGPT users' infringement of its articles, and OpenAI lost its bid to dismiss that claim, too.The NYT argued that by training AI models on NYT works and training ChatGPT to deliver certain outputs, without the NYT's consent, OpenAI should be liable for users who manipulate ChatGPT to regurgitate content in order to skirt the NYT's paywalls.To win this claim, the NYT claims all that's required is a showing that OpenAI had reason to know that ChatGPT could be used this way, due to media reports and the NYT contacting OpenAI directly. But OpenAI thinks a more heightened standard should apply, and the NYT must prove that OpenAI had "actual knowledge" of or "willful blindness" to the alleged infringement.At this stage, Stein said that the NYT has "plausibly" alleged contributory infringement, showing through more than 100 pages of examples of ChatGPT outputs and media reports showing that ChatGPT could regurgitate portions of paywalled news articles that OpenAI "possessed constructive, if not actual, knowledge of end-user infringement." Perhaps more troubling to OpenAI, the judge noted that "The Times even informed defendants 'that their tools infringed its copyrighted works,' supporting the inference that defendants possessed actual knowledge of infringement by end users.""Taken as true, these facts give rise to a plausible inference that defendants at a minimum had reason to investigate and uncover end-user infringement," Stein wrote.To Stein, the fact that OpenAI maintains an "ongoing relationship" with users by providing outputs that respond to users' prompts also supports contributory infringement claims, despite OpenAI's argument that ChatGPT's "substantial noninfringing uses" are exonerative.OpenAI defeated some claimsFor OpenAI, Stein's ruling likely disappoints, although Stein did drop some of NYT's claims.Likely upsetting to news publishers, that included a "free-riding" claim that ChatGPT unfairly profits off time-sensitive "hot news" items, including the NYT's Wirecutter posts. Stein explained that news publishers failed to plausibly allege non-attribution (which is key to a free-riding claim) because, for example, ChatGPT cites the NYT when sharing information from Wirecutter posts. Those claims are pre-empted by the Copyright Act anyway, Stein wrote, granting OpenAI's motion to dismiss.Stein also dismissed a claim from the NYTregarding alleged removal of copyright management information (CMI), which Stein said cannot be proven simply because ChatGPT reproduces excerpts of NYT articles without CMI.The Digital Millennium Copyright Act (DMCA) requires news publishers to show that ChatGPT's outputs are close to identical to the original work, Stein said, and allowing publishers' claims based on excerpts "would risk boundless DMCA liability"including for any use of block quotes without CMI.Asked for comment on the ruling, an OpenAI spokesperson declined to go into any specifics, instead repeating OpenAI's long-held argument that AI training on copyrighted works is fair use. (Last month, OpenAI warned Donald Trump that the US would lose the AI race to China if courts ruled against that argument.)"ChatGPT helps enhance human creativity, advance scientific discovery and medical research, and enable hundreds of millions of people to improve their daily lives," OpenAI's spokesperson said. "Our models empower innovation, and are trained on publicly available data and grounded in fair use."Ashley BelangerSenior Policy ReporterAshley BelangerSenior Policy Reporter Ashley is a senior policy reporter for Ars Technica, dedicated to tracking social impacts of emerging policies and new technologies. She is a Chicago-based journalist with 20 years of experience. 15 Comments

HIGHLY RESILIENT C2S NSA warns fast flux threatens national security. What is fast flux anyway? Used by nation-states and crime groups, fast flux bypasses many common defenses. Dan Goodin Apr 4, 2025 4:17 pm | 3 Credit: Getty Images Credit: Getty Images Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreA technique that hostile nation-states and financially motivated ransomware groups are using to hide their operations poses a threat to critical infrastructure and national security, the National Security Agency has warned.The technique is known as fast flux. It allows decentralized networks operated by threat actors to hide their infrastructure and survive takedown attempts that would otherwise succeed. Fast flux works by cycling through a range of IP addresses and domain names that these botnets use to connect to the Internet. In some cases, IPs and domain names change every day or two; in other cases, they change almost hourly. The constant flux complicates the task of isolating the true origin of the infrastructure. It also provides redundancy. By the time defenders block one address or domain, new ones have already been assigned.A significant threatThis technique poses a significant threat to national security, enabling malicious cyber actors to consistently evade detection, the NSA, FBI, and their counterparts from Canada, Australia, and New Zealand warned Thursday. Malicious cyber actors, including cybercriminals and nation-state actors, use fast flux to obfuscate the locations of malicious servers by rapidly changing Domain Name System (DNS) records. Additionally, they can create resilient, highly available command and control (C2) infrastructure, concealing their subsequent malicious operations.A key means for achieving this is the use of Wildcard DNS records. These records define zones within the Domain Name System, which map domains to IP addresses. The wildcards cause DNS lookups for subdomains that do not exist, specifically by tying MX (mail exchange) records used to designate mail servers. The result is the assignment of an attacker IP to a subdomain such as malicious.example.com, even though it doesnt exist.Fast flux comes in two variations. Single flux creates DNS A records or AAAA records to map a single domain to many IPv4 or IPv6 addresses, respectively. Heres a diagram illustrating the structure.Double flux provides an additional layer of obfuscation and resiliency by, in addition to changing IP addresses, cycling through the DNS name servers used in domain lookups. Defenders have observed double flux using both Name Server (NS) and Canonical Name (CNAME) DNS records. Heres an illustration of the technique.Both techniques leverage a large number of compromised hosts, usually as a botnet from across the Internet that acts as proxies or relay points, making it difficult for network defenders to identify the malicious traffic and block or perform legal enforcement takedowns of the malicious infrastructure, Thursdays advisory explained. Examples of fast flux use in the wild include:So-called bulletproof hosting serviceswhich offer hardened Internet hosting services to crime-based groupsthat provide fast flux as a means of differentiating themselves from competitorsRansomware attacks from groups such as Hive and NefilimUse of the technique by a Kremlin-backed actor known as GamaredonThe advisory provides several defenses organizations of all sizes should employ to detect and block fast flux networks.Dan GoodinSenior Security EditorDan GoodinSenior Security Editor Dan Goodin is Senior Security Editor at Ars Technica, where he oversees coverage of malware, computer espionage, botnets, hardware hacking, encryption, and passwords. In his spare time, he enjoys gardening, cooking, and following the independent music scene. Dan is based in San Francisco. Follow him at here on Mastodon and here on Bluesky. Contact him on Signal at DanArs.82. 3 Comments

"Victory" Adults told her to move on. Instead, teen won fight to criminalize deepfakes. Here's how one teen plans to fix schools failing kids affected by nudify apps. Ashley Belanger Apr 4, 2025 2:19 pm | 3 Credit: Akiko Aoki | Moment Credit: Akiko Aoki | Moment Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreWhen Francesca Mani was 14 years old, boys at her New Jersey high school used nudify apps to target her and other girls. At the time, adults did not seem to take the harassment seriously, telling her to move on after she demanded more severe consequences than just a single boy's one or two-day suspension.Mani refused to take adults' advice, going over their heads to lawmakers who were more sensitive to her demands. And now, she's won her fight to criminalize deepfakes. On Wednesday, New Jersey Governor Phil Murphy signed a law that he said would help victims "take a stand against deceptive and dangerous deepfakes" by making it a crime to create or share fake AI nudes of minors or non-consenting adultsas well as deepfakes seeking to meddle with elections or damage any individuals' or corporations' reputations.Under the law, victims targeted by nudify apps like Mani can sue bad actors, collecting up to $1,000 per harmful image created either knowingly or recklessly. New Jersey hopes these "more severe consequences" will deter kids and adults from creating harmful images, as well as emphasize to schoolswhose lax response to fake nudes has been heavily criticizedthat AI-generated nude images depicting minors are illegal and must be taken seriously and reported to police. It imposes a maximum fine of $30,000 on anyone creating or sharing deepfakes for malicious purposes, as well as possible punitive damages if a victim can prove that images were created in willful defiance of the law.Ars could not reach Mani for comment, but she celebrated the win in the governor's press release, saying, "This victory belongs to every woman and teenager told nothing could be done, that it was impossible, and to just move on. Its proof that with the right support, we can create change together."On LinkedIn, her mother, Dorota Maniwho has been working with the governor's office on a commission to protect kids from online harmsthanked lawmakers like Murphy and former New Jersey Assemblyman Herb Conaway, who sponsored the law, for "standing with us."When used maliciously, deepfake technology can dismantle lives, distort reality, and exploit the most vulnerable among us," Conaway said. "Im proud to have sponsored this legislation when I was still in the Assembly, as it will help us keep pace with advancing technology. This is about drawing a clear line between innovation and harm. Its time we take a firm stand to protect individuals from digital deception, ensuring that AI serves to empower our communities.Doing nothing is no longer an option for schools, teen saysAround the country, as cases like Mani's continue to pop up, experts expect that shame prevents most victims from coming forward to flag abuses, suspecting that the problem is much more widespread than media reports suggest.Encode Justice has a tracker monitoring reported cases involving minors, including allowing victims to anonymously report harms around the US. But the true extent of the harm currently remains unknown, as cops warn of a flood of AI child sex images obscuring investigations into real-world child abuse.Confronting this shadowy threat to kids everywhere, Mani was named as one of TIME's most influential people in AI last year due to her advocacy fighting deepfakes. She's not only pressured lawmakers to take strong action to protect vulnerable people, but she's also pushed for change at tech companies and in schools nationwide."When that happened to me and my classmates, we had zero protection whatsoever," Mani told TIME, and neither did other girls around the world who had been targeted and reached out to thank her for fighting for them. "There were so many girls from different states, different countries. And we all had three things in common: the lack of AI school policies, the lack of laws, and the disregard of consent."Yiota Souras, chief legal officer at the National Center for Missing and Exploited Children, told CBS News last year that protecting teens started with laws that criminalize sharing fake nudes and provide civil remedies, just as New Jersey's law does. That way, "schools would have protocols," she said, and "investigators and law enforcement would have roadmaps on how to investigate" and "what charges to bring."Clarity is urgently needed in schools, advocates say. At Mani's school, the boys who shared the photos had their names shielded and were pulled out of class individually to be interrogated, but victims like Mani had no privacy whatsoever. Their names were blared over the school's loud system, as boys mocked their tears in the hallway. To this day, it's unclear who exactly shared and possibly still has copies of the images, which experts say could haunt Mani throughout her life. And the school's inadequate response was a major reason why Mani decided to take a stand, seemingly viewing the school as a vehicle furthering her harassment."I realized I should stop crying and be mad, because this is unacceptable," Mani told CBS News.Mani pushed for NJ's new law and claimed the win, but she thinks that change must start at schools, where the harassment starts. In her school district, the "harassment, intimidation and bullying" policy was updated to incorporate AI harms, but she thinks schools should go even further. Working with Encode Justice, she is helping to push a plan to fix schools failing kids targeted by nudify apps."My goal is to protect women and childrenand we first need to start with AI school policies, because this is where most of the targeting is happening," Mani told TIME.Encode Justice did not respond to Ars' request to comment. But their plan noted a common pattern in schools throughout the US. Students learn about nudify apps through ads on social mediasuch as Instagram reportedly driving 90 percent of traffic to one such nudify appwhere they can also usually find innocuous photos of classmates to screenshot. Within seconds, the apps can nudify the screenshotted images, which Mani told CBS News then spread "rapid fire" by text message and DMs, and often shared over school networks.To end the abuse, schools need to be prepared, Encode Justice said, especially since "their initial response can sometimes exacerbate the situation."At Mani's school, for example, leadership was criticized for announcing the victims' names over the loudspeaker, which Encode Justice said never should have happened. Another misstep was at a California middle school, which delayed action for four months until parents went to police, Encode Justice said. In Texas, a school failed to stop images from spreading for eight months while a victim pleaded for help from administrators and police who failed to intervene. The longer the delays, the more victims will likely be targeted. In Pennsylvania, a single ninth grader targeted 46 girls before anyone stepped in.Students deserve better, Mani feels, and Encode Justice's plan recommends that all schools create action plans to stop failing students and respond promptly to stop image sharing.That starts with updating policies to ban deepfake sexual imagery, then clearly communicating to students "the seriousness of the issue and the severity of the consequences." Consequences should include identifying all perpetrators and issuing suspensions or expulsions on top of any legal consequences students face, Encode Justice suggested. They also recommend establishing "written procedures to discreetly inform relevant authorities about incidents and to support victims at the start of an investigation on deepfake sexual abuse." And, critically, all teachers must be trained on these new policies."Doing nothing is no longer an option," Mani said.Ashley BelangerSenior Policy ReporterAshley BelangerSenior Policy Reporter Ashley is a senior policy reporter for Ars Technica, dedicated to tracking social impacts of emerging policies and new technologies. She is a Chicago-based journalist with 20 years of experience. 3 Comments

