Theo on tech · June 14th

Welcome back to the Abstract! Here are the studies this week that died in the deep, let nature call, tossed a galactic salad, and became interstellar voyeurs. First, there’s a whale necropolis under the sea that is packed with ancient carcasses and teeming with new species. Then: a bygone world preserved in poop, the fruits of the universe’s labor, and a zoom lens for distant planets. As always, for more of my work, check out my book First Contact: The Story of Our Obsession with Aliens, or subscribe to my personal newsletter the BeX Files. Scientists have discovered an unprecedented underwater “necropolis” that contains the remains of hundreds of whales that died over the past five million years, scattered across 745 miles. During dives in a deep sea submersible, researchers spotted whale bones submerged under more than four miles of the Diamantina Zone in the Indian Ocean, making this site the geographically largest, deepest, and oldest whale necropolis ever found. The graveyard is also teeming with species that may be “new to science” and subsist on these fortuitous “whale falls,” according to a new study. “The discovery of whale-fall communities in the Diamantina Zone at depths exceeding 6,700 meters establishes one of the deepest known whale-fall ecosystems in the ocean, extending the known depth range of such habitats by more than 2,500 meters,” said researchers co-led by Xiaotong Peng of China’s Institute of Deep-sea Science and Engineering. “This area has a deep and extensive accumulation comprising five modern natural whale-fall communities and 476 fossil cetaceans recorded,” the team said. Peng and his colleagues first spotted the necropolis during dives in early 2023 using the Fendouzhe submersible, which is capable of bringing crews to depths of nearly seven miles. The team quickly realized they had tapped into a scientific motherlode, complete with an immense fossil archive of extinct animals—mostly deep-diving beaked whales—along with recent whale falls that still support thriving ecosystems of crustaceans, molluscs, worms, and microbes. “Bone-eating worms, gastropods, vesicomyid bivalves and brittle stars dominate the megafauna (more than several centimetres in size), reaching local densities up to 2,840 individuals per square metre,” the team said. “Most recovered taxa may be new to science.” As for why this vast necropolis formed, beaked whales may be attracted to these deep waters due to the abundance of prey sources, such as squid and fish. Some might accidentally dive so deep that they experience decompression sickness or fatal exhaustion, becoming bonus bodies for seafloor ecosystems. The sinking carcasses are then funnelled into the Diamantina Zone because of its V-shaped topography, serving up a figurative feast for scientists (and a literal one for marine biota). “As beaked whales are known primarily from rare strandings, their abundance, distribution and ecology remain poorly understood overall,” Peng and his colleagues concluded. “Our discovery of an accumulation of skeletal remains…provides an unparalleled source of information on these largely enigmatic cetaceans.” Mariners have long dreaded ending up in Davy Jones’ locker, the proverbial resting ground of drowned sailors. It turns out that whales have a whole locker room down in the deep, where the bodies of countless leviathans blossom into fleeting hotspots of life. In other news… The Klondike region of Canada’s Yukon territory is famous for the 19th-century gold rush that led hopeful prospectors to riches, ruin, and early graves. But now, scientists have found a very different type of valuable nugget in Klondike soil—ancient squirrel poops made by ancient squirrel bums as early as 700,000 years ago. Scientists sequenced ancient environmental DNA (aeDNA) from these permafrosted scats, thereby opening up a poopy portal into the past. The fossilized feces, known as coprolites, contained genetic traces of mammoth, saber-tooth cat, horse, and bison, suggesting that these Ice Age rodents may have gnawed on the corpses of much larger megafauna. The coprolites also preserved DNA from hundreds of plant species, several insects, and a bevy of microbial and fungal strains. “The diversity and abundance of aeDNA recovered from the permafrost preserved, ground squirrel coprolites presented here underscores the immense value of Arctic rodent middens as repositories of Quaternary ecosystems,” said researchers led by Tyler J. Murchie of the Hakai Institute and McMaster University.

