Anthropic’s Restraint Is a Terrifying Warning Sign

Normally right now I would be writing about the geopolitical implications of the war with Iran, and I am sure I will again soon. But I want to interrupt that thought to highlight a stunning advance in artificial intelligence — one that arrived sooner than expected and that will have equally profound geopolitical implications.

The artificial intelligence company Anthropic announced Tuesday that it was releasing the newest generation of its large language model, dubbed Claude Mythos Preview, but to only a limited consortium of roughly 40 technology companies, including Google, Broadcom, Nvidia, Cisco, Palo Alto Networks, Apple, JPMorganChase, Amazon and Microsoft. Some of its competitors are among these partners because this new A.I. model represents a “step change” in performance that has some critically important positive and negative implications for cybersecurity and America’s national security.

The good news is that Anthropic discovered in the process of developing Claude Mythos that the A.I. could not only write software code more easily and with greater complexity than any model currently available, but as a byproduct of that capability, it could also find vulnerabilities in virtually all of the world’s most popular software systems more easily than before.

The bad news is that if this tool falls into the hands of bad actors, they could hack pretty much every major software system in the world, including all those made by the companies in the consortium.

This is not a publicity stunt. In the run-up to this announcement, representatives of leading tech companies have been in private conversation with the Trump administration about the implications for the security of the United States and all the other countries that use these now vulnerable software systems, technologists involved told me.

For good reason. As Anthropic said in its written statement on Tuesday, in just the past month, “Mythos Preview has already found thousands of high-severity vulnerabilities, including some in every major operating system and web browser. Given the rate of A.I. progress, it will not be long before such capabilities proliferate, potentially beyond actors who committed to deploying them safely. The fallout — economics, public safety and national security — could be severe.’’

Project Glasswing, Anthropic’s name for the consortium, is an undertaking to work with the biggest and most trusted tech companies and critical infrastructure providers, including banks, “to put these capabilities to work for defensive purposes,” the company added, and to give the leading technology firms a head start in finding and patching those vulnerabilities.

“We do not plan to make Claude Mythos Preview generally available, but our eventual goal is to enable our users to safely deploy Mythos-class models at scale — for cybersecurity purposes, but also for the myriad other benefits that such highly capable models will bring,” Anthropic said.

My translation: Holy cow! Superintelligent A.I. is arriving faster than anticipated, at least in this area. We knew it was getting amazingly good at enabling anyone, no matter how computer literate, to write software code. But even Anthropic reportedly did not anticipate that it would get this good, this fast, at finding ways to find and exploit flaws in existing code.

Anthropic said it found critical exposures in every major operating system and Web browser, many of which run power grids, waterworks, airline reservation systems, retailing networks, military systems and hospitals all over the world.

If this A.I. tool were, indeed, to become widely available, it would mean the ability to hack any major infrastructure system — a hard and expensive effort that was once essentially the province only of private-sector experts and intelligence organizations — will be available to every criminal actor, terrorist organization and country, no matter how small. [...]

That is why Anthropic is giving carefully controlled versions to key software providers so they can find and fix the vulnerabilities before the bad guys do — or your kids.

At moments like this I prefer to do a deep dive with my technology tutor, Craig Mundie, a former director of research and strategy at Microsoft, a member of President Barack Obama’s President’s Council of Advisors on Science and Technology and an author, with Henry Kissinger and Eric Schmidt, of a book on A.I. called “Genesis.”

In our view, no country in the world can solve this problem alone. The solution — this may shock people — must begin with the two A.I. superpowers, the U.S. and China. It is now urgent that they learn to collaborate to prevent bad actors from gaining access to this next level of cyber capability.

Such a powerful tool would threaten them both, leaving them exposed to criminal actors inside their countries and terrorist groups and other adversaries outside. It could easily become a greater threat to each country than the two countries are to each other.

Indeed, this is potentially as fundamental and significant a turning point as was the emergence of mutually assured destruction and the need for nuclear nonproliferation. The U.S. and China need to work together to protect themselves, as well the rest of the world, from humans and autonomous A.I.s using this technology — a lot more than they need to worry about Russia.

This is so important and urgent that it should be a top subject on the agenda for the summit between Trump and President Xi Jinping in Beijing next month.