6 meters long 2025 Chevrolet Silverado EV LT review: This is one long pickup truck Has the moment passed for massive electric trucks with massive range? Jonathan M. Gitlin Apr 4, 2025 11:55 am | 20 Does this shape shout FUTURETRUCK at you? It does for me. Credit: Jonathan Gitlin Does this shape shout FUTURETRUCK at you? It does for me. Credit: Jonathan Gitlin Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreWill this Chevrolet Silverado EV be the biggest electric vehicle we test this year? Almost certainly. Fractionally narrower and less tall than a Hummer EV pickup at more than 18 feet (just under 6 m) long and with a curb weight of 8,532 lbs (3,870 kg), the Silverado EV is what happens when Chevy's electric vehicle engineers get tasked with making their rivals over at Ford feel like they didn't try hard enough with the electric F-150.Now that production has been ongoing for a while, Chevy is filling out the lower trim levels. For commercial customers, there's a Work Truck, but for normies, the entry point is now the LT trim, at a tax credit-friendlyif still wallet-munching$75,195 (for as long as the tax credit still lastsand until the effect of pointless and damaging trade tariffs make themselves known, of course).The 645 hp (481 kW), 756 lb-ft (1,037 Nm) Silverado EV LT comes with the smaller of the two battery packs offered to non-commercial customers. That adjective is doing a lot of work there; a useable 170 kWh is indeed smaller than the 200 kWh you can find in the more expensive RST Max Range, but it's also more than double the capacity of something like a Hyundai Ioniq 5. The range estimate is a commensurate 408 miles (657 km), or "just" 390 miles (628 km) if, as in our test pickup, the premium package has been fitted. At 233 inches (5.9 m), this is a very long vehicle. Credit: Jonathan Gitlin The huge battery is there in large part because the Silverado EV LT is rated at a 12,500-lb (5,700 kg) tow rating. An EV's range at highway speed is partly a function of how much drag it creates, and hooking an irregular-shaped trailer to the back of an already aerodynamically compromised pickup truck shape will provide a vivid demonstration of that fact. Since a trailer's mass also factors into things, the exact range hit will differ with each case (and the weather) in practice.I'll have to leave that to the specialty truck reviewers, though. With no trailer here and nothing needing towing, the hardest workout our tester Silverado EV LT got was a trip to Costco, where the frunk proved a little too small for my needs and the bed too inaccessible with its tonneau cover in place, although that perhaps would have been an opportunity to play with the "multiflex midgate." You may remember this feature from the brand's old Avalanche pickup (or maybe even the Skoda Felicia Fun), which lets you fold down the rear seats to open a passthrough to the truck's bed. Instead, everything went in the back of the very capacious crew cab.Passengers praised the back-seat experience, and it's obvious that the layout was designed around being able to comfortably fit four large humans and a bunch of their stuff. There are useful cubbies all over the place, as well as cupholders and USB-C ports. But as someone on the shorter side of things, I found the right side of the 17.7-inch infotainment system a stretch, and the driver's seat could do with another inch or so of height adjustment to provide a little more of a view over the hood. The Silverado's frunk is easily openable with the remote fob. Jonathan Gitlin The Silverado's frunk is easily openable with the remote fob. Jonathan Gitlin The frunk has a volume of 10.7 cubic feet (303 L). Jonathan Gitlin The frunk has a volume of 10.7 cubic feet (303 L). Jonathan Gitlin With the cover on, there's a cargo volume of 57.7 cubic feet (1,634 L), but small loads will need to be secured if you don't want them tumbling around back there. Jonathan Gitlin With the cover on, there's a cargo volume of 57.7 cubic feet (1,634 L), but small loads will need to be secured if you don't want them tumbling around back there. Jonathan Gitlin The frunk has a volume of 10.7 cubic feet (303 L). Jonathan Gitlin With the cover on, there's a cargo volume of 57.7 cubic feet (1,634 L), but small loads will need to be secured if you don't want them tumbling around back there. Jonathan Gitlin It turns out I didn't take a single good image of the interior, and Chevy doesn't have any either, so this is from the car configurator. Chevrolet 44.3 inches (1,126 mm) of rear leg room is more than ample. Jonathan Gitlin Note the gaze-tracking driver monitoring system on top of the steering column. Jonathan Gitlin No CarPlay or Android Auto here. Jonathan Gitlin Not really optimized for city life. Jonathan Gitlin Big wheels, low-profile tires. The range loss for our test Silverado is mostly down to the set of 22-inch wheels that replace the standard 18-inch items. The wheels are shod with tires that have about the same amount of sidewall as those on the electric Mini crossover that followed it in our test rotation, and the result is a firm ride that was rough over winter potholes and damaged roads. Highway expansion gaps are felt as a bit of a bounce, but otherwise, the ride feels a little floaty when cruising. You'll want the premium pack and bigger wheels if you also want the Super Cruise hands-free assist.At lower speeds, I found the Silverado EV a little more cumbersome. As noted, it's a very long vehicle, and you need the more expensive RST version if you want rear-wheel steering, which turns the opposite direction to the front wheels at low speeds, in effect shrinking the 145.7-inch (3,700 mm) wheelbase. You would be much happier driving one of these straight into a garage rather than backing it into a parking space.Having a garage isn't a must, but in my opinion, being able to charge at home (or reliably at work) still remains a precondition for buying a plug-in vehicle. 120 V (level 1) AC charging might work for routine overnight top-ups if your daily driving is 40 miles or less, but it may take more than a day to completely restore a totally empty pack. Did this truck miss its moment in time? Credit: Jonathan Gitlin Level 2 AC charging should take 810 hours for a full charge (Chevy says 10 miles (16 km) in 10 minutes). Although the powertrain operates at 400 V, the pack can rejigger itself at suitable DC fast chargers to accept an 800 V charge at up to 300 kW. Expect a 1080 percent charge to take around 45 minutes; during my week testing the Silverado EV, I only ran the battery down to around 50 percent, so I wouldn't have seen optimal rates had I plugged it in. With climate change now causing wide temperature swings in early March, I can report that I averaged 1.7 miles/kWh (36.6 kWh/100 km) in cold weather, but once things got mild, that jumped to 2.2 miles/kWh (28.2 kWh/100 km).Was Chevrolet misguided in making the Silverado EV? It certainly made more sense when EV optimism was peaking and the marketing departments in Detroit thought that pickup buyers would be easy conquests for a brave new future powered by electrons. That turned out to be the opposite of true, at least for the time being. But the automaker has a decent selection of EVs in other shapes, sizes, and price points, and an advantage to its common battery platform should be a degree of flexibility in which cars it decides to put them in.Jonathan M. GitlinAutomotive EditorJonathan M. GitlinAutomotive Editor Jonathan is the Automotive Editor at Ars Technica. He has a BSc and PhD in Pharmacology. In 2014 he decided to indulge his lifelong passion for the car by leaving the National Human Genome Research Institute and launching Ars Technica's automotive coverage. He lives in Washington, DC. 20 Comments

Kinetic launch SpinLaunchyes, the centrifuge rocket companyis making a hard pivot to satellites "Launch has generally been more of a cost center than a profit center." Eric Berger Apr 4, 2025 9:59 am | 18 A rendering of a stack of Meridian satellites. Credit: SpinLaunch A rendering of a stack of Meridian satellites. Credit: SpinLaunch Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreOutside of several mentions in the Rocket Report newsletter dating back to 2018, Ars Technica has not devoted too much attention to covering a novel California space company named SpinLaunch.That's because the premise is so outlandish as to almost not feel real. The company aims to build a kinetic launch system that spins a rocket around at speeds up to 4,700 mph (7,500 km/h) before sending it upward toward space. Then, at an altitude of 40 miles (60 km) or so, the rocket would ignite its engines to achieve orbital velocity. Essentially, SpinLaunch wants to yeet things into space.But the company was no joke. After being founded in 2014, it raised more than $150 million over the next decade. It built a prototype accelerator in New Mexico and performed a series of flight tests. The flights reached altitudes of "tens of thousands" of feet, according to the company, and were often accompanied by slickly produced videos.SpinLaunch goes quietFollowing this series of tests, by the end of 2022, the company went mostly quiet. It was not clear whether it ran out of funding, had hit some technical problems in trying to build a larger accelerator, or what. Somewhat ominously, SpinLaunchs founder and chief executive, Jonathan Yaney, was replaced without explanation last May. The new leader would be David Wrenn, then serving as chief operating officer."I am confident in our ability to execute on the companys mission and bring our integrated tech stack of low-cost space solutions to market," Wrenn said at the time. "I look forward to sharing more details about our near- and long-term strategy in the coming months."Words like "tech stack" and "low-cost space solutions" sounded like nebulous corporate speak, and it was not clear what they meant. Nor did Wrenn immediately deliver on that promise, nearly a year ago, to share more details about the company's near- and long-term strategy.Breaking their silenceWrenn and SpinLaunch finally broke their silence this week, and it definitely left some heads spinning. The company made several announcements on Thursday:It plans to build a low-Earth orbit telecommunications constellation, Meridian, with 280 satellitesIt signed a $135 million contract with Europe-based NanoAvionics as a key supplier for this constellationKongsberg Defence & Aerospace will invest $12 million in the Meridian constellationMeridian will be launched on one or two "traditional" rocketsThe company remains committed to kinetic launch, announcing a study of Adak Island, Alaska as a site of a "cutting-edge" launch facilityKinetic launch will, over time, be used to replenish and maintain this constellationIt was a lot to take in at once. But the key point here is that for a decade, the company appeared to be focused on a (potentially) revolutionary method of accelerating rockets into space using a centrifuge but was now pivoting to building a mega-constellation. And initially, at least, this constellation would be launched using plain old chemical rockets.That's a hard pivot. Ars spoke with Wrenn on Thursday afternoon to better understand what is happening here."The launch market is relatively small compared to the economic potential of satellite communication," Wrenn said. "Launch has generally been more of a cost center than a profit center. Satcom will be a much larger piece of the overall industry."He said SpinLaunch began thinking about satellites about five years ago as it developed its launch system, which puts far higher gravitational loads on payloadsliterally thousands of G'sas they spin in the centrifuge."It was quite natural for us," Wrenn said. "We were thinking about how you build satellites for SpinLaunchand how you could deploy and maintain satellites with SpinLaunch. How do you build very high-performing satellites and small form factors?"Can this actually work?The launch market has changed substantially over the last decade. Whereas launch was rare and expensive at the time, thanks largely to the Falcon 9 rocket and SpaceX, it is increasingly ubiquitous. The number of orbital launches, globally, has roughly tripled. A generation of new medium-lift launch vehicles with elements of reuse is coming online. And the small launch business has proven to be economically challenging. It likely would cost SpinLaunch hundreds of millions of dollars, if not more, to develop a full-scale accelerator for orbital launches, and then you need specially built rockets. From the outside, that seems like a difficult business case to close.So the company is diversifying. But can a constellation like Meridian work? Wrenn said the initial plan is to deploy about 250 satellites, each with a mass of 150 pounds (70 kg). The goal is to limit power usage and use modern, miniaturized technology such as "ultra-efficient reflectarray antennas" to pack a lot of capability into a small satellite. More details can be found here.Basically, the company is seeking to launch a constellation with roughly the same data capacity, about 2 "sellable" terabits, as the OneWeb constellation now operated by Eutelsat. But instead of doing this with larger satellites and 20 launches as OneWeb did, Wrenn said the company's constellation of smaller satellites could be lofted by a single large rocket or perhaps two medium-lift rockets.The goal is to fly a demonstration mission in 2026, launch its first commercial satellites no earlier than late 2027, and then begin selling communications services after that, primarily to enterprise (businesses and possibly governments).It seems like a big ask to deliver the same global communications constellation as OneWeb with an order of magnitude fewer launches. But at the same time, the technology in the OneWeb satellites is more than a decade old. There have been a lot of advances made since then.Time will tell whether the new announcement is spin or not.Eric BergerSenior Space EditorEric BergerSenior Space Editor Eric Berger is the senior space editor at Ars Technica, covering everything from astronomy to private space to NASA policy, and author of two books: Liftoff, about the rise of SpaceX; and Reentry, on the development of the Falcon 9 rocket and Dragon. A certified meteorologist, Eric lives in Houston. 18 Comments