ABOVE: In celebration of the return of the U.S. Green Berets, China In Arms is offering t-shirts and coffee mugs with this emblem. I recommend getting the image on the back of a black t-shirt. Click HERE. China In Arms BOOKSTORE and GIFT SHOP! Twitter and YouTube Page and LinkedIn Subscribe: $5 Monthly; Cancel Anytime 14 June 2026 (Sunday) By Wendell Minnick (Whiskey Mike) 顏文德 TAIPEI - New evidence confirms that Green Berets from the 1st Special Forces Group (based in Okinawa) actively train Taiwanese special operation forces, far more than the simple observation and evaluation that many had speculated. Social media posts further reveal that U.S. Special Forces actively train Taiwan’s Army Sniper Teams, National Police Agency Special Operations Group (NPA SOG), and Army Aviation and Special Forces Command (ASFC) in Close Quarters Battle (CQB). Taiwan continues to use English on many of its military emblems, likely a lasting legacy of the U.S. military presence on the island from the late 1940s to 1979 through the U.S.-Taiwan Defense Command and other American airbases and naval facilities. The U.S. military and the CIA actively used Taiwan as a base for both special operations and overt missions during the Korean War and the Vietnam War. The CIA operated front airlines such as Civil Air Transport (CAT) and Air Asia, and it ran logistics support for its covert activities through a front company called Western Enterprises on the island. The return of U.S. Special Forces to Taiwan was blocked for decades by the U.S. State Department’s One-China Policy. I will not identify the source material of these images to protect his career, but they do appear genuine and are from the same source. These events appear to have occurred in the 2023 timeframe (Year of the Rabbit). These require some explanation to the uninitiated, so excuse my exposition. For reference, this is the official Taiwan Army Aviation and Special Forces Command (ASFC) emblem: The official US Army Special Forces Group (Airborne) stationed in Okinawa, Japan (Operational Detachment Alpha = ODA): I can confirm that the first two numbers of this Green Beret Operational Detachment Alpha (ODA) are “14,” indicating ODA 14xx. This identifies it as a 12-man A-Team from the 1st Special Forces Group (1), 4th Battalion (4). The third and fourth numbers (Company and Specialty/Team designation) remain unclear. The first patch shared by the U.S. source features Traditional Chinese text with the standard orange-and-yellow instructor tab: Close Quarters Battle Instructor (限制空間戰鬥教官). CQB (Close Quarters Battle) is required training for all Taiwan special forces, military police, and special operations units. Operators master room clearing, urban combat, and confined-space tactics. Analysts sometimes refer to it as Close Quarters Combat (CQC). The other item on display is the traditional green beret with the 1st SFG insignia with two Taiwan military emblems: Taiwan Army Special Forces (bottom right) and the Taiwan Army Sniper School (middle). Close Up of Both: Also from the source material is the image below. Taiwan Army Snipers also wear the center emblem below that reads “One Shot, One Kill” at the bottom. The top features “SFC,” which stands for Special Forces Command. The source also displays the challenge coin below. It belongs to Taiwan’s National Police Agency Special Operations Group (NPA SOG). This elite unit proudly carries the motto “Others Guard, We Assault” (餘皆牧犬，吾為兇狼 — “The rest are shepherd dogs; I am the fierce wolf”). The NPA SOG, also known as the Wei-An Forces (officially the Safety Maintenance Special Operations Group or 維安特勤隊), serves as Taiwan’s primary national-level counter-terrorism and tactical unit. Operating under the Special Police First Headquarters of the Ministry of the Interior, it handles high-risk operations, hostage rescues, and VIP protection. It has already been confirmed by China In Arms that the 1st Special Forces Group has established a liaison relationship with the NPA SOG. Plain clothes members of the 1st SFG are often sighted at the Sheraton Grand Taipei Hotel (aka Lai Lai Sheraton) and is a five minute walk from the NPA building. The image below proved difficult to identify, even though the “ODA” clearly stands for Operational Detachment Alpha. A secondary source in Taiwan confirms it belongs to ODA 14. The image below is difficult to discern, but the central design features what appears to be a stylized skull or tactical emblem with crossed elements (common in U.S. Special Forces ODA patches) The image below remains a mystery, even though the Taiwan and U.S. flags stand out clearly. The prominent animal figure, possibly a rabbit, looks particularly strange. If anyone recognizes this patch, please leave a comment below.