“What used to be the province of big countries, big militaries, big companies and big criminal organizations with big budgets — this ability to develop sophisticated cyberhacking operations — could become easily available to small actors,” explained Mundie. “What we are about to see is nothing short of the complete democratization of cyberattack capabilities.”

It means that responsible governments, in concert with the companies that build these A.I. tools and software infrastructure, need to do three things urgently, Mundie argues.

For starters, he says, we need to “carefully control the release of these new superintelligent models and make sure they only go to the most responsible governments and companies.”

Then we need to use the time this buys us to distribute defensive tools to the good actors “so that the software that runs their key infrastructure can have all their flaws found and fixed before hackers inevitably get these tools one way or another.” (By the way, the cost of fixing the vulnerabilities that are sure to be discovered in legacy software systems, like those of telephone companies, will be significant. Then multiply that across our whole industrial base.)

by Thomas Friedman, NY Times |  Read more:

Image: Vincent Forstenlechner/Connected Archives
[ed. No shit Sherlock. Basically, everything that runs on software is vulnerable (including all forms of infrastructure). It's only what everyone's been saying for months now, if not years. Maybe this will finally get someone's attention, but who? Congress can't even rouse itself to engage with a war and a mentally unstable President. So all the enablers (politicians, banks, hedge funds, corporations) will finally get to meet their Frankenstein and are appropriately freaking out. See also: Anthropic Claims Its New A.I. Model, Mythos, Is a Cybersecurity ‘Reckoning’ (NYT):]

***

Claude Mythos Preview is already capable of carrying out autonomous security research, including scanning for and exploiting so-called zero-day vulnerabilities in critical software programs, flaws that are unknown even to the software’s developer. These efforts can often be triggered by amateurs with simple prompts. The company claims that the new model has already identified “thousands” of bugs and vulnerabilities in popular software programs, including every major operating system and browser.

One of the vulnerabilities Claude found, the company said, was a 27-year-old bug in OpenBSD, an open-source operating system that was designed to be difficult to hack. Many internet routers and secure firewalls incorporate OpenBSD’s technology. Another was a longstanding issue in a piece of popular video software that automated testing tools had scanned five million times, without finding any problems.

“This model is good at finding vulnerabilities that would be well understood and findable by security researchers,” Mr. Graham said. “At the same time, it has found vulnerabilities, and in some cases crafted exploits, sophisticated enough that they were both missed by literally decades of security researchers, as well as all the automated tools designed to find them.”

[ed. Probably a good idea to take a few screenshots of your bank accounts before they disappear.]

Intraventricular CARv3-TEAM-E T Cells in Recurrent Glioblastoma

In this first-in-human, investigator-initiated, open-label study, three participants with recurrent glioblastoma were treated with CARv3-TEAM-E T cells, which are chimeric antigen receptor (CAR) T cells engineered to target the epidermal growth factor receptor (EGFR) variant III tumor-specific antigen, as well as the wild-type EGFR protein, through secretion of a T-cell–engaging antibody molecule (TEAM). Treatment with CARv3-TEAM-E T cells did not result in adverse events greater than grade 3 or dose-limiting toxic effects. Radiographic tumor regression was dramatic and rapid, occurring within days after receipt of a single intraventricular infusion, but the responses were transient in two of the three participants. (Funded by Gateway for Cancer Research and others; INCIPIENT ClinicalTrials.gov number, NCT05660369.)

***

Glioblastoma is the most aggressive primary brain tumor, and the prognosis for recurrent disease is exceedingly poor with no effective treatment options. Chimeric antigen receptor (CAR) T cells represent a promising approach to cancer because of their proven efficacy against refractory lymphoid malignant neoplasms, for which they have become the standard of care. However, the use of CAR T cells in solid tumors such as glioblastomas has been limited to date, largely owing to the challenge in targeting a single antigen in a heterogeneous disease and to immunosuppressive mechanisms associated with the tumor microenvironment. 