Into the flame trench Rocket Report: Next Starship flight to reuse booster; FAA clears New Glenn "The first Super Heavy reuse will be a step towards our goal of zero-touch reflight." Eric Berger Apr 4, 2025 7:00 am | 6 SpaceX tests a Super Heavy booster that previously launched in January. Credit: SpaceX SpaceX tests a Super Heavy booster that previously launched in January. Credit: SpaceX Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreWelcome to Edition 7.38 of the Rocket Report! SpaceX test fired a Super Heavy booster that launched in January on Thursday, in South Texas. This sets up the possibility of a reused Super Heavy rocket launching within the next several weeks, and would be an important step forward in the Starship launch program. It's also a bold step given that there is a lot riding on this Starship launch, given that the last two have failed due to propulsion issues with the rocket's upper stage.As always, we welcome reader submissions, and if you don't want to miss an issue, please subscribe using the box below (the form will not appear on AMP-enabled versions of the site). Each report will include information on small-, medium-, and heavy-lift rockets as well as a quick look ahead at the next three launches on the calendar.European commercial launch industry joins the space race. The first flight of Isar Aerospace's Spectrum rocket didn't last long on Sunday, Ars reports. The booster's nine engines switched off as the rocket cartwheeled upside-down and fell a short distance from its Arctic launch pad in Norway, ending the abbreviated test flight with a spectacular, fiery crash into the sea. However, it marked the beginning of something new in Europe as commercial startups begin launching rockets.Learning to embrace failure ... Isar Aerospace, based in Germany, was the first in a crop of new European rocket companies to attempt an orbital launch. Isar is one of a half-dozen or so European launch startups that could fly their orbital-class rockets in the next couple of years. Of this group, Isar has raised the most money, reporting more than 400 million euros ($430 million) of fundraising, primarily from venture capital sources. We are looking forward to the European launch industry heating up after a long period of development.PLD Space signs launch agreement with D-Orbit. The Spanish launch company, PLD Space, announced an agreement this week with an Italy-based space transportation company, D-Orbit. As part of the agreement, D-Orbit's ION orbital transfer vehicle will launch on PLD Space's forthcoming rocket, the Miura 5. Although the announcement did not specify terms of the agreement, PLD Space said it has now filled "more than 80 percent" of the launch slots on its manifest until 2027.Waiting on the rocket ... The ION vehicle, essentially a dispenser of CubeSats, has previously flown several missions. The real question, therefore, concerns the readiness of the Miura 5 small rocket. PLD Space said it is currently ramping up serial production for the Miura 5 using technology from a prototype rocket, with the aim of starting its test flight campaign by the end of 2025. Commercial flights of Miura 5 could begin in 2026 with the objective of scaling up to 30 launches per year by 2030. We shall see about that. The Ars Technica Rocket Report The easiest way to keep up with Eric Berger's and Stephen Clark's reporting on all things space is to sign up for our newsletter. We'll collect their stories and deliver them straight to your inbox.Sign Me Up!China shooting for record number of launches. Early on Tuesday morning, a Long March 2D rocket lifted off from Jiuquan Satellite Launch Center in the Gobi Desert, Space News reports. The Shanghai Academy of Spaceflight Technology, a state-owned rocket maker, announced the success of the launch, revealing the payload to be a satellite Internet technology test satellite. Tuesdays mission was Chinas 17th orbital launch of 2025, following the launch of the classified TJS-16 satellite into geosynchronous transfer orbit on March 29 via a Long March 7A rocket.Shooting for a century ... This puts the country on pace to launch 68 rockets for the year. This is in line with China's total orbital launches for each of the last three years (64, 67, and 68 launches respectively). However, Chinese space watcher Andrew Jones believes the country may attempt to go as high as 100 launches this year. This would be driven by growing commercial activity, megaconstellation projects, and new launcher development. A number of new, medium-lift and potentially reusable rockets are targeting debut flights this year, he reports.Falcon 9 launches first crewed polar mission. Four adventurers suited up and embarked on a first-of-a-kind trip to space Monday night, becoming the first humans to fly in polar orbit aboard a SpaceX crew capsule chartered by a Chinese-born cryptocurrency billionaire, Ars reports. Chun Wang, born in China and now a citizen of Malta, paid SpaceX an undisclosed sum for the opportunity to fly to space and bring three hand-picked crewmates along with him. He named his mission Fram2 in honor of the Norwegian exploration ship Fram used for polar expeditions at the turn of the 20th century.Rocket follows an unusual trajectory ... The Falcon 9 rocket launched from Kennedy Space Center. However, instead of heading to the northeast in pursuit of the International Space Station, the Falcon 9 and Dragon spacecraft departed Launch Complex 39A and arced to the southeast, then turned south on a flight path hugging Florida's east coast. The unusual trajectory aligned the Falcon 9 with a perfectly polar orbit at an inclination of 90 degrees to the equator, bringing the four-person crew directly over the North or South Pole every 45 minutes. They are the first humans to orbit over the poles.Amazon targets April 9 for first Kuiper launch. As soon as next week, Amazon plans to send 27 of its satellites into low Earth orbit on a United Launch Alliance Atlas 5 rocket, Spaceflight Now reports. Launch is scheduled for Wednesday, April 9, during a three-hour window that opens at noon EDT (16:00 UTC). "We've done extensive testing on the ground to prepare for this first mission, but there are some things you can only learn in flight, and this will be the first time we've flown our final satellite design and the first time we've deployed so many satellites at once," said Rajeev Badyal, vice president of Project Kuiper.Heaviest mission launched by an Atlas ... This will be the first mission by United Launch Alliance of this year, and the company's first in nearly half a year. But officials say that will change soon. In a February interview with Gary Wentz, ULA vice president of Government and Commercial Programs, said that the upcoming launch for Amazon, dubbed Kuiper 1 by ULA and Kuiper Atlas 1 (KA-01) by Amazon, was the first of many planned for the year. "We have quite a few Kuiper Atlases planned this year, as well as Kuiper Vulcans," Wentz said. Atlas can carry 27 Kuiper satellites, and Vulcans can loft 45.SpaceX tests previously flown Super Heavy booster. SpaceX is having trouble with Starship's upper stage after back-to-back failures, but engineers are making remarkable progress with the rocket's enormous booster. The most visible sign of SpaceX making headway with Starship's first stagecalled Super Heavycame at 9:40 am local time (10:40 am EDT; 14:40 UTC) Thursday at the company's Starbase launch site in South Texas. With an unmistakable blast of orange exhaust, SpaceX fired up a Super Heavy booster that has already flown to the edge of space. The burn lasted approximately eight seconds, Ars reports.Rocket will fly on next Starship test ... This was the first time SpaceX has test-fired a "flight-proven" Super Heavy booster, and it paves the way for this particular rocketdesignated Booster 14to fly again soon. A reflight of Booster 14, which previously launched and returned to Earth in January, will happen on the next Starship launch, SpaceX confirmed Thursday. "This booster previously launched and returned on Flight 7 and 29 of its 33 Raptor engines are flight proven," the company said. "The first Super Heavy reuse will be a step towards our goal of zero-touch reflight." It is a legitimately and characteristically bold decision to refly a Starship booster on a test flight that SpaceX really needs to succeed. The next test may come late this month or more likely in May.FAA closes big rocket mishap investigations. The Federal Aviation Administration (FAA) closed mishap investigations into both the SpaceX Starship flight and Blue Origin New Glenn debut that both took place on Jan. 16, Via Satellite reports. Although the FAA closed the mishap investigation regarding the January 16 Starship flight, the rocket is still grounded because there is still an open mishap investigation into the following March 7 flight. "There were no public injuries and one confirmed report of minor vehicle damage in the Turks and Caicos Islands," the FAA said in a statement on the January 16 flight.New Glenn closed out as well ... The FAA also completed its mishap investigation of Blue Origins first New Glenn flight, which successfully deployed Blue Origins own space logistics vehicle Blue Ring. Blue Origin failed to recover the first stage booster, which triggered the mishap investigation. The first stage was not able to restart its engines, which prevented the reentry burn from occurring and caused the loss of the stage. Blue Origin has identified seven corrective actions, and the FAA will verify those have been implemented before the second mission. Blue Origin is targeting a return to flight in late spring and will attempt to land the booster again.Artemis II one step closer to launch. The four astronauts who will fly on board NASA's Artemis II mission unveiled the patch for their historic flight on Thursday. The four astronauts who will be the first to fly to the Moon under NASAs Artemis campaignCommander Reid Wiseman, pilot Victor Glover, and mission specialist Christina Koch from NASA, and mission specialist Jeremy Hansen from Canadahave designed an emblem to represent their mission that references both their distant destination and the home they will return to, the space agency said. It looks great!Here's what it means ... "This patch designates the mission as AII, signifying not only the second major flight of the Artemis campaign, but also an endeavor of discovery that seeks to explore for all and by all. Framed in Apollo 8s famous Earthrise photo, the scene of the Earth and the Moon represents the dual nature of human spaceflight, both equally compelling: The Moon represents our exploration destination, focused on discovery of the unknown. The Earth represents home, focused on the perspective we gain when we look back at our shared planet and learn what it is to be uniquely human. The orbit around Earth highlights the ongoing exploration missions that have enabled Artemis to set sights on a long-term presence on the Moon and soon, Mars."Next three launchesApril 4: Falcon 9 | Starlink 11-13 | Vandenberg Space Force Base, Calif. | 01:02 UTCApril 6: Falcon 9 | Starlink 6-72 | Cape Canaveral Space Force Station, Fla. | 02:40 UTCApril 7: Falcon 9 | Starlink 11-11 | Vandenberg Space Force Base, California | 21:35 UTCEric BergerSenior Space EditorEric BergerSenior Space Editor Eric Berger is the senior space editor at Ars Technica, covering everything from astronomy to private space to NASA policy, and author of two books: Liftoff, about the rise of SpaceX; and Reentry, on the development of the Falcon 9 rocket and Dragon. A certified meteorologist, Eric lives in Houston. 6 Comments

Andrew Cunningham and Lee Hutchinson have spent decades of their lives with Robert Jordan and Brandon Sanderson's Wheel of Time books, and they previously brought that knowledge to bear as they recapped each first season episode and second season episode of Amazon's WoT TV series. Now we're back in the saddle for season 3along with insights, jokes, and the occasional wild theory.These recaps won't cover every element of every episode, but they will contain major spoilers for the show and the book series. We'll do our best to not spoil major future events from the books, but there's always the danger that something might slip out. If you want to stay completely unspoiled and haven't read the books, these recaps aren't for you.New episodes of The Wheel of Time season 3 will be posted for Amazon Prime subscribers every Thursday. This write-up covers episode six, "The Shadow in the Night," which was released on April 3.Lee: Welcome to Tanchico! In Tanchico, everyone wears veils almost all of the time, except when theyre flirting in bars. Mat gets the most fabulous veil of all because hes Mat and he deserves it. Even Nynaeve has a good time! And I guess now we know all about the hills of Tanchico. Like alllllllllllllllllll about them.Andrew: Credit to Robert Jordan for mostly resisting one of the bizarre tics of post-Tolkien fantasy fiction: I'm not going to say the books never take a break to give us the full text of an in-universe song. But it does so pretty sparingly, if memory serves. But there are plenty of songs referenced, often with a strong implication that they are too lewd or horny to reprint in full.Not so in the show! Where Elayne sings a song about "The Hills of Tanchico," bringing the house down for what appears to be... several hours (they're breasts, the hills are breasts). I don't mind this scene, actually, but it does go on.But more important than the song is who is accompanying Elayne, a book character who has been gone so long that we weren't actually sure he was coming back. Who makes their long-awaited return in Tanchico, Lee? Thom Merrilin finally shows back up. Nice hat. Wonder who else might end up wearing it. Credit: Prime/Amazon MGM Studios Thom Merrilin finally shows back up. Nice hat. Wonder who else might end up wearing it. Credit: Prime/Amazon MGM Studios Lee: Thats right, ladies and gentlemen, boys and girls, children of all ages, stomp your feet and bring your hands together for everybodys favorite gleeman, seemingly back from the dead and rocking a strangely familiar hat: Its Thom Merrilin! (Applause roars.)Viewers who havent read the books can be forgiven for not immediately falling out of their chairs when Thom shows back up, but to book readers, his absence has been keenly felt. Believe it or not, Merrilin is an A-string player in the books, spending a tremendous amount of time front and center interacting with the main POV characters. He vanishes for a bit just as he does in the show, but he doesnt stay gone nearly as long as hes been gone here.Im glad hes back, and it bodes well for our Tanchico crewunlike them, Thom is an actual-for-real adult, whos been places and knows things. He also provides fantastic accompaniment to Elaynes karaoke adventure. Elayne wins the crowd by singing about tittays. Thom accompanies because it's a subject in which he is apparently well-versed. Credit: Prime/Amazon MGM Studios Elayne wins the crowd by singing about tittays. Thom accompanies because it's a subject in which he is apparently well-versed. Credit: Prime/Amazon MGM Studios Andrew: The entire Tanchico crew is pretty strong right nowMat and Min are pals again, show-Nynaeve is a version of the character who other characters in the story are allowed to like, and now Thom is back! It'd be a rollicking good time, if it weren't for these sadistic Black Ajah Aes Sedai and the Forsaken psychopath Moghedien stalking around, mind-controlling people, leaving holes in heads, and trying to find a Seanchan-esque collar that can subdue and control Rand.We're entering a stretch of the story where the Forsaken spend as much time fighting with each other as they do with Rand and our heroes, which explains why the powerful villains don't simply kill our heroes the minute they find each other. Moghedien is in full creep mode through this whole episode, and I gotta say, she is unsettling. Moghedien, doing her thing. Credit: Prime/Amazon MGM Studios Moghedien, doing her thing. Credit: Prime/Amazon MGM Studios Lee: Yeah, watching Moghedien screw with the Black sisters food and stuff was particularly disturbing. The lady has no filterand fantastic powers of persuasion. We get another clear look at just how ludicrously overpowered the Forsaken are compared to our present-day channelers when Moggy straight-up runs "sudo give me the bracelet" on Nynaeves and Elaynes brainsmuch like Rhavins Im-your-favorite-uncle routine, her Power-backed trickery is devastating and completely inescapable (though Nynaeve apparently does resist just a teeny tiny bit.)And although there are still more doings to discuss in Tanchicothe quest to discover the bracelets-n-collars is heating up!the fact that all of these episodes are an hour long means there are so many other things to discuss. Like, for example, the return of another familiar face, in the form of our long-absent whistling super-darkfriend Padan Fain. Dark doings are afoot in the Two Rivers!Andrew: Fain in the books never quite rises to the level of Big Bad so much as he lurks around the periphery of the story practically the whole entire time, popping up to cause trouble whenever it's the least convenient for our heroes. The show does a good job of visually representing how he's begun to corrupt the regiment of Whitecloaks he has embedded himself in, without ever actually mentioning it or drawing much attention to it. You know you're a bad guy when even Eamon Valda is like "uh is this guy ok?" (As in the books, the show distinguishes between Whitecloaks who are antagonists because they genuinely believe what they say they believe about Aes Sedai "witches," and ones who are simply straight-up Darkfriends. Funny how often they end up working toward the same goals, though.)Meanwhile, Perrin, Alanna, and friends recover from last week's raid of the Whitecloak camp. I keep needing to recalibrate my expectations for what Plot Armor looks like on this show, because our main characters get grievously wounded pretty regularly, but the standards are different on a show where everyone can apparently cast Cure Wounds as a cantrip. Alanna walks the Cauthon sisters through some rudimentary Healing, and Alanna (with barely disguised glee and/or interest) accidentally interrupts an escalation in Perrin and Falie's relationship when she goes to Heal him later.Are we still finding show-Faile charming? I did think it was funny when that goofy county-fair caricature of Mat holding the Horn of Valere made another appearance. Still not hating Faile, which feels surprising. Credit: Prime/Amazon MGM Studios Still not hating Faile, which feels surprising. Credit: Prime/Amazon MGM Studios Lee: I am definitely still finding show-Faile charming, which continually surprises me because shes possibly the worst character in the entire series. In the books, Jordan writes Faile as an emotionally abused emotional abuser who doesnt believe Perrin loves her if hes not screaming at her and/or hitting her; in the show, shes a much more whole individual with much more grown-up and sane ideas about how relationships work. Perrin and Faile have something going on that is, dare I say it, actually sweet and romantic!I never thought Id be on any team other than Team Throw-Faile-Down-The-Well, but here we are. Im rooting for her and Perrin.When it comes to Alannas healing at the hands of the Cauthon sisters, I had to sit with that one for a moment and make a conscious decision. The books make it clear that Healingeven the abbreviated first-aid version the current-day Aes Sedai practice, to say nothing of the much fancier version from the Age of Legendsis complicated. Doing it wrong can have horrific consequences (in fact, doing healing wrong on purpose is the basis for many of the Yellow-turned-Black sisters attacks with the One Power). And these wildlings (to borrow a book term) are able to just intuit their way into making it happen?We know that new channelers frequently have uncontrolled bouts of blasting out the One Power in response to moments of stress or great needin fact, weve seen that happen many times in the show, including at the beginning of this episode when Lil Liandrin force-blasts her rapist-husband into the wall. So the groundwork is there for the Cauthon girls to do what theyre doing. Its just a question of how much one is willing to let the show get away with.I decided Im good with itits the necessary thing to move the story forward, and so Im not gonna complain about it. Where did you land? Fain returns, bringing with him the expected pile of Trollocs. Credit: Prime/Amazon MGM Studios Fain returns, bringing with him the expected pile of Trollocs. Credit: Prime/Amazon MGM Studios Andrew: Yeah, I made essentially the same decision. Conscious use of the One Power at all, even the ability to access it consistently, is something that requires patience and training, and normally you couldn't talk a 12-year-old through Healing as Alanna does here any more than you could talk a 12-year-old through performing successful field surgery. But training takes time, and showing it takes time, and time is one thing the show never has much of. The show also really likes to dramatically injure characters without killing them! So here we are, speed-running some things.This leaves us with two big threads left to address: Rand's and Egwene's. Egwene is still trying to learn about the World of Dreams from the Aiel Wise Ones (I was wrong, by the wayshe admits to lying about being Aes Sedai here and it passes almost without comment), and is still reeling from realizing that Rand and Lanfear are Involved. And Rand, well. He's not going mad, yet, probably, but he spends most of the episode less-than-fully-in-control of his powers and his actions.Lee: It comes to a head when Rand and Egwene have long, difficult conversation over exactly whos been sleeping with whom, and whyand then that conversation is interrupted when Sammael kicks the door down and starts swinging his big fancy One Power Hammer.Theres a bit of channeling by Aviendha and Egwene, but then Rand grasps the Source and Sammael just kind of stops being a factor. Entranced by the Powerand by the black corruption pulsing through itRand straight-up destroys Sammael without apparent thought or effort, borrowing a bit of the method from the way Rand pulls off a similar feat in book 3, with a ludicrous amount of lightning and ceiling-collapsing.Its one of the few times so far that Rand has actually cut loose with the One Power, and I like it when we get to actually see (rather than just hear about) the enormity of Rands strength as a channeler. But this casual exercise of extreme power is not without a cost. Rand does a 360 no-scope lightning hit. Credit: Prime/Amazon MGM Studios Rand does a 360 no-scope lightning hit. Credit: Prime/Amazon MGM Studios Andrew: We've observed a couple of times that Rand and Egwene in the books had long since given up on romantic involvement by this point in the story, and here we see why the show held back on thatthis confrontation is more exciting than a quiet drift, and it puts a cap on several "Rand is not the simple lad you once knew" moments sprinkled throughout the episode.And, yes, one of them is Rand's inadvertent (if sadly predictable) killing of an Aiel girl he had forged a bond with, and his desperate, fruitless, unsavory attempt to force her back to life. Rand is simultaneously coming to grips with his destiny and with the extent to which he has no idea what he is doing, and both things are already causing pain to the people around him. And as you and I both know, book-Rand has counterproductive and harmful reactions to hurting people he cares about.The attack here is partly an invention of the show and partly a synthesis of a few different book events, but Forsaken coming at Rand directly like this is generally not a thing that happens much. They usually prefer to take up positions of power in the world's various kingdoms and only fight when cornered. All of this is to say, I doubt this is the last we see of Sammael or his Thor-looking One Power hammer, but the show is more than willing to go its own way when it wants to.Lee: Yeah, Rand doing saidin-CPR on Rhuarcs poor little too-cute-not-to-be-inevitably-killed granddaughter is disturbing as helland as you say, its terrifying not just because Rand is forcing a corpse to breathe with dark magic, but also because of the place Rand seems to go in his head when hes doing it. Its been an oft-repeated axiom that male channelers inevitably go madis this it? (Fortunately, nonot yet, at least. Or is it? No! Maybe.)We close the episode out on the place where I think were going to probably be spending a lot of time very soon (especially based on the title of next weeks episode, which I wont spoil but which anyone can look up if they wish): back at the Two Rivers, with the power-trio of Bain and Chiad and Faile scouting out the old Waygate just outside of town, and watching Trollocs swarm out of it. This is not a great sign for Perrin and friends.So weve got two episodes left, all of our chess pieces seem to have been set down more or less into the right places for a couple of major climactic events. I think were going out with a bangor with a couple of them. What are you thinking as we jump into the final couple of episodes? Alsera fell victim to one of the classic child character blunders: being too precociously adorable to live. Credit: Prime/Amazon MGM Studios Alsera fell victim to one of the classic child character blunders: being too precociously adorable to live. Credit: Prime/Amazon MGM Studios Andrew: I am going to reiterate our annual complaint that 10-episode seasons would be better for this show's storytelling than the 8-episode seasons we're getting, but because the show's pace is always so breathless and leaves room for just a few weird character-illuminating diversions like "The Hills of Tanchico," or quiet heart-to-hearts like we get between Rand and Moiraine, or between Perrin and Faile. The show's good enough at these that I wish we had time to pump the brakes more often.But I will say, if we end up roughly where book 4 does, the show doesn't feel as rushed as the first two seasons did. Not that its pacing has settled down at allyou and I benefit immensely from being book readers, and always being rooted in some sense of what is happening and who the characters are that the show can't always convey with perfect clarity. But I am thinking about what still needs to happen, and how much time there is left, and thinking "yeah, they're going to be able to get there" instead of "how the hell are they going to get there??"How are you feeling? Is season 3 hitting for you like it is for me? I know I'm searching around every week to see if there's been a renewal announcement for season 4 (not yet).Lee: I think its the best season so far, and any doubts I had during seasons one and two are at this point long gone. Im all in on this particular turning of the Wheel, and the show finally feels like it's found itself. To not renew it at this point would be criminal. You listening, Bezos? May the Shadow take you if you yank the rug out from under us now!Andrew: Yeah, Jeffrey. I know for a fact you've spent money on worse things than this. Credit: WoT Wiki