May 2026 We currently live in multiple, overlappling fields in the Western world, when it comes to how different people think about AI, Artificial General Intelligence (AGI), and Artificial Superintelligence (ASI). For a quick update on these terms and how some frontier AI lab researchers think of them, consider this table from Google Deepmind: Current AI systems are jagged in their performance and reliability, depending on the task and domain, and range from Level 1 to Level 4, and so these systems are confusing to use even for researchers and super-users. However, I already believe the current Claude Fable to GPT 5.5 Codex level models are super-intelligent in many ways, albeit jagged, and that future models trained on Rubins and Feynman chips will operate as swarms and be smarter, more capable (autonomous and agentic over long horizons), and reliable. Here is how people in the developed world tend to think of AI: Normies: They discount AI and think of it as unimportant and over-hyped, or at best a “normal technology” like laptops and cell-phones. At best they think AI adds 10 to 30bps per year or productivity or GDP growth, per Acemoglu’s macroeconomics of AI work. This world see AI assistants are marginally useful, agents as doing some work in a slow changing economy, and technological growth similar to the last 40 years, roughly 2% income and productivity growth per year, or 2.2%. The tech tree: chatbots, AI-copilots, better Uber and Lyft, a handful of new drugs and consumer devices, etc; no major health or lifespan changes. AI-pilled: They think AI will have large consequences for our business, society, economics, and politics, and will lead to large wealth creation and political changes (maybe 30-80bps of additional productivity and GDP growth per year). The average Fortune 500 business exec or Silicon Valley tech leader is here, along with McKinsey Global Institute or Erik Brynjolfsson, and maybe the Pope with his recent encyclical. Economic growth could range from 2% to 4% (see the Goldman Sachs estimates here and here), and the big changes are how fast we can get self-driving cars on the road, increase airplane slots and products, build up the grid and more factories, and generally use AI tools to augment existing workers to make them more productive. The tech tree: teams of humans manage larger teams of AI agents, everyone drives Waymo level cars, air travel drops costs by 50% of more and humans travel much more, dozens of new drugs and transformative consumer devices over a decade, average US lifespans rise to 90 or 100. AGI-pilled: They think we are close to expert human-level AI agents (within 2 to 4 years), across 90% or more of all cognitive jobs and tasks that humans use a computer to do. So for the tasks that are mostly or completely done on a computer, agents can take over, nudging or pushing humans to pick up more physical or non-computer tasks. Hence the nature of work will drastically change, and there will be a major social, economic, and political upheaval (but much larger and faster than AI-pilled people believe - perhaps GDP growth goes into an annual 5% to 19% range in the US (see Chad Jones’ work) or other heavy adopter countries as AGI takes off). This is where the average frontier AI lab (eg. OpenAI, Anthropic, Meta MSL, etc) employee is, as they see AI replace jobs and tasks inside their company that used to require hours or days of work for conscientious, 120 to 150 IQ workers. Large hyperscalers and tech companies are ramping up capex from the current $600bn annual levels to $1trn or higher, expecting to reap a massive reward from AGI. The tech tree: teams of humans manage larger teams of AI agents, until the AGI agents get so much better the humans aren’t needed other than occasionally to come discuss preferences and cars; everyone drives Waymo level cars, which can also fly distances up to 200 miles; air travel drops costs by 80% of more and humans travel much more in business and first class like planes that get to destinations at half or a third of the speed; hundreds of new drugs and transformative consumer devices over a decade, average US lifespans rise to 100 to 120, and we find ways to reverse or pause ageing and a tail upside is much longer lifespans and healthspans (gene editing treaments and easy organ replacements let you revert your biological age back to 25 or 30). ASI-pilled: They think we are close to an AGI world (3 to 10 years), with hundreds of millions of genius-level AI agents (Dario’s “datacenters of geniuses”). But the AGI equilibrium won’t last (it may only be around for a year to a decade). ASI occurs as the AGIs start to recursively self-improve and evolve, both becoming much more intelligent and effective with current levels of AI infrastructure investment, while also designing better hardware, robots, datacenters, and energy systems, while producing them in such a way that capex goes much higher than $1 trillion.