In a previous clinical trial, we found that peripheral infusion of epidermal growth factor receptor (EGFR) variant III–specific CAR T cells (CART-EGFRvIII) safely mediated on-target effects in patients with glioblastoma. Despite this activity, no radiographic responses were observed, and recurrent tumor cells expressed wildtype EGFR protein and showed heavy intratumoral infiltration with suppressive regulatory T cells. To address these barriers, we developed an engineered T-cell product (CARv3-TEAM-E) that targets EGFRvIII through a second-generation CAR while also secreting T-cell–engaging antibody molecules (TEAMs) against wildtype EGFR, which is not expressed in the normal brain but is nearly always expressed in glioblastoma. We found in preclinical models that TEAMs secreted by CAR T cells act locally at the site where cognate antigen is engaged by the CAR T cells in the treatment of heterogeneous tumors. We also found in vitro that these molecules have the capacity to redirect even regulatory T cells against tumors. On the basis of these data, we initiated a first-in-human, phase 1 clinical study to evaluate the safety of CARv3-TEAM-E T cells in patients with recurrent or newly diagnosed glioblastoma. Here, we report the findings from a prespecified interim analysis involving the first three participants treated with this approach. [...]

Discussion

This study shows that antitumor CAR-mediated responses can be rapidly obtained in patients with glioblastoma, even in those with advanced, intraparenchymal cerebral disease. This finding contrasts with a previous report of a complete response that was observed in a patient with recurrent leptomeningeal disease who received treatment with 16 intracranial infusions of monospecific interleukin-13 receptor alpha 2 CAR T cells. It was hypothesized by the investigators of that study that the involvement of glioblastoma in the leptomeninges may have rendered the disease more responsive to intraventricular therapy. Our experience in the current study suggests that even a single dose of intraventricularly administered living drugs such as CAR T cells also have the capacity to access and mediate activity against infiltrative, parenchymal glioblastoma.

by Bryan D. Choi, M.D., Ph.D., Elizabeth R. Gerstner, M.D., Matthew J. Frigault, M.D., Mark B. Leick, M.D., Christopher W. Mount, M.D., Ph.D., Leonora Balaj, Ph.D., Sarah Nikiforow, M.D., Ph.D., Bob S. Carter, M.D., Ph.D., William T. Curry, M.D., Kathleen Gallagher, Ph.D., and Marcela V. Maus, M.D., Ph.D. NIH, National Center for Biotechnology Information |   Read more:

Image: via

[ed. Only three patients (so far) and it appears sustained treatments are needed to prevent recurrence. But still, pretty interesting.]

Man vs Mist vs Mountain

Something big is happening, but nothing big is happening to me.

Throughout my "career" as a "statistician", 13 years and counting now (but how much longer?), I've always been great at stopping myself from doing useful work. At first, I worried that I didn't know enough yet to tackle interesting problems---until I've started feeling that I forgot too much to do "real" statistics. With LLMs that barrier is now gone and I've been finding them very useful. I have just enough context and experience to pose good questions and understand the explanations.

(BTW, I am surprised by how little students and my peers seem to use them. I am usually, willingly, cast in the role of the nay-sayer. So what's happening? Are they using them surreptitiously? Or else, why do I get more utility than others?)...

So, obviously I decided to make this situation even worse and dip my toes into THE AGENTS this month (starting with the OpenAI one). In case you haven't encountered them, these are the ~latest craze in the LLM world.

Yes, just like with chatbots, you just describe what you'd like, in words, and it gets coded. But it's not just coding programs. You can do (some parts of) academic research or you can just make small, fun ideas come to life. I recently met a girl who vibe coded a Chinese medicine app that took photo of your tongue and told you seven things that were fucked up about your bladder.

Ultimately, however, my problem---because obviously I wouldn't bother to write this to just conclude that they're alright, would I---is that these tools are designed for people who like manipulating mental symbols in a certain way, you know, the screen-starers. Obviously this is a ridiculous complaint, not least because I am one of them... but as I get older, screen-staring part of my brain feels like the one I want to be visiting least often. And I think it was no coincidence that I had most fun playing with these tools when my mood was lowest.

In fact, they are addictive as hell, like a video game can feel. Everyone keeps reporting this. They dial difficulty down so much, that things get a bit muddy. People who talk about these models the most often seem to me maniacal, but I think these agents can stop you from getting actual work done. When I tried using these agents for my work, I ended up solving a lot of problems, but none of these problems seemed very important in retrospect.

Clearly that is a skill issue. I have no doubt that I'll get better at it. And if your work is mediated through screens and you're good at defining what you do and don't like, these agents may be great for me.