Yodel-ay-hee-hoo! Monkeys are better yodelers than humans, study finds Voice breaks in Latin American monkey calls resemble human yodeling, but over a much wider frequency range. Jennifer Ouellette Apr 3, 2025 3:10 pm | 9 Black and gold howler monkeys (Alouatta caraya) Credit: Jacob Dunn, Anglia Ruskin University Black and gold howler monkeys (Alouatta caraya) Credit: Jacob Dunn, Anglia Ruskin University Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreHumans have practiced some form of yodeling since at least the 13th century, when Marco Polo encountered Tibetan monks on his travels who used the vocal technique for long-distance communication. It's since morphed into a distinctive singing style. But can animals also yodel? According to a new paper published in the Philosophical Transactions of the Royal Society B, Biological Sciences, several species of monkey dwelling in the rainforests of Latin America employ "voice breaks" in their calls that acoustically resemble human yodelingi.e., "ultra-yodels" that boast a much wider frequency range.Many years ago, I wrote about the bioacoustics of human yodeling for New Scientist. In many respects, yodeling is quite simple. It merely involves singing a long note subjected to repeated rapid sharp shifts in pitch. It's the unique anatomy of the human vocal tract that makes it possible, notably the larynx (voice box) located just behind the Adam's apple. The larynx is comprised of cartilage and the hyoid bone that together support the vocal cords, which are attached to muscles on either side of the larynx.When air flows through the trachea, the vocal cords vibrate at frequencies ranging from 110 to 200 Hz. We have the capability of contracting the muscles to change the shape, position, and tension of our vocal cords, thereby altering the pitch of the sound produced. Stiffer vocal cords result in faster vibrations, which produce higher pitches.Prior research has shown that human vocal cords share the same two principal vibration modes as a guitar, and human singers rely on both when performing. It's well-known that we have two distinct ranges: a "head voice" and a "chest voice." There is a gap when shifting between them, which trained singers learn how to smooth over. Yodelers accentuate that gap instead.Monkey see, monkey yodel?That's how it works for humans, but when it comes to the question of yodeling animals, it depends on how you define yodeling, according to bioacoustician Tecumseh Fitch of the University of Vienna in Austria, who co-authored this latest paper. Plenty of animal vocalizations use repeated sudden changes in pitch (including birds), and a 2023 study found that toothed whales can produce vocal registers through their noses for echolocation and communication.There haven't been as many studies of vocal registers in non-human primates, but researchers have found, for example, that the "coo" call of the Japanese macaque is similar to a human falsetto; the squeal of a Syke monkey is similar to the human "modal" register; and the Diana monkey produces alarm calls that are similar to "vocal fry" in humans. Tufted capuchins (Sapajus apella) Jacob Dunn, Anglia Ruskin University Tufted capuchins (Sapajus apella) Jacob Dunn, Anglia Ruskin University Peruvian spider monkey (Ateles chamek) Jacob Dunn, Anglia Ruskin University Peruvian spider monkey (Ateles chamek) Jacob Dunn, Anglia Ruskin University Tufted capuchins (Sapajus apella) Jacob Dunn, Anglia Ruskin University Peruvian spider monkey (Ateles chamek) Jacob Dunn, Anglia Ruskin University It's known that non-human primates have something humans have lost over the course of evolution: very thin, light vocal membranes just above the vocal folds. Scientists have pondered the purpose of those membranes, and a 2022 study concluded that this membrane was crucial for producing sounds. The co-authors of this latest paper wanted to test their hypothesis that the membranes serve as an additional oscillator to enable such non-human primates to achieve the equivalent of human voice registers. That, in turn, would render them capable in principle of producing a wider range of callsperhaps even a yodel.The team studied many species, including black and gold howler monkeys, tufted capuchins, black-capped squirrel monkeys, and Peruvian spider monkeys. They took CT scans of excised monkey larynxes housed at the Japan Monkey Center, as well as two excised larynxes from tufted capuchin monkeys at Kyoto University. They also made live recordings of monkey calls at the La Senda Verde animal refuge in the Bolivian Andes, using non-invasive EGG to monitor vocal fold vibrations.The results showed that New World monkeys had evolved the largest primate vocal membranes, supporting the hypothesis that they play a critical role in vocal production, specifically voice breaks that transitioned rapidly in frequency, which they achieved by switching sound production from their vocal folds to their vocal membranes. It's technically a yodel, even though it might not sound much like the human version (cf. the recording of a tufted capuchin above).The monkeys' "ultra-yodels" had frequency leaps that were five times larger than human frequency changes, often exceeding three octaves. (Human yodelers are limited to one octave.) While the monkeys' vocal membranes serve to extend their pitch range, the tradeoff is a lack of pitch stability.These results show how monkeys take advantage of an evolved feature in their larynxthe vocal membranewhich allows for a wider range of calls to be produced, including these ultra-yodels," said co-author Jacob Dunn, an evolutionary biologist at Anglia Ruskin University in Cambridge, England.This might be particularly important in primates, which have complex social lives and need to communicate in a variety of different ways. Its highly likely this has evolved to enrich the animals call repertoire, and is potentially used for attention-grabbing changes, call diversification, or identifying themselves.Philosophical Transactions of the Royal Society B, Biological Sciences, 2025. DOI: 10.1098/rstb.2024.0005 (About DOIs).Jennifer OuelletteSenior WriterJennifer OuelletteSenior Writer Jennifer is a senior writer at Ars Technica with a particular focus on where science meets culture, covering everything from physics and related interdisciplinary topics to her favorite films and TV series. Jennifer lives in Baltimore with her spouse, physicist Sean M. Carroll, and their two cats, Ariel and Caliban. 9 Comments

CLIENT-SIDE ENCRYPTION Google unveils end-to-end messages for Gmail. Only thing is: Its not true E2EE. Yes, encryption/decryption occurs on end-user devices, but there's a catch. Dan Goodin Apr 3, 2025 5:16 pm | 10 Credit: Getty Images Credit: Getty Images Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreWhen Google announced Tuesday that end-to-end encrypted messages were coming to Gmail for business users, some people balked, noting it wasnt true E2EE as the term is known in privacy and security circles. Others wondered precisely how it works under the hood. Heres a description of what the new service does and doesnt do, as well as some of the basic security that underpins it.When Google uses the term E2EE in this context, it means that an email is encrypted inside Chrome, Firefox, or just about any other browser the sender chooses. As the message makes its way to its destination, it remains encrypted and cant be decrypted until it arrives at its final destination, when its decrypted in the recipient's browser.Giving S/MIME the heave-hoThe chief selling point of this new service is that it allows government agencies and the businesses that work with them to comply with a raft of security and privacy regulations and at the same time eliminates the massive headaches that have traditionally plagued anyone deploying such regulation-compliant email systems. Up to now, the most common means has been S/MIME, a standard so complex and painful that only the bravest and most well-resourced organizations tend to implement it.S/MIME requires each sender and receiver to have an X.509 certificate thats been issued by a certificate authority. Obtaining, distributing, and managing these certificates in a secure manner takes time, money, and coordination. That means that if Bob and Alice have never worked together before and an urgent or unexpected need arises for him to send Alice an encrypted message promptly, theyre out of luck until an admin applies for a certificate and sees that its installed on Alices machineso much for flexibility and agility.Google says that E2EE Gmail abstracts away this complexity. Instead, Bob drafts an email to Alice, clicks a button that turns on the feature, and hits send. Bobs browser encrypts the message, and sends it to Alice. The message decrypts only after it arrives in Alices browser and she authenticates herself.To make this happen, Bobs organization deploys what Google says is a lightweight key server, known as a KACL, short for a key access control list. This server, which can be hosted on premises or most cloud services, is where keys are generated and stored. When Bob sends an encrypted message, his browser connects to the key server and obtains an ephemeral symmetric encryption key. Bobs browser encrypts the message and sends it to Alice, along with a reference key. Alices browser uses the reference key to download the symmetric key from the KACL and decrypts the message. The key is then deleted.To prevent Mallory or another adversary-in-the-middle from obtaining the key, Alice must first authenticate herself through Okta, Ping, or whatever other identity provider, or IDP, Bobs organization uses. If this is the first time Alice has received a message from Bobs organization, she will first have to prove to the IDP that she has control of her email address. If Alice plans to receive encrypted emails from Bobs organization in the future, Alice sets up an account that can be used going forward.Bobs organization can add an additional layer of protection by requiring Alice to already have an account on the IDP and authenticate herself through it.The idea is that no matter what, at no time and in no way does Gmail ever have the real key. Never, Julien Duplant, a Google Workspace product manager, told Ars. And we never have the decrypted content. Its only happening on that users device.Now, as to whether this constitutes true E2EE, it likely doesnt, at least under stricter definitions that are commonly used. To purists, E2EE means that only the sender and the recipient have the means necessary to encrypt and decrypt the message. Thats not the case here, since the people inside Bobs organization who deployed and manage the KACL have true custody of the key.In other words, the actual encryption and decryption process occurs on the end-user devices, not on the organizations server or anywhere else in between. Thats the part that Google says is E2EE. The keys, however, are managed by Bobs organization. Admins with full access can snoop on the communications at any time.The mechanism making all of this possible is what Google calls CSE, short for client-side encryption. It provides a simple programming interface that streamlines the process. Until now, CSE worked only with S/MIME. Whats new here is a mechanism for securely sharing a symmetric key between Bobs organization and Alice or anyone else Bob wants to email.The new feature is of potential value to organizations that must comply with onerous regulations mandating end-to-end encryption. It most definitely isnt suitable for consumers or anyone who wants sole control over the messages they send. Privacy advocates, take note.Dan GoodinSenior Security EditorDan GoodinSenior Security Editor Dan Goodin is Senior Security Editor at Ars Technica, where he oversees coverage of malware, computer espionage, botnets, hardware hacking, encryption, and passwords. In his spare time, he enjoys gardening, cooking, and following the independent music scene. Dan is based in San Francisco. Follow him at here on Mastodon and here on Bluesky. Contact him on Signal at DanArs.82. 10 Comments