Hi, fellow future and current Data Leaders; Ben here 👋 It’s 2026, and everyone wants to talk about AI engineers, FDEs, and agents. But in almost every company I work with, the real bottleneck is still the same thing it's often been…messy data, weak data foundations, and teams that need help making data valuable. Now before we jump in to talking about the data world in 2026, I wanted to share a bit about Estuary, a platform I’ve used to help make clients’ data workflows easier and am an adviser for. Estuary helps teams easily move data in real-time or on a schedule, from databases and SaaS apps to data lakes and warehouses, empowering data leaders to focus on strategy and impact rather than getting bogged down by infrastructure challenges. If you want to simplify your data workflows, check them out today. Now let’s jump into the article! When I first started in the data world, the sexiest job of the 21st century was a data scientist. Of course, that’s what I wanted to do! Even when it came to consulting, that’s what I was looking for. Data science projects! But you know what I found out...every data science project was +80 percent data engineering. And really, the skills that got you far were really solid fundamentals. In 2015, if you knew SQL, Python, and data modeling well, you could get a job at a lot of companies. Even Facebook’s interview was really just SQL, Python, and data modeling for the most part. But it’s starting to feel like what is being demanded of new engineers and analysts is even more than that as if the world is shifting away. The coolest job? It’s somewhere between AI Engineer and FDE. And that requires a whole new set of skills. But does it? I wanted to share some of my thoughts in terms of what I am seeing in the data world in 2026 and what you should do to break in. Even if the job titles change, many of the fundamental skills and needs of the business don’t. If we roll back time by ten to fifteen years, and you were wanting to learn the basics as a programmer and developer, you likely picked up a lot of ancillary skills you didn’t even realize. Spinning up Docker containers, ingesting CSV files with wonky formats and jagged rows, working with SFTP or other ancient technologies. All so you could set up Airflow. Ok, maybe not Airflow, but you’d likely find yourself having to do a lot of pre-work. In turn, you’d pick up what I like to call “glue technical skills”. The stuff that’s in between running data pipelines and data warehouses. The stuff that actually makes everything run. The stuff you assume everyone knows, but no one who hasn’t had to struggle with it for four hours on a weekend knows. It’s also likely what will give you a good sense when AI gives you garbage vs something that looks like it’s going in the right direction. The basic skills you pick up are your foundation(I should also add that data modeling is also critical, perhaps that is another future article). Sure, they might not always be the coolest skills. But they are often the most useful. They let you understand how to build more reliable AI systems and workflows. In fact, many of the skills I see people incorporate into these systems…are traditional programming and system design in one way or another. I do tend to have a bias here. I went through a similar hype cycle in the culinary industry. And as one of my chefs used to say: Maybe it’s also why I recently wrote the post below: Because sure, we might, in the future, get further abstracted away from some of the work we are doing today. But I don’t think it will ever completely disappear. The same way, to this day, there is still a need to write in lower-level languages. Along with the basics still being relevant, here is another thing I still see as relevant. Or at least, the act of centralizing data in a single location and making it more accessible to multiple users(and now machines) to help improve performance, reduce costs, and create a more consistent set of naming and entities. Don’t get me wrong. I am sure there are plenty of people saying, “ Hey, just leave your data in the source system or in some very raw state, and then have your AI go over that. Yeah, we did that in 2010; it was called schema on read. It went really poorly. It’ll also likely drive your token costs through the roof. So the labs will be happy. I am not saying a data warehouse or lake house is the answer to everything. There are plenty of people who are doing just fine on reporting on database replicas. But as your company grows and has increasing data and complexity. It is often a good choice. And even now, in the era of AI. I am seeing the same pattern. Companies’ data is disparate, and they feel limited on the questions they can answer, so they don’t dig too deeply since everyone is busy. Then, someone finally does centralize their data, and then the floodgates are open. The challenge then becomes focusing on the right use cases, ensuring there is some governance.

Apple announced iOS 27 at WWDC this week, coming with a ton of new Siri and Apple Intelligence features. One of the biggest changes here is the addition of an all-new Siri app. This marks the first time that Siri has been pre-installed as a standalone app for iPhone users …

I was digging into the FCC filings and found the first hint that Beats is already shipping a two‑tone over‑ear model, even though we haven’t seen it on store shelves yet. The data shows a fresh SKU, which lines up with the pair Antonee Robinson flaunted at the World Cup—so Beats is definitely using influencers to tease the drop.