But at the same time, it feels like a general manifestation of any sort of "life-improving" technology, which is often just about channeling of mental disturbances. So, no, I am not banking on it making me a Nietzschean ubermensch next month; nor helping me start a billion dollar company, or even on having a better time. Right now it still feels net zero: for every bit of busy work that it may rescue me from, it feels like it has potential to rob meaningful work or meaning---or maybe even my life of life itself. "Projects" that I really care about in my life are not app-shaped or list-shaped, and in doing things, technology is always an afterthought.

by Witold Więcek, Monthly Witold | Read more:
Image: Strawberry in ASCII by Claude Sonnet 4.5 via:

[ed. More from Witold, about keeping a journal:]

***

I have been keeping a somewhat regular journal for close to 15 years now, a few pages per week usually. Most of it is very mundane, too, not even an attempt at recollection of what happened, more of a microcatalogue of internal states that feel new---I'd be less embarrassed by someone getting their paws on it as sorry for them.

Why do it? I used to call this project "long Witek", extracting what is slow-moving or semi-permanent from the detritus, the more transitory elements. I use these journals to sometimes jump back an arbitrary number of years and try to recognise myself again. In other words, I try to make myself legible to myself.

Yasushi Kishida (岸田 保), Old folk house in Muroo
via:

Sam Altman May Control Our Future—Can He Be Trusted?

[ed. A must read, possibly historic. Unfortuntately, the accompanying visual is too weird to include here. For a more concise summary see: A history and a proposal (DWAtV)]

In the fall of 2023, Ilya Sutskever, OpenAI’s chief scientist, sent secret memos to three fellow-members of the organization’s board of directors. For weeks, they’d been having furtive discussions about whether Sam Altman, OpenAI’s C.E.O., and Greg Brockman, his second-in-command, were fit to run the company. Sutskever had once counted both men as friends. In 2019, he’d officiated Brockman’s wedding, in a ceremony at OpenAI’s offices that included a ring bearer in the form of a robotic hand. But as he grew convinced that the company was nearing its long-term goal—creating an artificial intelligence that could rival or surpass the cognitive capabilities of human beings—his doubts about Altman increased. As Sutskever put it to another board member at the time, “I don’t think Sam is the guy who should have his finger on the button.”

At the behest of his fellow board members, Sutskever worked with like-minded colleagues to compile some seventy pages of Slack messages and H.R. documents, accompanied by explanatory text. The material included images taken with a cellphone, apparently to avoid detection on company devices. He sent the final memos to the other board members as disappearing messages, to insure that no one else would ever see them. “He was terrified,” a board member who received them recalled. The memos, which we reviewed, have not previously been disclosed in full. They allege that Altman misrepresented facts to executives and board members, and deceived them about internal safety protocols. One of the memos, about Altman, begins with a list headed “Sam exhibits a consistent pattern of . . .” The first item is “Lying.”

Many technology companies issue vague proclamations about improving the world, then go about maximizing revenue. But the founding premise of OpenAI was that it would have to be different. The founders, who included Altman, Sutskever, Brockman, and Elon Musk, asserted that artificial intelligence could be the most powerful, and potentially dangerous, invention in human history, and that perhaps, given the existential risk, an unusual corporate structure would be required. The firm was established as a nonprofit, whose board had a duty to prioritize the safety of humanity over the company’s success, or even its survival. The C.E.O. had to be a person of uncommon integrity. According to Sutskever, “any person working to build this civilization-altering technology bears a heavy burden and is taking on unprecedented responsibility.” But “the people who end up in these kinds of positions are often a certain kind of person, someone who is interested in power, a politician, someone who likes it.” In one of the memos, he seemed concerned with entrusting the technology to someone who “just tells people what they want to hear.” If OpenAI’s C.E.O. turned out not to be reliable, the board, which had six members, was empowered to fire him. Some members, including Helen Toner, an A.I.-policy expert, and Tasha McCauley, an entrepreneur, received the memos as a confirmation of what they had already come to believe: Altman’s role entrusted him with the future of humanity, but he could not be trusted. [...]