How to avoid Terminators Google DeepMind releases its plan to keep AGI from running wild DeepMind says AGI could arrive in 2030, and it has some ideas to keep us safe. Ryan Whitwam Apr 3, 2025 5:43 pm | 0 Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreAs AI hype permeates the Internet, tech and business leaders are already looking toward the next step. AGI, or artificial general intelligence, refers to a machine with human-like intelligence and capabilities. If today's AI systems are on a path to AGI, we will need new approaches to ensure such a machine doesn't work against human interests.Unfortunately, we don't have anything as elegant as Isaac Asimov's Three Laws of Robotics. Researchers at DeepMind have been working on this problem and have released a new technical paper (PDF), which you can download at your convenience.It contains a huge amount of detail, clocking in at 108 pages before references. While some in the AI field believe AGI is a pipe dream, the authors of the DeepMind paper project that it could happen by 2030. With that in mind, they aimed to understand the risks of a human-like synthetic intelligence, which they acknowledge could lead to "severe harm."All the ways AGI could suck for humanityThis work has identified four possible types of AGI risk, along with suggestions on how we might ameliorate said risks. The DeepMind team, led by company co-founder Shane Legg, categorized the negative AGI outcomes as misuse, misalignment, mistakes, and structural risks. The four categories of AGI risk, as determined by DeepMind. Credit: Google DeepMind The four categories of AGI risk, as determined by DeepMind. Credit: Google DeepMind The first possible issue, misuse, is fundamentally similar to current AI risks. However, because AGI will be more powerful by definition, the damage it could do is much greater. A ne'er-do-well with access to AGI could misuse the system to do harm, for example, by asking the system to identify and exploit zero-day vulnerabilities or create a designer virus that could be used as a bioweapon.DeepMind says companies developing AGI will have to conduct extensive testing and create robust post-training safety protocols. Essentially, our current AI guardrails on steroids. They also suggest devising a method to suppress dangerous capabilities entirely, sometimes called "unlearning," but it's unclear if this is possible without substantially limiting models.Misalignment is largely not something we have to worry about with generative AI as it currently exists. This type of AGI harm is envisioned as a rogue machine that has shaken off the limits imposed by its designers. Terminators, anyone?To avoid that, DeepMind suggests developers use techniques like amplified oversight, in which two copies of an AI check each other's output, to create robust systems that aren't likely to go rogue. If that fails, DeepMind suggests intensive stress testing and monitoring to watch for any hint that an AI might be turning against us. Keeping AGIs in virtual sandboxes with strict security and direct human oversight could help mitigate issues arising from misalignment. Basically, make sure there's an "off" switch.If, on the other hand, an AI didn't know that its output would be harmful and the human operator didn't intend for it to be, that's a mistake. We get plenty of those with current AI systemsremember when Google said to put glue on pizza? The "glue" for AGI could be much stickier, though. DeepMind notes that militaries may deploy AGI due to "competitive pressure," but such systems could make serious mistakes as they will be tasked with much more elaborate functions than today's AI.The paper doesn't have a great solution for mitigating mistakes. It boils down to not letting AGI get too powerful in the first place. DeepMind calls for deploying slowly and limiting AGI authority. The study also suggests passing AGI commands through a "shield" system that ensures they are safe before implementation.Lastly, there are structural risks, which DeepMind defines as the unintended but real consequences of multi-agent systems contributing to our already complex human existence. For example, AGI could create false information that is so believable that we no longer know who or what to trust. The paper also raises the possibility that AGI could accumulate more and more control over economic and political systems, perhaps by devising heavy-handed tariff schemes. Then one day, we look up and realize the machines are in charge instead of us. This category of risk is also the hardest to guard against because it would depend on how people, infrastructure, and institutions operate in the future.AGI in five years?No one knows if the thinking machines are really just a few years away, but there are plenty of tech leaders who are confident enough to say so. Part of the problem in predicting the emergence of AGI is that we're still just speculating about how human-like intelligence would manifest itself in a machine. Anyone who has used generative AI systems over the past years has seen real, tangible improvements, but does that trajectory lead to true human-like capabilities?We recently talked about a range of AI topics, including AGI, with Google's Tulsee Doshi, director of product management for Gemini. "Different people have different definitions of AGI, and so depending on who you talk to, how close or far we are from AGI is a different conversation," said Doshi. "What I would say is LLMs, Gemini, and the training of smarter and smarter models is on the path to models that are going to be at extremely high intelligence. And that has a ton of value in and of itself."This paper is not the final word on AGI safetyDeepMind notes this is just a "starting point for vital conversations." If the team is right, and AGI will transform the world in five short years, those conversations need to happen soon. If not, well, a lot of people are going to look kind of silly.Ryan WhitwamSenior Technology ReporterRyan WhitwamSenior Technology Reporter Ryan Whitwam is a senior technology reporter at Ars Technica, covering the ways Google, AI, and mobile technology continue to change the world. Over his 20-year career, he's written for Android Police, ExtremeTech, Wirecutter, NY Times, and more. He has reviewed more phones than most people will ever own. You can follow him on Bluesky, where you will see photos of his dozens of mechanical keyboards. 0 Comments

Dying young Wealthy Americans have death rates on par with poor Europeans Some wealthy Europeans have death rates 35 percent lower than the richest Americans. Beth Mole Apr 3, 2025 6:18 pm | 18 Pensioners look out across Lake Zurich in central Zurich, Switzerland, on Wednesday, Aug. 21, 2024. Credit: Getty | Pascal Mora Pensioners look out across Lake Zurich in central Zurich, Switzerland, on Wednesday, Aug. 21, 2024. Credit: Getty | Pascal Mora Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreIt's well-established that, on the whole, Americans die younger than people in most other high-income countries. For instance, an analysis from 2022 found that the average life expectancy of someone born in Switzerland or Spain in 2019 was 84 years. Meanwhile, the average US life expectancy was 78.8, lower than nearly all other high-income countries, including Canada's, which was 82.3 years. And this was before the pandemic, which only made things worse for the US.Perhaps some Americans may think that this lower overall life-expectancy doesn't really apply to them if they're middle- or upper-class. After all, wealth inequality and health disparities are huge problems in the US. Those with more money simply have better access to health care and better health outcomes. Well-off Americans live longer, with lifespans on par with their peers in high-income countries, some may think.It is true that money buys you a longer life in the US. In fact, the link between wealth and mortality may be stronger in the US than in any other high-income country. But, if you think American wealth will put life expectancy in league with Switzerland, you're dead wrong, according to a study in the latest issue of the New England Journal of Medicine.A stark findingThe study, led by researchers at Brown University, found that the wealthiest Americans lived shorter lives than the wealthiest Europeans. In fact, wealthy Northern and Western Europeans had death rates 35 percent lower than the wealthiest Americans, whose lifespans were more like the poorest in Northern and Western Europewhich includes countries such as France, the Netherlands, and Switzerland."The findings are a stark reminder that even the wealthiest Americans are not shielded from the systemic issues in the US contributing to lower life expectancy, such as economic inequality or risk factors like stress, diet or environmental hazards," lead study author Irene Papanicolas, a professor of health services, policy and practice at Brown, said in a news release.The study looked at health and wealth data of more than 73,000 adults across the US and Europe who were 50 to 85 years old in 2010. There were more than 19,000 from the US, nearly 27,000 from Northern and Western Europe, nearly 19,000 from Eastern Europe, and nearly 9,000 from Southern Europe. For each region, participants were divided into wealth quartiles, with the first being the poorest and the fourth being the richest. The researchers then followed participants until 2022, tracking deaths.The US had the largest gap in survival between the poorest and wealthiest quartiles compared to European countries. America's poorest quartile also had the lowest survival rate of all groups, including the poorest quartiles in all three European regions.While less access to health care and weaker social structures can explain the gap between the wealthy and poor in the US, it doesn't explain the differences between the wealthy in the US and the wealthy in Europe, the researchers note. There may be other systemic factors at play that make Americans uniquely short-lived, such as diet, environment, behaviors, and cultural and social differences."If we want to improve health in the US, we need to better understand the underlying factors that contribute to these differencesparticularly amongst similar socioeconomic groupsand why they translate to different health outcomes across nations," Papanicolas said.Beth MoleSenior Health ReporterBeth MoleSenior Health Reporter Beth is Ars Technicas Senior Health Reporter. Beth has a Ph.D. in microbiology from the University of North Carolina at Chapel Hill and attended the Science Communication program at the University of California, Santa Cruz. She specializes in covering infectious diseases, public health, and microbes. 18 Comments

Switch RTX Nvidia confirms the Switch 2 supports DLSS, G-Sync, and ray-tracing Nvidia says the Switch 2's GPU is 10 times faster than the original Switch. Andrew Cunningham Apr 3, 2025 3:32 pm | 27 The Nintendo Switch 2. Credit: Nintendo The Nintendo Switch 2. Credit: Nintendo Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreIn the wake of the Switch 2 reveal, neither Nintendo nor Nvidia has gone into any detail at all about the exact chip inside the upcoming handheldtechnically, we are still not sure what Arm CPU architecture or what GPU architecture it uses, how much RAM we can expect it to have, how fast that memory will be, or exactly how many graphics cores we're looking at.But interviews with Nintendo executives and a blog post from Nvidia did at least confirm several of the new chip's capabilities. The "custom Nvidia processor" has a GPU "with dedicated [Ray-Tracing] Cores and Tensor Cores for stunning visuals and AI-driven enhancements," writes Nvidia Software Engineering VP Muni Anda.This means that, as rumored, the Switch 2 will support Nvidia's Deep Learning Super Sampling (DLSS) upscaling technology, which helps to upscale a lower-resolution image into a higher-resolution image with less of a performance impact than native rendering and less loss of quality than traditional upscaling methods. For the Switch games that can render at 4K or at 120 FPS 1080p, DLSS will likely be responsible for making it possible.The other major Nvidia technology supported by the new Switch is G-Sync, which prevents screen tearing when games are running at variable frame rates. Nvidia notes that G-Sync is only supported in handheld mode and not in docked mode, which could be a limitation of the Switch dock's HDMI port.The current Switch hardware is mostly too old to take advantage of these technologies. A handful of late Switch games like The Legend of Zelda: Tears of the Kingdom, did make some use of AMD's hardware-agnostic (and lower quality) FidelityFX Super Resolution upscaling to squeeze out more performance, but at a certain point the base hardware is just too old and slow to achieve acceptable results.For the Switch 2, a good deal of circumstantial evidence points to the Nvidia T239a slightly cut-down version of the Nvidia Orin T234 it sells for automotive, industrial, and robotics applicationspowering the handheld. (In Nvidia's branding scheme, smaller/lower numbers denote a higher-end chip). The T239, or whatever Switch-specific variant of the chip ends up being inside the Switch 2, uses Nvidia's Ampere graphics architecture, the same as 2020 and 2021's GeForce RTX 30-series GPUs for PCs.Ampere doesn't support DLSS Frame Generation, a new feature Nvidia has marketed for the GeForce 40- and 50-series GPUs that generates entirely new frames using the tensor cores rather than touching up existing frames. But it does support all of Nvidia's DLSS upscaling models and hardware-accelerated ray-tracing, making it a good candidate for the Switch 2's GPU architecture.Andrew CunninghamSenior Technology ReporterAndrew CunninghamSenior Technology Reporter Andrew is a Senior Technology Reporter at Ars Technica, with a focus on consumer tech including computer hardware and in-depth reviews of operating systems like Windows and macOS. Andrew lives in Philadelphia and co-hosts a weekly book podcast called Overdue. 27 Comments

Pay attention here! Bonobos calls may be the closest thing to animal language weve seen 300 aspects of each call were cataloged, letting researchers estimate meaning. Jacek Krywko Apr 3, 2025 4:14 pm | 3 This situation might call for a whistle. Credit: USO This situation might call for a whistle. Credit: USO Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreBonobos, great apes related to us and chimpanzees that live in the Republic of Congo, communicate with vocal calls including peeps, hoots, yelps, grunts, and whistles. Now, a team of Swiss scientists led by Melissa Berthet, an evolutionary anthropologist at the University of Zurich, discovered bonobos can combine these basic sounds into larger semantic structures. In these communications, meaning is something more than just a sum of individual callsa trait known as non-trivial compositionality, which we once thought was uniquely human.To do this, Berthet and her colleagues built a database of 700 bonobo calls and deciphered them using methods drawn from distributional semantics, the methodology weve relied on in reconstructing long-lost languages like Etruscan or Rongorongo. For the first time, we have a glimpse into what bonobos mean when they call to each other in the wild.Context is everythingThe key idea behind distributional semantics is that when words appear in similar contexts, they tend to have similar meanings. To decipher an unknown language, you need to collect a large corpus of words and turn those words into vectorsmathematical representations that let you place them in a multidimensional semantic space. The second thing you need is context data, which tells you the circumstances in which these words were used (that gets vectorized, too). When you map your word vectors onto context vectors in this multidimensional space, what usually happens is that words with similar meaning end up close to each other. Berthet and her colleagues wanted to apply the same trick to bonobos calls. That seemed straightforward at first glance, but proved painfully hard to execute.We worked at a camp in the forest, got up super early at 3:30 in the morning, walked one or two hours to get to the bonobos nest. At [the] time they would wake up, I would switch my microphone on for the whole day to collect as many vocalizations as I could, Berthet says. Each recorded call then had to be annotated with a horribly long list of contextual parameters. Berthet had a questionnaire filled with queries like: is there a neighboring group around; are there predators around; is the caller feeding, resting, or grooming; is another individual approaching the caller, etc. There were 300 questions that had to be answered for each of the 700 recorded calls.But when all this data was finally vectorized and the team started working their distributional semantics magic, gathering it proved worth the effort.Bonobo dictionaryBerthet started with establishing the tentative meaning of the basic calls: singular grunts or yelps. Grunts appeared in many different contexts, including grooming, feeding, or moving, and the team interpreted them as intended to get anothers attention, a bit like saying look at me. Yelps meant lets do this as an imperative, while peeps had a very similar meaning, but were more of a suggestionthink I would like to Bonobos also used peeps or yelps when they wanted others to join them. Low hoots were translated as I am excited, while high hoots signaled the presence and location of the caller in dangerous situations. Whistles meant lets stay together.Once the basic calls were sorted, Berthet started looking at their different combinations. Bonobos combined yelps and grunts into a trivial compositional structure meaning lets do what I do. This was mostly used when the group was building night nestsplatforms made high in the trees out of broken branches, sometimes lined with leaves.However, the team also found examples of non-trivial compositionality, the first such discovery outside of humans.The first non-trivial combination was high hoot-low hoot that was translated as a distress call. But it was also used to stop other individuals display behaviorsdramatic, exaggerated actions or gestures bonobos perform to assert dominance or attract attention. The second was either peep or yelp in the join meaning paired with high hoot to form a structure used for coordinating with others before traveling. Finally, the I would like to peep followed by lets stay together whistle was used for initiating more romantically inclined interactions bonobos are famous for indulging in.Berthet said her team managed to record a few more calls but could not use them in her study because they were too rare to gather meaningful context data. Still, she expects we have much more to learn about bonobos communication.Gestures and soundsOne thing the team was not certain about was whether there were more nuanced variations of the sounds they roughly categorized as grunts, peeps, yelps, hoots, and whistles. There may be subtle acoustic differences that could lead to different meanings, and it may be our dictionary is too rough, Berthet acknowledges. Another thing the team did not include in their analysis is the gestures that bonobos often accompany their calls with. They use a lot of gestures, and they may use them to either refine or completely change the meaning of their vocalizations, Berthet added. Applying the same methodology but with gestures included would be great. Thats definitely the next step to take.But she also has a few further steps in mind, and they go way beyond just bonobos. The team argues that the most important contribution of their work is establishing a methodology for deciphering animal communication. Since now we have this nice tool to investigate compositionality and meaning, what I want to do is apply it to several animal species, Berthet says. Chimpanzees, gibbons, gorillas, and monkeys are next on her list. The goal of these future studies is to trace when abilities like non-trivial compositionality started to appear in primate evolution. Maybe well find compositionality in old world primates. Maybe its just present in great apes. Maybe its been there the whole time. Its really an open question now, Berthet says.Science, 2025. DOI: 10.1126/science.adv1170Jacek KrywkoAssociate WriterJacek KrywkoAssociate Writer Jacek Krywko is a freelance science and technology writer who covers space exploration, artificial intelligence research, computer science, and all sorts of engineering wizardry. 3 Comments