What’s odd is the lack of detail on the finish. Earlier photos only showed a single shade, but Robinson’s set mixes colors, and there’s no word on whether you can swap ear cups or pick other combos. That suggests a possible customization angle, or maybe it’s a limited‑run piece made just for a handful of creators.

The filing doesn’t clarify if this is a revamped Beats Studio Pro or an entirely new line. The specs look similar to the Studio Pro, yet the branding could be pointing to a fresh product family.

All we know for now is that the launch timing is still a mystery, and the color palette and availability are being kept under wraps. I’ll keep an eye on the FCC updates and let you know when anything concrete surfaces.

Wine‑Staging 11.11 lands with almost 300 patches layered on top of the fresh upstream code, and the most noticeable shift is how the Wayland driver now handles window scaling. The driver’s internal buffer management was reworked, so apps that previously jittered when you moved between monitors now stay smooth, and the compositor gets a cleaner handshake that reduces latency.

Beyond the Wayland tweaks, the patch set brings a batch of low‑level fixes that tighten DirectX 11 translation. A few of the new shims correct how shader constants are packed, which clears up a handful of texture glitches that showed up in newer games. There’s also a modest bump in Vulkan support: the DXVK‑like layer now respects a few more Vulkan extensions, letting some titles that were stalling on older builds finally start.

On the stability side, the maintainers trimmed a handful of regressions that had slipped in after the last upstream release. Crash‑guard patches around thread synchronization were added, and a couple of memory‑leak guards were tightened, meaning the runtime stays a bit more resilient when you push it with heavy workloads.

All in all, if you’re already on Wine‑Staging, updating to 11.11 gives you a smoother Wayland experience, a few extra DirectX quirks ironed out, and a quieter, more reliable runtime. It’s the kind of incremental polish that makes the whole stack feel a little less experimental and a lot more usable.

"Battery breakthroughs will lessen AI's demand on the electricity grid," argues The Washington Post's editoral board, arguing that GM's latest moves "offer a fresh reminder that resource constraints can be solved by innovation." Or As Fortune put it, "America's electric grid is buckling under extreme weather, aging infrastructure, and an AI build-out that is quietly rewriting U.S. power demand — and General Motors wants to turn that crisis into a business." They describe GM's plan as offering itself "as a distributed utility in disguise... stitching together hundreds of thousands of battery-powered cars, new grid-scale storage, and a unified charging platform into what amounts to a virtual fleet of power plants." The bet puts GM on a collision course with Ford's newly branded Ford Energy unit as both Detroit rivals race to repurpose underused EV capacity for a more urgent problem: keeping the lights on in the AI era. GM's case rests on three planks. The first is its existing fleet. GM says more than 250,000 of its EVs on U.S. roads can already charge bidirectionally — pulling electricity from the grid and sending it back. "Every evening, a quiet transformation occurs across the American landscape," GM Energy vice president Wade Sheffer writes in an open letter to utilities and regulators, describing the EVs sitting in driveways as "a massive opportunity to aggregate energy storage capacity." A firmware update is rolling out to customers with GM Energy's vehicle-to-home hardware, converting those systems into full vehicle-to-grid assets with no new hardware and turning home backup systems into grid resources when utilities need them. GM is piloting the idea in Michigan with DTE Energy at 30 employee homes, and has sketched a 2030 vision with Pacific Gas & Electric in which more than 52,000 GM EVs help balance the grid out of a projected 130,000 vehicles in the area. GM is also "seeking partnerships with utility companies nationwide to assist in offering such vehicle-to-grid services for customers," reports CNBC, noting it's one of two moves "meant to address concerns about rising energy costs amid an artificial intelligence boom." Forbes reports that GM's second goal "is to leapfrog the dominant battery cell tech used for energy storage packs right now" — right past the LFP (lithium-iron phosphate) stage, "which is dominated by China." Sodium batteries are cheaper to use than LFP because they don't need an additional cooling system. They also have a 20-year usable life and are made from materials that can be sourced from within the U.S., the company said at a briefing in San Francisco on Tuesday. "Sodium-ion actually is the better chemistry for that application. And when I say sodium-ion is better, I mean GM's version of sodium-ion," Kurt Kelty, GM's battery chief and a long-time Tesla battery executive, told Forbes. He said GM is seeing great results from its prototypes, even at scorching temperatures of 55 Celsius (131 Fahrenheit). "Sodium-ion-powered energy storage systems have the potential to operate without active cooling and with much less system complexity," Kurt Kelty, GM's vice president of battery and sustainability, said Tuesday in a blog post. "In large energy storage systems, that matters." Not having to cool the battery cells could lead to lower upfront costs as well as operating costs, the automaker said. TechCrunch reports on GM's big new partnership with energy-storage startup Peak Energy to develop GM's sodium-ion battery chemistry for grid-scale deployments: GM wouldn't share with TechCrunch how much money it is investing in this energy-storage effort. But we do know the company has committed $900 million to commercialize new battery chemistries, an investment that includes a new battery-development center. .. The first GM cells are expected to enter trial production at the company's Battery Cell Development Center in 2028. "Our next-generation sodium-ion cell development will drive energy density higher," promises GM's blog post, arguing they're extending the company's battery expertise and technical infrastructure "into the electrical grid itself. If we get this right, we will not just build better batteries. We will help create a more resilient, more affordable and more flexible energy future... Every improvement we make strengthens the development stack that supports both EVs and energy storage." "The message: GM isn't just selling cars into a stressed grid; it's supplying the batteries to stabilize it," argues Fortune. And GM also announced they're augmenting their apps with an "Energy Pass" offering "seamless access to Tesla Supercharger, IONNA, Electrify America, and soon, ChargePoint and EVgo networks." Their goal is to simplify the charging experience with an app "that covers nearly 70% of all DC fast chargers in the United States, plus many Level 2 chargers, all through one app."Read more of this story at Slashdot.