The day that Altman was fired, he flew back to his twenty-seven-million-dollar mansion in San Francisco, which has panoramic views of the bay and once featured a cantilevered infinity pool, and set up what he called a “sort of government-in-exile.” Conway, the Airbnb co-founder Brian Chesky, and the famously aggressive crisis-communications manager Chris Lehane joined, sometimes for hours a day, by video and phone. Some members of Altman’s executive team camped out in the hallways of the house. Lawyers set up in a home office next to his bedroom. During bouts of insomnia, Altman would wander by them in his pajamas. When we spoke with Altman recently, he described the aftermath of his firing as “just this weird fugue.”

With the board silent, Altman’s advisers built a public case for his return. Lehane has insisted that the firing was a coup orchestrated by rogue “effective altruists”—adherents of a belief system that focusses on maximizing the well-being of humanity, who had come to see A.I. as an existential threat. (Hoffman told Nadella that the firing might be due to “effective-altruism craziness.”) Lehane—whose reported motto, after Mike Tyson, is “Everyone has a game plan until you punch them in the mouth”—urged Altman to wage an aggressive social-media campaign. Chesky stayed in contact with the tech journalist Kara Swisher, relaying criticism of the board.

Altman interrupted his “war room” at six o’clock each evening with a round of Negronis. “You need to chill,” he recalls saying. “Whatever’s gonna happen is gonna happen.” But, he added, his phone records show that he was on calls for more than twelve hours a day. At one point, Altman conveyed to Mira Murati, who had given Sutskever material for his memos and was serving as the interim C.E.O. of OpenAI in that period, that his allies were “going all out” and “finding bad things” to damage her reputation, as well as those of others who had moved against him, according to someone with knowledge of the conversation. (Altman does not recall the exchange.) [...]

In a series of increasingly tense calls, Altman demanded the resignations of board members who had moved to fire him. “I have to pick up the pieces of their mess while I’m in this crazy cloud of suspicion?” Altman recalled initially thinking, about his return. “I was just, like, Absolutely fucking not.” Eventually, Sutskever, Toner, and McCauley lost their board seats. Adam D’Angelo, a founder of Quora, was the sole original member who remained. As a condition of their exit, the departing members demanded that the allegations against Altman—including that he pitted executives against one another and concealed his financial entanglements—be investigated. They also pressed for a new board that could oversee the outside inquiry with independence. But the two new members, the former Harvard president Lawrence Summers and the former Facebook C.T.O. Bret Taylor, were selected after close conversations with Altman. “would you do this,” Altman texted Nadella. “bret, larry summers, adam as the board and me as ceo and then bret handles the investigation.” (McCauley later testified in a deposition that when Taylor was previously considered for a board seat she’d had concerns about his deference to Altman.)

Less than five days after his firing, Altman was reinstated. Employees now call this moment “the Blip,” after an incident in the Marvel films in which characters disappear from existence and then return, unchanged, to a world profoundly altered by their absence. But the debate over Altman’s trustworthiness has moved beyond OpenAI’s boardroom. The colleagues who facilitated his ouster accuse him of a degree of deception that is untenable for any executive and dangerous for a leader of such a transformative technology. “We need institutions worthy of the power they wield,” Murati told us. “The board sought feedback, and I shared what I was seeing. Everything I shared was accurate, and I stand behind all of it.” Altman’s allies, on the other hand, have long dismissed the accusations. After the firing, Conway texted Chesky and Lehane demanding a public-relations offensive. “This is REPUTATIONAL TO SAM,” he wrote. He told the Washington Post that Altman had been “mistreated by a rogue board of directors.”

OpenAI has since become one of the most valuable companies in the world. It is reportedly preparing for an initial public offering at a potential valuation of a trillion dollars. Altman is driving the construction of a staggering amount of A.I. infrastructure, some of it concentrated within foreign autocracies. OpenAI is securing sweeping government contracts, setting standards for how A.I. is used in immigration enforcement, domestic surveillance, and autonomous weaponry in war zones.

Altman has promoted OpenAI’s growth by touting a vision in which, he wrote in a 2024 blog post, “astounding triumphs—fixing the climate, establishing a space colony, and the discovery of all of physics—will eventually become commonplace.” His rhetoric has helped sustain one of the fastest cash burns of any startup in history, relying on partners that have borrowed vast sums. The U.S. economy is increasingly dependent on a few highly leveraged A.I. companies, and many experts—at times including Altman—have warned that the industry is in a bubble. “Someone is going to lose a phenomenal amount of money,” he told reporters last year. If the bubble pops, economic catastrophe may follow. If his most bullish projections prove correct, he may become one of the wealthiest and most powerful people on the planet.