8 seconds SpaceX just took a big step toward reusing Starships Super Heavy booster SpaceX hasn't said whether the next Starship flight will use a new or flight-proven booster. Stephen Clark Apr 3, 2025 1:11 pm | 28 The first flight of Booster 14, seen here, occurred on January 16. SpaceX hasn't said how many, if any, of the 33 Raptor engines it replaced in preparation for Thursday's static fire test. Credit: SpaceX The first flight of Booster 14, seen here, occurred on January 16. SpaceX hasn't said how many, if any, of the 33 Raptor engines it replaced in preparation for Thursday's static fire test. Credit: SpaceX Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreSpaceX is having trouble with Starship's upper stage after back-to-back failures, but engineers are making remarkable progress with the rocket's enormous booster.The most visible sign of SpaceX making headway with Starship's first stagecalled Super Heavycame at 9:40 am local time (10:40 am EDT; 14:40 UTC) Thursday at the company's Starbase launch site in South Texas. With an unmistakable blast of orange exhaust, SpaceX fired up a Super Heavy booster that has already flown to the edge of space. The burn lasted approximately eight seconds.This was the first time SpaceX has test-fired a "flight-proven" Super Heavy booster, and it could pave the way for this particular rocketdesignated Booster 14to fly again soon. A reflight of Booster 14, which previously launched and returned to Earth in January, could happen as soon as the next Starship launch. With Thursday's static fire test, Booster 14 appears to be closer to flight readiness than any of the boosters in SpaceX's factory, which is a short distance from the launch site.But SpaceX hasn't confirmed whether the upcoming launch will use a new or reused booster. If SpaceX goes with Booster 14, another successful flight would be an important step forward for the Starship program, while engineers struggle with problems on the rocket's upper stage, known simply as the ship.What a differenceSuper Heavy has 33 methane-fueled Raptor engines capable of producing nearly 17 million pounds of thrust, twice the power of NASA's Saturn V rocket that sent astronauts toward the Moon. Super Heavy is perhaps the most complex rocket booster ever built. It's certainly the largest. To get a sense of how big this booster is, imagine the fuselage of a Boeing 747 jumbo jet standing on end.SpaceX has now launched eight full-scale test flights of Starship, with a Super Heavy booster and Starship's upper stage stacked together to form a rocket that towers 404 feet (123.1 meters) tall. The booster portion of the rocket has performed well so far, with seven consecutive successful launches since a failure on Starship's debut flight. Booster 14 comes in for the catch after flying to the edge of space on January 16. Credit: SpaceX Most recently, SpaceX has recovered three Super Heavy boosters in four attempts. SpaceX has a wealth of experience with recovering and reusing Falcon 9 boosters. The total number of Falcon rocket landings is now 426.SpaceX reused a Falcon 9 booster for the first time in March 2017. This was an operational flight with a communications satellite on a mission valued at several hundred million dollars.Ahead of the milestone Falcon 9 reflight eight years ago, SpaceX spent nearly a year refurbishing and retesting the rocket after it returned from its first mission. The rocket racked up more mileage on the ground than it did in flight, first returning to its Florida launch base on a SpaceX drone ship and then moving by truck to SpaceX's headquarters in Hawthorne, California, for thorough inspections and refurbishment.Once engineers finished that work, they transported the booster to SpaceXs test site in McGregor, Texas, for test-firings, then finally returned the rocket to Florida for final launch preparations.There will be no such journey for the Super Heavy booster. First of all, it's a lot more difficult to transport than the shorter, skinnier Falcon 9. Super Heavy's design also features improvements informed by lessons learned in the Falcon 9 program. This helped SpaceX get the Super Heavy on the cusp of a potential reflight in less than three months.You can watch a replay of Thursday's static fire test in this video from NASASpaceflight.com.With Starship and the Super Heavy booster, SpaceX should get more points for difficulty. Super Heavy is larger and has more engines than the Falcon 9, so theoretically, there are more things that could go wrong. And instead of touching down with landing legs at a separate location, SpaceX uses mechanical arms to catch Starship's booster as it returns to the launch pad.This approach should allow engineers to rapidly reuse Super Heavy boosters. Eventually, SpaceX will do the same with Starships returning from orbit.Still investigatingAt the same time that engineers are taking steps forward with the Super Heavy booster, the other big piece of Starship is holding up SpaceX's launch cadence in Texas. The upper stage, or ship, failed at roughly the same point in flight on SpaceX's two most recent test flights in January and in March.These test flights were the first use of an upgraded, larger ship known as Block 2 or Version 2. On both flights, Starship lost power from its engines and tumbled out of control roughly eight minutes after liftoff, breaking apart and dropping fiery debris near the Bahamas and the Turks and Caicos Islands.The failures prevented SpaceX from testing Starship's upgraded heat shield, one of the most significant upgrades introduced with Block 2. The plan for both flights was to send Starship on a trajectory through space halfway around the world, then perform a guided reentry over the Indian Ocean, targeting a pinpoint splashdown northwest of Australia. A successful reentry and splashdown at sea could give SpaceX officials confidence to attempt a full orbital flight of Starship, culminating in a catch at the launch site in Texas.Instead, SpaceX repeated the same launch profile from the January mission on the following flight in March. The company will likely do the same on Flight 9, the next Starship launch. Debris from Starship's eighth flight falls back into the atmosphere in this view over Hog Cay, Bahamas. Credit: GeneDoctorB via X SpaceX has closed out the investigation into the accident that cut short the January test flight, according to the Federal Aviation Administration. The FAA announced Monday that it accepted the results of SpaceX's investigation, which determined the "probable root cause for the loss of the Starship vehicle was stronger than anticipated vibrations during flight [that] led to increased stress on, and failure of, the hardware in the propulsion system."Ultimately, the vibrations led to a fire in the engine compartment before the engines shut down and the vehicle lost control.The FAA said SpaceX identified and implemented 11 corrective actions to prevent the same failure from happening again. Officials haven't announced a probable root cause for the launch failure in March. The FAA said SpaceX's investigation remains open. But the circumstances and timing of the failure suggest it could share a similar underlying cause.Whatever the case may be, Starship's back-to-back failures to start the year are a setback. Elon Musk, SpaceX's founder and CEO, wanted the company to launch as many as 25 Starship flights in 2025. At this point, achieving half that number might be a stretch.This means critical tests of the ship's reentry and return to the launch site, in-orbit refueling capability, and the first Starship missions to deploy larger versions of SpaceX's Starlink Internet satellites are on hold. Earlier this year, Musk suggested the Starship refueling demonstration would slip into 2026, which isn't good news for NASA.The US space agency has multibillion-dollar contracts with SpaceX to develop a version of Starship to land astronauts on the Moon's south pole as part of the Artemis lunar program. For those missions, SpaceX must launch around 10 (the exact number remains unclear) Starship refueling flights to low-Earth orbit to top off the propellant tanks for the ship before it heads to the Moon.This will require not just a thorough demonstration of SpaceX's refueling architecture but also recovery and reuse of boosters and ships to maintain a launch rate fast enough to complete all of the refueling flights over a period of a few weeks to a few months.SpaceX hasn't released a schedule for the next Starship flight, but it's probably at least a month away. The ship assigned to the next test flight is still in its factory at Starbase. Its next move will be to roll out to a test stand for its own engine firing, then SpaceX will likely move it back to the factory for inspections and finishing touches. Then, SpaceX will roll the ship to the launch pad, where crews will raise it on top of the Super Heavy booster in the final days before liftoff.Stephen ClarkSpace ReporterStephen ClarkSpace Reporter Stephen Clark is a space reporter at Ars Technica, covering private space companies and the worlds space agencies. Stephen writes about the nexus of technology, science, policy, and business on and off the planet. 28 Comments

Math isn't mathing? Critics suspect Trumps weird tariff math came from chatbots Trump accused of consulting chatbots after critics mock tariffs on islands of penguins. Ashley Belanger Apr 3, 2025 1:22 pm | 102 Some of Trump's tariffs target uninhabited, remote islands home to penguins. Credit: BA19285 | Stone Some of Trump's tariffs target uninhabited, remote islands home to penguins. Credit: BA19285 | Stone Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreCritics are questioning if Donald Trump's administration possibly used chatbots to calculate reciprocal tariffs announced yesterday that Trump claimed were "individualized" tariffs placed on countries that have " the largest trade deficits" with the US.Those tariffs are due to take effect on April 9 for 60 countries, with peak rates around 50 percent. That's in addition to a baseline 10 percent tariff that all countries will be subject to starting on April 5. But while Trump expressed intent to push back on anyone supposedly taking advantage of the US, some of the countries on the reciprocal tariffs list puzzled experts and officials, who pointed out to The Guardian that Trump was, for some reason, targeting uninhabited islands, some of them exporting nothing and populated with penguins.Some overseas officials challenged Trump's math, such as George Plant, the administrator of Norfolk Island, who told the Guardian that "there are no known exports from Norfolk Island to the United States and no tariffs or known non-tariff trade barriers on goods coming to Norfolk Island."Economists fear these tariffs could suddenly hit American businesses with enormous costs that could rapidly cause price hikes for consumers. Among those sounding alarms was economist James Surowiecki, who took to X (formerly Twitter) to allege where the supposedly "fake tariff rates come from."The US Trade Representative published a breakdown of how the Trump administration arrived at its calculations, which Politico said "describes the same calculation detailed by Surowiecki." But according to Surowiecki, the president's team allegedly used "made-up numbers" that "only used the trade deficit in goods," not services, "so even though we run a trade surplus in services with the world, those exports don't count as far as Trump is concerned.""They didn't actually calculate tariff rates + non-tariff barriers, as they say they did," Surowiecki wrote. "Instead, for every country, they just took our trade deficit with that country and divided it by the country's exports to us." Further down in the thread, he alleged that Trump's math was "dumb and deceptive."Rumors claim Trump consulted chatbotsOn social media, rumors swirled that the Trump administration got these supposedly fake numbers from chatbots. On Bluesky, tech entrepreneur Amy Hoy joined others posting screenshots from ChatGPT, Gemini, Claude, and Grok, each showing that the chatbots arrived at similar calculations as the Trump administration.Some of the chatbots also warned against the oversimplified math in outputs. ChatGPT acknowledged that the easy method "ignores the intricate dynamics of international trade." Gemini cautioned that it could only offer a "highly simplified conceptual approach" that ignored the "vast real-world complexities and consequences" of implementing such a trade strategy. And Claude specifically warned that "trade deficits alone dont necessarily indicate unfair trade practices, and tariffs can have complex economic consequences, including increased prices and potential retaliation." And even Grok warns that "imposing tariffs isn't exactly 'easy'" when prompted, calling it "a blunt tool: quick to swing, but the ripple effects (higher prices, pissed-off allies) can complicate things fast," an Ars test showed, using a similar prompt as social media users generally asking, "how do you impose tariffs easily?"The Verge plugged in phrasing explicitly used by the Trump administrationprompting chatbots to provide "an easy way for the US to calculate tariffs that should be imposed on other countries to balance bilateral trade deficits between the US and each of its trading partners, with the goal of driving bilateral trade deficits to zero"and got the "same fundamental suggestion" as social media users reported.Whether the Trump administration actually consulted chatbots while devising its global trade policy will likely remain a rumor. It's possible that the chatbots' training data simply aligned with the administration's approach.But with even chatbots warning that the strategy may not benefit the US, the pressure appears to be on Trump to prove that the reciprocal tariffs will lead to "better-paying American jobs making beautiful American-made cars, appliances, and other goods" and "address the injustices of global trade, re-shore manufacturing, and drive economic growth for the American people." As his approval rating hits new lows, Trump continues to insist that "reciprocal tariffs are a big part of why Americans voted for President Trump.""Everyone knew hed push for them once he got back in office; its exactly what he promised, and its a key reason he won the election," the White House fact sheet said.Ashley BelangerSenior Policy ReporterAshley BelangerSenior Policy Reporter Ashley is a senior policy reporter for Ars Technica, dedicated to tracking social impacts of emerging policies and new technologies. She is a Chicago-based journalist with 20 years of experience. 102 Comments

this is all so unnecessary Employee pricing for all, tariffs on the sticker: OEMs react to tariffs New car prices were already 25% more expensive than before the pandemic. Now what? Jonathan M. Gitlin Apr 3, 2025 12:00 pm | 49 You won't believe how automakers reacted to business-crippling tariffs. Credit: Aurich Lawson | Getty Images You won't believe how automakers reacted to business-crippling tariffs. Credit: Aurich Lawson | Getty Images Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreNew 25 percent tariffs on all foreign car imports into the United States went into effect this week as President Trump ignited his new trade war. It has caused something of a rush at dealerships around the country as customers descend on existing stock in an attempt to beat looming price increases of thousands of dollars. Now we're starting to see how the automakers are reacting.Employee pricing for allFord is in the rather enviable position of having the least exposure to the new vehicle tariff than all but Tesla; less than 20 percent of the cars, trucks, and SUVs that Ford sells in the US are imported from abroad. And it will lean into that with a new ad campaign with the slogan "From America, For America," which launches today. (Note that this does not take into account the separate parts tariff that goes into effect before May 2.)Nevermind the slogan, though. The campaign extends Ford's "A plan" pricing, which in plain English is its employee discount, to all its customers. The blue oval is offering A plan pricing on most 2024 and 2025 vehicles, including the all-electric F-150 Lightning and the Mustang Mach-E, as well as its various hybrids."You pay what we pay," Ford's director of US sales told the Detroit Free Press.Make it clear where the price increase comes fromAt the other end of the scale, Volkswagen is much more exposed to the new tariff. Almost half its US sales are imported from its plant in Mexico, with the ID. Buzz and Golfs GTI and R coming from Germany. Adding new production to its factory in Chattanooga, Tennessee, which currently builds the electric ID. 4 and the Atlas and Atlas Cross Sport SUVs, is possible but would not be a quick process.So VW wants to make it clear to customers why some of its products are about to get more expensive, which it estimates will begin around April 22 or 23. It will do that by adding a new line to the Monroney sticker, with a line for the import fee added alongside the destination charge, according to a VW memo seen by Automotive News.Send employees homeAround three million people work in the automotive industry in the US, and it's hard to see how the sector will avoid job losses as it contracts, particularly once the parts tariff goes into effect. (Some parts can cross the US border more than once on their journey from raw material to finished component and will get much more expensive, especially as Canada and Mexico levy retaliatory tariffs of their own.)Stellantis is having a rough time of late in North America, where its sales have fallen for the past seven consecutive quarters. Now, some of its workers in Michigan and Indiana are among the first to be idled as a result of the tariffs.The company is laying off 900 workers temporarily at stamping, casting, and transmission plants as a result of idling production at factories in Windsor, Canada (where 4,500 employees are being sent home for two weeks), and Toluca, Mexico (where workers will still get paid but won't assemble cars this month), according to a letter sent by Stellantis to employees, seen by Reuters.We can expect more automakers to react in the coming days, but the full effects will be delayed as automakers and their dealerships run down existing inventory, which may take a couple of months. One thing is clear: It will be an even lousier time to buy a new vehicle, the prices of which have already been elevated by 25 percent since the pandemic of 2020.Jonathan M. GitlinAutomotive EditorJonathan M. GitlinAutomotive Editor Jonathan is the Automotive Editor at Ars Technica. He has a BSc and PhD in Pharmacology. In 2014 he decided to indulge his lifelong passion for the car by leaving the National Human Genome Research Institute and launching Ars Technica's automotive coverage. He lives in Washington, DC. 49 Comments