Shutterstock’s new suite stitches its huge catalog of contributor photos and videos straight into an AI workflow, so the moment you type a prompt the system can pull relevant assets, suggest tweaks, and even spin out fresh images or clips without you leaving the same interface. What’s different under the hood is the way the platform automatically picks the right model for each task and layers a conversational search that feels more like chatting with a colleague than hunting through folders.

The tools also let you remix existing content—crop, recolor, or add elements with a few clicks—while the royalty engine still tracks back to the original creators. That safety net is meant to keep the human side visible, even as the AI does most of the heavy lifting.

It’s not a solo act; Adobe, Canva and a swarm of startups are already weaving similar AI helpers into their pipelines. Shutterstock’s angle is the tight integration of its licensed library with the generative engine, aiming to shave the time from idea to finished piece in one place.

The Wall Street Journal reports: The Trump administration's decision to halt all foreign use of Anthropic's most capable AI models was prompted by conversations between Amazon Chief Executive Andy Jassy and U.S. officials including Treasury Secretary Scott Bessent, people familiar with the matter said. Researchers at Amazon had used a series of prompts to get Anthropic's Fable 5 model to provide them with information that could be used to aid cyberattacks and was supposed to be off limits, Jassy told the officials, according to people familiar with the matter. Tech industry executives have been in regular touch with the administration about the power of cutting-edge AI tools. Shortly afterward, White House officials held a meeting to discuss how to respond and security researchers began testing Amazon's claims. The officials asked Anthropic to fix the vulnerabilities or take down the model, according to administration officials. The officials decided that the most direct way to address that risk was by preventing foreign governments, companies and individuals from accessing the tool, the people said. President Trump later signed off on the action despite reservations about it hindering innovation, a senior White House official said. The administration had long felt that Anthropic, one of the leaders in America's AI race, couldn't be trusted to manage the security risks its new model presented. Friday's call between some administration officials and Anthropic Chief Executive Dario Amodei reinforced that feeling, the people said... Anthropic has said that the vulnerabilities like those flagged by Amazon are relatively basic. The company has said that other publicly available models are capable of discovering them and that they don't represent a full so-called jailbreak, a point of view shared by some security researchers familiar with Amazon's research. The article points out that Amazon is "a big investor in Anthropic, supply Anthropic with chips for data centers.Read more of this story at Slashdot.