In a tense call after Altman’s firing, the board pressed him to acknowledge a pattern of deception. “This is just so fucked up,” he said repeatedly, according to people on the call. “I can’t change my personality.” Altman says that he doesn’t recall the exchange. “It’s possible I meant something like ‘I do try to be a unifying force,’ ” he told us, adding that this trait had enabled him to lead an immensely successful company. He attributed the criticism to a tendency, especially early in his career, “to be too much of a conflict avoider.” But a board member offered a different interpretation of his statement: “What it meant was ‘I have this trait where I lie to people, and I’m not going to stop.’ ” Were the colleagues who fired Altman motivated by alarmism and personal animus, or were they right that he couldn’t be trusted?

One morning this winter, we met Altman at OpenAI’s headquarters, in San Francisco, for one of more than a dozen conversations with him for this story. The company had recently moved into a pair of eleven-story glass towers, one of which had been occupied by Uber, another tech behemoth, whose co-founder and C.E.O., Travis Kalanick, seemed like an unstoppable prodigy—until he resigned, in 2017, under pressure from investors, who cited concerns about his ethics. (Kalanick now runs a robotics startup; in his free time, he said recently, he uses OpenAI’s ChatGPT “to get to the edge of what’s known in quantum physics.”)

An employee gave us a tour of the office. In an airy space full of communal tables, there was an animated digital painting of the computer scientist Alan Turing; its eyes tracked us as we passed. The installation is a winking reference to the Turing test, the 1950 thought experiment about whether a machine can credibly imitate a person. (In a 2025 study, ChatGPT passed the test more reliably than actual humans did.) Typically, you can interact with the painting. But the sound had been disabled, our guide told us, because it wouldn’t stop eavesdropping on employees and then butting into their conversations. Elsewhere in the office, plaques, brochures, and merchandise displayed the words “Feel the AGI.” The phrase was originally associated with Sutskever, who used it to caution his colleagues about the risks of artificial general intelligence—the threshold at which machines match human cognitive capacities. After the Blip, it became a cheerful slogan hailing a superabundant future.

We met Altman in a generic-looking conference room on the eighth floor. “People used to tell me about decision fatigue, and I didn’t get it,” Altman told us. “Now I wear a gray sweater and jeans every day, and even picking which gray sweater out of my closet—I’m, like, I wish I didn’t have to think about that.” Altman has a youthful appearance—he is slender, with wide-set blue eyes and tousled hair—but he is now forty, and he and Mulherin have a one-year-old son, delivered by a surrogate. “I’m sure, like, being President of the United States would be a much more stressful job, but of all the jobs that I think I could reasonably do, this is the most stressful one I can imagine,” he said, making eye contact with one of us, then with the other. “The way that I’ve explained this to my friends is: ‘This was the most fun job in the world until the day we launched ChatGPT.’ We were making these massive scientific discoveries—I think we did the most important piece of scientific discovery in, I don’t know, many decades.” He cast his eyes down. “And then, since the launch of ChatGPT, the decisions have gotten very difficult.”

by Ronan Farrow and Andrew Marantz, New Yorker | Read more:
Image: via

[ed. See also: “The problem is Sam Altman”: OpenAI Insiders don’t trust CEO (Ars Technica).]

Anton Elfilter, Shifting Tides

via:

China's AI Education Experiment

A deep dive.

Pilot schools in China are already using AI to grade children’s artwork, monitor their facial expressions during lectures, and screen them for psychological problems — and the Ministry of Education (MOE) wants schools across the country to follow suit.

Integrating AI into the education system has rapidly become a top priority of the Chinese central government, which is betting that AI tools can eliminate China’s vast educational inequities and make the next generation of workers more productive. The State Council highlighted education as a key area of focus in the “AI+” plan, it received a shout-out in the 15th Five-Year Plan, and in May 2025, the Ministry of Education (MOE) released a white paper on AI for education. This MOE document proclaims that 2025 marks the dawn of an era (“智慧教育元年”), the beginning of a system-wide effort to “intelligentize” 智能化 education using AI tools. The MOE’s goal: universalize basic AI access in primary and secondary schools by 2030. Industry received that signal and responded rapidly, with Alibaba Cloud releasing its own AI+education white paper the following month. But the gap between Beijing’s (and Hangzhou’s) techno-optimism and rural China’s reality is enormous.