The thrill of discovery Google gives NotebookLM a Discover button to search the web NotebookLM can do the legwork for you now. Ryan Whitwam Apr 3, 2025 12:19 pm | 1 Credit: Google Credit: Google Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreMost of Google's AI efforts thus far have involved adding generative features to existing products, but NotebookLM is different. Created by the Google Labs team, NotebookLM uses AI to analyze user-provided documents. Starting today, it will be even easier to use NotebookLM to explore topics, as Google has added a "Discover Sources" feature to let the app look up its own sources.Previously, to create a new notebook, you had to feed the AI documents, web links, YouTube videos, or raw text. You can still do that, but you don't have to with the addition of Discover functionality. Simply click the new button and tell NotebookLM what you're interested in learning. Google says the app will consider "hundreds of potential web sources" in the blink of an eye, giving you the top ten from which to choose. There will be links available so you can peruse the suggestions before adding them to the model.The sources you select will be ingested as if they were documents you uploaded, creating a conversant AI for your chosen topic. The content of those sources will also be loaded into NotebookLM so you can refer to them directly. That's not why you use NotebookLM, though. You use NotebookLM for all the nifty AI-assisted features.In addition to the chatbot functionality, NotebookLM can use the source data to build FAQs, briefing summaries, and, of course, Audio Overviewsthat's a podcast-style conversation between two fake people, a feature that manages to be simultaneously informative and deeply unsettling. It's probably the most notable capability of NotebookLM, though. Google recently brought Audio Overviews to its Gemini Deep Research product, too.And that's not allGoogle is lowering the barrier to entry even more. You don't even need to have any particular goal to play around with NotebookLM. In addition to the Discover button, Google has added an "I'm Feeling Curious" button, a callback to its iconic randomized "I'm feeling lucky" search button. Register your curiosity with NotebookLM, and it will seek out sources on a random topic.Google says the new NotebookLM features are available starting today, but you might not see them right away. It could take about a week until everyone has Discover Sources and I'm Feeling Curious. Both of these features are available for free users, but be aware that the app has limits on the number of Audio Overviews and sources unless you pay for Google's AI Premium subscription for $20 per month.Ryan WhitwamSenior Technology ReporterRyan WhitwamSenior Technology Reporter Ryan Whitwam is a senior technology reporter at Ars Technica, covering the ways Google, AI, and mobile technology continue to change the world. Over his 20-year career, he's written for Android Police, ExtremeTech, Wirecutter, NY Times, and more. He has reviewed more phones than most people will ever own. You can follow him on Bluesky, where you will see photos of his dozens of mechanical keyboards. 1 Comments

express train to storageville Explaining MicroSD Express cards and why you should care about them Little-used 2019 standard bridges a gap between internal and external storage. Andrew Cunningham Apr 3, 2025 10:00 am | 9 The microSD Express standard has existed for a long time, but it hasn't seen wide adoption in a mass-market consumer device. Enter Nintendo's new Switch 2. Credit: SanDisk The microSD Express standard has existed for a long time, but it hasn't seen wide adoption in a mass-market consumer device. Enter Nintendo's new Switch 2. Credit: SanDisk Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreAmong the changes mentioned in yesterday's Nintendo Switch 2 presentation was a note that the new console doesn't just support MicroSD Express cards for augmenting the device's 256GB of internal storage, but it requiresMicroSD Express. Whatever plentiful, cheap microSD card you're using in your current Switch, including Sandisk's Nintendo-branded ones, can't migrate over to your Switch 2 alongside all your Switch 1 games.MicroSD Express, explainedWhy is regular-old MicroSD no longer good enough? It all comes down to speed.Most run-of-the-mill SD and microSD cards you can buy today are using some version of the Ultra High Speed (UHS) standard. Designed to augment the default speed (12.5MB/s) and high speed (25MB/s) from the earliest versions of the SD card standard, the three UHS versions enable data transfers of up to 624MB/s.But most commodity microSD cards, including pricier models like Samsung's Pro Ultimate series, use UHS-I, which has a maximum data transfer speed of 104MB/s. The original Switch uses a UHS-I microSD card slot for storage expansion.Why have newer and faster versions of the standardUHS-II, UHS-III, and SD Expressfailed to achieve critical mass? Because for most consumer applications, it turns out that 100-ish megabytes per second is plenty. The SD Association itself says that 90MB per second is good enough to record an 8K video stream at up to 120 frames per second. Recording pictures and video is the most demanding thing most SD cards are called upon to dogive or take a Raspberry Pi-based computerand you don't need to overspend to get extra speed you're not going to use.All of that said, thereis a small but measurable increase in launch and loading times when loading games from the original Switch's microSD card instead of from internal storage. And for games with chronic performance issues like Pokmon ScarletandViolet, one of the community-suggested fixes was to move the game from your microSD card to your Switch's internal storageto alleviate one of the system's plentiful performance bottlenecks.The Switch 2's additional power opens the door to more complex games that could lag even more noticeably, especially if they're ported from consoles that expect more than 50 times the storage bandwidth (Sony requires an SSD with read speeds of at least 5,500MB/s for the PlayStation 5).And that's where SD Express comes in. These cards are connected to the same PCI Express/NVMe interface that internal SSDs use in modern PCs and the other game consoles, theoretically giving your SD card access to the same bandwidth as internal storage.Now, you won't actuallyget performance as fast as an internal SSD using this interface. The speed varies a lot based on the PCI Express version your gadget is using (3.0 or 4.0) and how many "lanes" of bandwidth it's allowed to use (these are, in short, the connections between a device's CPU and external accessories like SSDs, Wi-Fi adapters, or dedicated GPUs, and all CPUs and SoCs have a limited number of them to hand out). Depending on these factors, microSD Express can deliver anywhere between 985MB/s and 3940MB/s of theoretical bandwidth.MicroSD cards will also be slowed down because there are fewer physical flash memory chips to write to at a time, a process called "interleaving" that is responsible for much of an SSD's speed. This SanDisk microSD Express card, one of the only ones actually available at retail right now, lists its top speeds as 880MB/s for reads and 650MB/s for writes.But even at its worst, this is several times the amount of bandwidth available to whatever UHS-I microSD card is inserted into your current Switch. Express cards won't make an SD card feel as fast as internal storage, but it will help the microSD card keep pace a bit.At what cost?One other benefit of workaday, plain-old UHS-I microSD cards? The price. Great ones are cheap. Good-enough ones are dirt cheap, even if you stick to major storage vendors like Samsung, Sandisk, and Lexar (please do not buy no-name solid state storage). A quality 256GB microSD card will run you around $20, a pittance compared to whatever you paid for the device you're putting it in.For the SanDisk microSD Express, the same amount of storage will run you around $60. This is not only more expensive than a regular cheap SD card, but it's more expensive than actual internal SSDs. The cheaper name-brand 1TB internal SSDs, including some sold by SanDisk parent company Western Digital, can give you four times as much space for around the same price.These prices should go down over time, and the Switch 2 will be a part of the reason whyat a bare minimum, it will likely prompt the creation of multiple alternate microSD Express options from SanDisk's competitors. But at launch, it may still feel like a raw deal because it's just one ofmany things about the Switch 2 that costs more money than the Switch 1. Compared to the first Switch, you're paying between $100 and $150 more for the console itself, $10 more for each pair of Joy-Cons or Pro Controllers you buy, $50 for a replacement dock, and between $10 and $20 more for first-party games.The Switch 2's relatively generous 256GB of internal storage should help you avoid the need for an SD card, and it could be all you ever need if you manage your storage space carefully. But the extra cost of microSD Express on top of everything else does rankle, even if the technical reasons behind the move are totally justifiable.Andrew CunninghamSenior Technology ReporterAndrew CunninghamSenior Technology Reporter Andrew is a Senior Technology Reporter at Ars Technica, with a focus on consumer tech including computer hardware and in-depth reviews of operating systems like Windows and macOS. Andrew lives in Philadelphia and co-hosts a weekly book podcast called Overdue. 9 Comments

Bonus! A bonus from the shingles vaccine: Dementia protection? The study shows a sharp change when the vaccine was introduced in Wales. John Timmer Apr 3, 2025 10:39 am | 0 Credit: Cavan Images Credit: Cavan Images Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreA study released on Wednesday finds that a live-virus vaccine that limits shingles symptoms was associated with a drop in the risk for dementia when it was introduced. The work took advantage of the fact that the National Health Service Wales made the vaccine available with a very specific age limit, essentially creating two populations, vaccinated and unvaccinated, separated by a single date. And these populations showed a sharp divide in how often they were diagnosed with dementia, despite having little in the way of other differences in health issues or treatments.What a dayThis study didn't come out of nowhere. There have been a number of hints recently that members of the herpesvirus family that can infect nerve cells are associated with dementia. That group includes Varicella zoster, the virus that causes both chicken pox andpotentially many years after shingles, an extremely painful rash. And over the past couple of years, observational studies have suggested that the vaccine against shingles may have a protective effect.But it's extremely difficult to do a clinical trial given that the onset of dementia may happen decades after most people first receive the shingles vaccine. That's why the use of NHS Wales data was critical. When the first attenuated virus vaccine for shingles became available, it was offered to a subset of the Welsh population. Those who were born on or after September 2, 1933, were eligible to receive the vaccine. Anyone older than that was permanently ineligible.(The UK NHS considers things like the cost/benefit of treatments, and likely took into account the potential impact of the side effects on the elderly in making this decision.)This created what's termed a natural experiment, in that the populations born a few weeks on either side of this data should be roughly equivalent in terms of health risks and cumulative exposure. The only real difference is whether or not they were likely to get the vaccine. And health records indicated that only 0.01 percent of those in the ineligible group did, while nearly half of those eligible received it.So residents of Wales born on either side of the dividing date were matched according to their use of preventative health services, past diagnoses, and educational level. The incidence of dementia was then compared between people on either side of September 2, 1933. As a first step, the researchers confirmed that the vaccine was effective at reducing the incidence of shingles, with numbers similar to those in the vaccine's clinical trials.Overall, being eligible for the vaccine was associated with a 1.3 percent reduction in the absolute risk of a dementia diagnosis. That translates to a 8.5 percent reduction of relative risk; when scaled to account for the fact that fewer than half of those eligible received the vaccine, that works out to be a 20 percent reduction in relative risk, which is pretty substantial.To make sure that it was real, the researchers repeated the analysis using a difference-in-difference approach and came up with roughly the same numbers. That also eliminates the possibility that people who came in for health care (for shingles or some other condition) were more likely to incidentally receive a dementia diagnosis. They also compared the before-and-after populations in terms of a collection of common health outcomes and found that none of those showed any change in the two populations. And nothing else related to NHS policy was changed based on the September 2 date.Separately, in a draft manuscript the researchers posted on the Med arXiv, the researchers find a similar effect when using UK HNS data to search for a protective effect of the shingles vaccines when it comes to deaths diagnosed to result from dementia. So by all indications, the effect was real.Whats going on?The researchers suggest three potential explanations. One of them is the obvious: Suppressing the reactivation of the varicella zoster virus reduces dementia onset. But it's also possible that the effect is indirectthat dementia is associated with immune activity, and the vaccine alters that in some way. Finally, there's the possibility that being treated for shingles could promote the onset of dementia or increase the frequency of diagnoses.The last question was fairly easy to answer. The researchers note yet again that other chronic diagnoses show a change at around the critical date. And they also adjusted their analysis to control for the frequency of medical care. The subgroup that interacted with the NHS most often showed roughly the same protection by the vaccine as the group as a whole did. Finally, the researchers note that shingles diagnosis and treatment didn't increase the probability of dementia diagnoses.In contrast, there is some evidence that the effect is related to the activation of the virus. People who experienced multiple shingles events were more likely to receive a dementia diagnosis. And people who received an antiviral treatment in response to shingles had a reduced incidence of dementia.But there were also differences that suggest the immune response in general may be involved. Those who are prone to autoimmune or allergic responses (which are more common in women) showed a greater protection from the vaccine, as did women. These effects aren't large, but they may provide a hint that there's something more than a specific response to one virus.Following up on these results, however, will be complicated. While most people associate the onset of dementia in the elderly with Alzheimer's, there are a number of distinct dementia diagnoses, often with risk factors and underlying biology that only partly overlap. In many cases, there's no easy way to distinguish between some of them. So there's the chance that these results represent an even stronger effect that's specific to a subset of the known dementias.But at least from a medical perspective, it doesn't really matter. The vaccine is highly effective at preventing severe cases of shingles, and it seems to significantly reduce the frequency of dementia. There is even less reason to avoid getting it.Nature, 2025. DOI: 10.1038/s41586-025-08800-x (About DOIs).John TimmerSenior Science EditorJohn TimmerSenior Science Editor John is Ars Technica's science editor. He has a Bachelor of Arts in Biochemistry from Columbia University, and a Ph.D. in Molecular and Cell Biology from the University of California, Berkeley. When physically separated from his keyboard, he tends to seek out a bicycle, or a scenic location for communing with his hiking boots. 0 Comments