This report explores why the Party wants to integrate AI into education, what applications the MOE is most optimistic about, and where the barriers to successful rollout lie. We’ll limit our analysis to K-12 education today, but university AI initiatives will be the focus of our next report in this series!

Institutional History

In official discourse, China is said to have entered a “post-equity era” 后均衡时代 since the MOE announced that all counties had met the baseline quality level for compulsory schooling in 2021. Now, the focus is shifting from access to education to improving the quality of that education. The 14th 5-year plan (2021-2025) prioritized expanding infrastructure in rural schools through the “county-level high school revitalization initiative” (县中振兴), part of which involved equipping classrooms with ‘smart hardware’ such as digitized blackboards. During this period, the party spent significant resources to provide nearly every school with an internet connection.

Still, rural education in China faces serious structural challenges. I spoke with Leo He — a research fellow at the Hoover Institution who did NGO work in rural China from 2019 to 2023 — for a firsthand account of the situation. Every locality, he explained, has designated “elite” schools that talented students from surrounding areas compete to transfer into. The result is a system where “educational resources are systematically sucked up to the center from the periphery, leaving rural areas incredibly depleted.” While this arguably gives academically gifted students opportunities to develop their talents, it deprives most students of educational resources.

According to China’s 2020 census, only 30.6% of the population has ever attended high school (including non-academic vocational secondary school), which Stanford professor Scott Rozelle notes, “is lower than South Africa, lower than Turkey and lower than Mexico.” In 2022, roughly 40% of China’s middle school graduates didn’t go on to attend high school of any kind, and among the students that do continue their education, national policy stipulates that roughly half (“五五分流”) are funneled into non-academic vocational high schools with no path to enter college.

To understand how AI could fit into this picture, we first need to understand the political and economic factors that incentivize Beijing to care about students in the countryside. It’s not clear that more investment in education will translate to high economic growth at this point in China’s development path — the real youth unemployment rate is probably still around 20%, and there are fewer entry-level positions available just as a record number of new graduates enter the workforce. Rather, this is a priority for the Party because improving the education system is so popular.

When Rozelle’s team surveyed 1,800 rural mothers and asked what they wanted their children to aspire to, over 95% said, “I want my child to go to college.” In China, a degree from an elite college doesn’t just translate to higher earnings — it unlocks better healthcare via the hukou system, cushy “iron rice bowl” 铁饭碗 jobs, and above all, social prestige. In 2023, researchers at Stanford found that Chinese families spent an average of 17.1% of their annual household income on education, which amounts to 7.9% of annual household expenditures. (Households in the US and Japan, by comparison, dedicate just 1-2% of annual expenditures to education.) The poorest quartile of families in China devotes a staggering 56.8% of income to education, and education spending is inelastic — that is, it’s prioritized as a necessary expense — across all income levels.

As Andrew Kipnis, the anthropologist who wrote Governing Educational Desire, explained to ChinaTalk, educational reform is a priority for the party “because it’s a way of keeping people happy. If they think there’s some hope their child will attend university, that gives them some investment in the system.” But not every child can become part of the elite: “People who have gone to university won’t work in factories,” as Kipnis put it. No matter how popular it would be, Beijing is not interested in building a system where a college education is available to anyone who wants one. But within this zero-sum system, where anyone who receives an advantage is inherently disadvantaging someone else, the party still needs to make parents feel like their child is getting ahead. Infrastructure is pretty much the perfect tool for this. It makes schools feel luxurious on the ground without changing the fundamentals that make the system so unfair. Shiny new facilities deliver popularity gains immediately, and if your child doesn’t get into university years later, it’s their own damn fault.