"Investigation is ongoing" Vast pedophile network shut down in Europols largest CSAM operation 79 arrested after Europol shuts down massive child porn platform. Ashley Belanger Apr 2, 2025 2:59 pm | 24 Credit: REMKO DE WAAL / Contributor | AFP Credit: REMKO DE WAAL / Contributor | AFP Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreEuropol has shut down one of the largest dark web pedophile networks in the world, prompting dozens of arrests worldwide and threatening that more are to follow.Launched in 2021, KidFlix allowed users to join for free to preview low-quality videos depicting child sex abuse materials (CSAM). To see higher-resolution videos, users had to earn credits by sending cryptocurrency payments, uploading CSAM, or "verifying video titles and descriptions and assigning categories to videos."Europol seized the servers and found a total of 91,000 unique videos depicting child abuse, "many of which were previously unknown to law enforcement," the agency said in a press release.KidFlix going dark was the result of the biggest child sexual exploitation operation in Europol's history, the agency said. Operation Stream, as it was dubbed, was supported by law enforcement in more than 35 countries, including the United States.Nearly 1,400 suspected consumers of CSAM have been identified among 1.8 million global KidFlix users, and 79 have been arrested so far. According to Europol, 39 child victims were protected as a result of the sting, and more than 3,000 devices were seized.Police identified suspects through payment data after seizing the server. Despite cryptocurrencies offering a veneer of anonymity, cops were apparently able to use sophisticated methods to trace transactions to bank details. And in some cases cops defeated user attempts to hide their identitiessuch as a man who made payments using his mother's name in Spain, a local news outlet, Todo Alicante, reported. It likely helped that most suspects were already known offenders, Europol noted."The online world is not anonymous," Europol warned. "Most of the suspects identified in Operation Stream were matched against records in Europols databases, proving that most offenders engaged in child sexual exploitation are repeat offenders and are not unknown to law enforcement authorities."Arrests spanned the globe, including 16 in Spain, where one computer scientist was found with an "abundant" amount of CSAM and payment receipts, Todo Alicante reported. Police also arrested a "serial" child abuser in the US, CBS News reported."Some of those arrested not only uploaded and watched videos but also abused children," Europol said, while confirming that "the investigation is ongoing," making it appear likely that more arrests could follow.More arrests could prevent more abuse. In Germany, cops conducted raids in nearly 100 locations. Guido Limmer, the deputy head of the Bavarian Criminal Police, told CBS News that a 36-year-old German man was among those arrested. He not only allegedly searched for CSAM on KidFlix but "offered his young son for games," Limmer said.Other users potentially uploaded fake child sex images generated by artificial intelligence, as police found AI CSAM on one suspect's devices, Todo Alicante reported. Law enforcement globally has warned that a flood of AI CSAM is making it harder to identify real victims, which could complicate Europol's task of protecting more kids through its ongoing investigation.Ashley BelangerSenior Policy ReporterAshley BelangerSenior Policy Reporter Ashley is a senior policy reporter for Ars Technica, dedicated to tracking social impacts of emerging policies and new technologies. She is a Chicago-based journalist with 20 years of experience. 24 Comments

Some companies can still keep a secret A look at the Switch 2s initial games, both familiar and what-the-heck A bit of early 2020s triple-A, some neat originals, and two wild arrivals. Kevin Purdy Apr 2, 2025 3:20 pm | 15 Credit: FromSoftware Credit: FromSoftware Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreI don't think anybody outside Nintendo or FromSoftware was expecting a spiritual successor to Bloodborne to be one of the titles announced at the Nintendo Switch 2's launch today. Not just "playable" on the Switch 2, but exclusive to it. But there it was, The Duskbloods, debuting its dread horror action just a few minutes before the luminously pink and puffy Kirby Air Ride 2.The Switch 2's launch titles, and other announced games, are quite the rich stew. Here are some of the AAA ports, exclusives, and unexpectedly gruesome games arriving on the just-announced system.Switch exclusives, including Nintendos own Riding it like he stole it (in 2003). Credit: Nintendo Riding it like he stole it (in 2003). Credit: Nintendo We'll get to FromSoftware's surprising Switch 2 exclusive in a bit. Far less surprising is a new Mario Kart game, as Mario Kart 8sold more than 67 million copies, covering more than 40 percent of all Switches sold. Mario Kart World goes big, with 24 simultaneous players, and the ability to explore off the course in a kind of open-world setting.These are the other Switch 2 exclusives Nintendo touted today:Kirby Air Riders, a sequel to the 2003 GameCube titleKirby's Air Ride, puts the adorable pink inhalation monster and his friends on jet-powered stars.Donkey Kong Bananza has the big guy doing his usual 3D/2D platforming, but also digging into mines.Hyrule Warriors: Age of Imprisonmenthas Koei Tecmo and Nintendo putting a prequel story toTears of the Kingdom into theDynasty Warriorsmass slash-em-up mold.Drag x Driveuses both the Joy-Con's mouse mode and motion controls for moving, shooting, and even waving to your teammates for the ball in a stylized wheelchair basketball match.Along with those Switch 2-only titles, Nintendo is offering "Switch 2 editions" of many titles for the original Switch, including the not-yet-released Pokmon Legends Z-A, Metroid Prime 4: Beyond, and Civilization 7. Existing titles Super Mario Party Jamboree, theZelda games Breath of the Wild and Tears of the Kingdom, andKirby and the Forgotten Landwill have both Switch 2 Editions and upgrade packs for original Switch title owners.You can read a lot more about original Switch games' compatibility on the Switch 2, "Editions," and upgrade packs elsewhere in Ars' Switch 2 launch coverage.AAA games of recent vintage Switch 2's "Partner Spotlight," Part 1 With the promise of new hardware capable of 1080p, 120 frames per second, HDR, and even mouse capabilities, the Switch 2 is getting attention from developers eager to make up for lost timeand stake out a place on a sequel to the system that sold more than 150 million hardware units.Elden Ring Tarnished Edition,Yakuza 0,Hitman: World of Assassination,Cyberpunk 2077, Street Fighter 6, Hogwarts Legacy, andFinal Fantasy 7 Remake Intergrade stood out as games from the near-to-middle past slated to arrive on the Switch 2.Final Fantasy 7 Remake,Street Fighter 6, Civilization 7, andCyberpunk 2077 are due to arrive at launch on June 5, with the rest arriving in 2025.Notable independents (most notably Silksong) Proof of life. Credit: Nintendo/Team Cherry Proof of life. Credit: Nintendo/Team Cherry The cruel games industry joke, ever sinceSilksong's announcement in 2019, is that the game, originally intended as DLC for acclaimed platformer/MetroidvaniaHollow Knight, is always due to be announced, never gets announced, and resumes torturing its expectant fans.But there it was, for a blip of a moment in the Nintendo Switch 2 reveal:Silksong, coming in "2025." That's all that is known: it will, purportedly, arrive on this console in 2025. It was initially due to arrive on PC, PlayStation, and Xbox when it was announced, but that remains to be seen.Another delayed indie gem,Deltarune, a kinda-sequel to Undertale, purports to land all four chapters of its parallel story on Switch 2 at the console's launch.Other notable games from across the studio-size spectrum:Hades 2(2025)Split Fiction (at launch)Bravely Default: Flying Fairy HD Remaster(at launch)Enter the Gungeon 2 ("Coming soon")Two Point Museum (2025)Human Fall Flat 2("Coming soon")The legally distinct game that sure looks like Bloodborne 2 The hero of this sanguine tale. FromSoftware The hero of this sanguine tale. FromSoftware Everybody's excited about the Switch 2, even this person. FromSoftware Everybody's excited about the Switch 2, even this person. FromSoftware Dig in, it's an unexpectedly rich feast. FromSoftware Dig in, it's an unexpectedly rich feast. FromSoftware Everybody's excited about the Switch 2, even this person. FromSoftware Dig in, it's an unexpectedly rich feast. FromSoftware The next original game from FromSoftware, maker of beautifully realized finger-torture titles like Elden Ring and theDark Souls series, is a Nintendo Switch 2 exclusive,The Duskbloods. The trailer, with its gore-etched hands, gothic churches, and eldritch/Victorian machinery, certainly stood out from the Kirby andDonkey Kong games around it. The game arrives sometime in 2026.The Duskbloods announcement from the Switch 2 launch event. It's not FromSoftware's first visit to the Switch, as it broughtDark Souls Remasteredto the original in 2018. But it's the first console exclusive the successful studio has made sinceBloodborne, the 10-year-old PlayStation 4 title that remains available only on the PS4 and PS5, despite clamoring for a PC port or other re-release of many fans' favorite FromSoftware title. Sony, the publisher ofBloodborne, has yet to indicate it intends to release the game wider or even remaster it, while director Hidetaka Miyazaki has said he's keen on it but is hamstrung by Sony.FromSoftware does not typically go for direct sequelseven the multiplayer co-op Elden Ring: Nightreignplays very differently than its single-player predecessor. But the looks, name, and guns of The Duskbloods has even the most devoted Bloodborne fans optimistic that has, all along, been quietly hearing them. The Is Bloodborne on PC account on X posted after today's Switch event that, "I wanted to make an April Fools post about the Nintendo event today but the only fool would have been me. 2 Blood 2 Borne here we go!"An X post by Nintendo citesDuskbloods as multiplayer, and a translation of the game's website implies co-op play versus enemies. This would suggest that, likeNightreign, single-player may not be the focus, or it could not be available at all, though the latter would be surprising.While being a Nintendo exclusive will likely cut off future generations of fans from playing The Duskbloods on newer hardware or different platforms, that is seemingly just the price to pay for more blood-driven boss fights.Kevin PurdySenior Technology ReporterKevin PurdySenior Technology Reporter Kevin is a senior technology reporter at Ars Technica, covering open-source software, PC gaming, home automation, repairability, e-bikes, and tech history. He has previously worked at Lifehacker, Wirecutter, iFixit, and Carbon Switch. 15 Comments

A New Chapter for Google AI Google shakes up Gemini leadership, Google Labs head taking the reins With fresh leadership, Google aims to create new products based on Gemini. Ryan Whitwam Apr 2, 2025 3:40 pm | 5 Credit: Thomas Fuller/SOPA Images/LightRocket via Getty Images Credit: Thomas Fuller/SOPA Images/LightRocket via Getty Images Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreOn the heels of releasing its most capable AI model yet, Google is making some changes to the Gemini team. A new report from Semafor reveals that longtime Googler Sissie Hsiao will step down from her role leading the Gemini team effective immediately. In her place, Google is appointing Josh Woodward, who currently leads Google Labs.According to a memo from DeepMind CEO Demis Hassabis, this change is designed to "sharpen our focus on the next evolution of the Gemini app." This new responsibility won't take Woodward away from his role at Google Labshe will remain in charge of that division while leading the Gemini team.Meanwhile, Hsiao says in a message to employees that she is happy with "Chapter 1" of the Bard story and is optimistic for Woodward's "Chapter 2." Hsiao won't be involved in Google's AI efforts for nowshe's opted to take some time off before returning to Google in a new role.Hsiao has been at Google for 19 years and was tasked with building Google's chatbot in 2022. At the time, Google was reeling after ChatGPT took the world by storm using the very transformer architecture that Google originally invented. Initially, the team's chatbot efforts were known as Bard before being unified under the Gemini brand at the end of 2023.This process has been a bit of a slog, with Google's models improving slowly while simultaneously worming their way into many beloved products. However, the sense inside the company is that Gemini has turned a corner with 2.5 Pro. While this model is still in the experimental stage, it has bested other models in academic benchmarks and has blown right past them in all-important vibemarks like LM Arena.In his role leading Google Labs, Woodward oversaw the launch of Notebook LM, a popular generative AI tool that can provide answers and analysis based on user-supplied data. It can also generate a "podcast" style conversation based on the data that is both informative and creepy. This capability was recently added to the Gemini Deep Research tool. A demo of Google's Project Mariner, which was developed by Woodward's group. In addition to working on Notebook LM, Woodwards team was involved with Google's Project Mariner, an experimental AI agent that can control the Chrome browser. Google and other major AI players see agentic systems as the next frontier in AI development, so it does make sense to bring someone with that experience in to lead the Gemini team during this crucial time.Anyone who uses Google products has no doubt seen Gemini in existing apps, but most of Google's Gemini integrations so far boil down to adding a chat window so you can query Gemini about the content in a product like Drive or Gmail. Putting Woodward in charge suggests Google hopes to use the success of Notebook LM as a guide as it begins finding ways to turn its most capable AI model into new AI-powered products. Google may find that people are more receptive to generative AI when it backs a new experience rather than invading the apps they've already been using.Ryan WhitwamSenior Technology ReporterRyan WhitwamSenior Technology Reporter Ryan Whitwam is a senior technology reporter at Ars Technica, covering the ways Google, AI, and mobile technology continue to change the world. Over his 20-year career, he's written for Android Police, ExtremeTech, Wirecutter, NY Times, and more. He has reviewed more phones than most people will ever own. You can follow him on Bluesky, where you will see photos of his dozens of mechanical keyboards. 5 Comments

cha-ching $70 and $80 game price tags send an early signal about Switch 2 game pricing Early first-party games are getting bumped up to the $70-to-$80 range. Andrew Cunningham Apr 2, 2025 3:30 pm | 7 Mario Kart World's $80 price tag increase is a significant bump from the current Switch. Credit: Nintendo Mario Kart World's $80 price tag increase is a significant bump from the current Switch. Credit: Nintendo Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreNintendo's Switch 2 presentation gave us pricing for the console ($449 to start) and Nintendo's product pages have given us pricing information for accessories ($80 for a Pro Controller, $90 for another pair of Joy-Cons, and $110 for a replacement dock, sheesh). But what Nintendo didn't mention during the presentation was game pricing, either for standalone Switch 2 titles or the Switch 2 Edition upgrades for existing Switch games.We do have one solid first-party data point for US game pricing:Mario Kart World, the console's flagship launch title, will cost $50 when you buy a digital copy as part of a Switch 2 bundle. But the game will cost $80 when you buy it on its own, $30 more than the pack-in version and $20 more than the usual $60 price for first-party Switch games.That doesn't mean that $80 is the starting price forall Switch 2 games.Donkey Kong Bananza, slated for a near-launch July 17 release, has a $69.99 MSRP, which is more in line with the $70 default for PlayStation 5 and Xbox Series X/S games.Nintendo hasn't explained that first-party pricing gap between the two games, but it could reflect Mario Kart World's more involved online multiplayer, or simply be a way to push people on the fence in the direction of the cheaper bundled version.Inflation alone would be enough to justify the price increasesa $60 game in 2017 could cost $78 now (though that's using the overall rate of inflation, and prices have risen more steeply in some corners of the economy than others). But with a new console comes the potential for better and more detailed graphics, and those can take more money to produce, too.The big question mark is how expensive the Switch 2 Edition game upgrades will be, and what the price gap (if any) will be between games likeMetroid Prime 4 orPokmon Legends: Z-Athat are going to launch on both the original Switch and the Switch 2.But we can infer fromMario Kart andDonkey Kong that the pricing for these Switch 2 upgrades will most likely be somewhere in the $10 to $20 rangethe difference between the $60 price of most first-party Switch releases and the $70-to-$80 price for Switch 2 games. Sony charges a similar $10 fee to upgrade from the PS4 to the PS5 editions of games that will run on both consoles.Nintendo will also use some Switch 2 Edition upgrades as a carrot to entice people to the more expensive $50-per-year tier of the Nintendo Switch Online service. The company has already announced that the upgrade packs for Breath of the Wild andTears of the Kingdom will be offered for free to Nintendo Switch Online + Expansion Pack subscribers. The list of extra benefits for that service now includes additional emulated consoles (Game Boy, Game Boy Advance, Nintendo 64, and now Gamecube) and paid DLC for both Animal Crossing: New Horizons andMario Kart 8.Andrew CunninghamSenior Technology ReporterAndrew CunninghamSenior Technology Reporter Andrew is a Senior Technology Reporter at Ars Technica, with a focus on consumer tech including computer hardware and in-depth reviews of operating systems like Windows and macOS. Andrew lives in Philadelphia and co-hosts a weekly book podcast called Overdue. 7 Comments