Those incentives are shaping the world’s largest AI education experiment. China is not the only country betting that AI will transform education, but the scale and style of China’s ambitions are unmatched globally. While China started with pilot programs, South Korea’s government led with inflexible national-level implementation, spending US$850 million on an ambitious AI textbook initiative that collapsed after just 4 months. India’s edtech ecosystem is private-sector-led with little top-down guidance or regulation, which resulted in the high-profile implosion of Byju’s and a proliferation of predatory practices targeting low-income families. Japan, unlike China, pledged to make sure every student had a device before implementing AI teaching tools.

Ultimately, China stands out globally for the sheer scale of its AI education ambitions — and the scope of applications its edtech industry is targeting for AI integration.

by Lily Ottinger, China Talks |  Read more:

Image: via

[ed. See also: Massive budget cuts for US science proposed again by Trump administration (Nature). National Science Foundation.]

Dating Apps: Giving Men What They Want But Not What They Need

Dating apps were built on the bones of Grindr. I have been known to joke that everything wrong with dating apps is divine retribution for culturally appropriating them from the gays.

Gay men, specifically, that’s important - the overwhelming majority of people making apps are still men, and most of those are still straight men, and while I don’t exactly have insider knowledge on this, it couldn’t be clearer to me that some open-ish minded straight tech boy heard from one of his gay male friends about being able to summon sex partners to his bed from the immediate vicinity after filtering on a bunch of lewd photos and thought: “There isn’t a straight man alive who wouldn’t consider giving up his left hand to have this experience with women. I could make a billion dollars making straight Grindr.”

And thus Tinder was born. Blah blah blah lust and greed sullying the purity of romantic and sexual love; a direction I could go, but instead we’re going to talk about the ways that playing to male preferences in the short term can easily ruin their entire lives, even when it was men’s idea.

Dating apps aggressively reflect male preferences, sexuality neutral. They’re long on photos, short on text. They filter primarily on location, which has some usefulness, but is most useful if the question is “who’s geographically close enough to me that walking to my place for sex is a realistic option” .

Men love flipping through photos of people they’re attracted to - that alone drove much of the traffic to Facebook’s precursor, Hot or Not. This app is built to give men a sexual scrolling experience as soothingly magnetic as any social media site while providing enough mystery to feel less degenerate than porn (the better for large doses and intermittent rewards).

For women, it’s grim. Yes, they get matches much more often than men do (largely because these extremely male-centric UI decisions lure vastly more male users than women; what economist could have predicted this problem with a heterosexual dating app). They don’t enjoy using these apps, not nearly to the degree or as often as men do. For most women, sifting through men feels dehumanizing, and sorting on pictures feels painfully limited (the male equivalent might be having to swipe based on photos of a woman’s favorite outfit, laid out on her bed. Vaguely boring and frustrating to have to make important decisions with so little information about the things you care about).

This isn’t just because of blackpill stuff about how men aren’t hot to women - that topic has been covered to death, yes women find men physically hot but no it doesn’t always work in such a way that static photos capture, so men are impossibly screwed by efforts to appeal to women with photos alone. There’s also the fact that men suck at taking pictures, because the market for photos of people is overwhelmingly men as buyers and women as suppliers, with the demand being for sexually attractive photos of women. Looking at photos of men is like driving a Nissan truck: it couldn’t be clearer that it is not your specialty and significantly worse than other products that your entire factory line was designed for.

You might think that dating apps are bad for men because they lead to men experiencing significant rejection - even the way my post is framed up until this point sort of implies as much. That framework, like much about dating apps, gets the whole picture subtly, insidiously wrong in a way that leaves people who take them at face value much worse off. You know who takes things at face value most often? You’re not going to believe this,

No, the greatest deprivation created by dating apps is specifically denying women and men the opportunity for women to keep men around in a general capacity. (If this idea makes you freak out about the friend zone, I’m almost impressed with you because young people seem to do so little socializing that no one complains about the friend zone anymore. Pat yourself on the back for having friends if you’ve managed to develop a resentment complex around the friend zone).

Most women develop attraction to men via proximity and time. Force a woman to choose if she wants the option to sleep with a man the second she meets him, and she will default to no in almost every single case. For many men, this means that any men who enjoy the attention of women who are open to sleeping with them at first glance are the only men women authentically want. Respectfully, you’re thinking like a guy, and if you believe that men and women are extremely different, I’m going to need you to trust that women develop affection for men differently than men do for women, such that you’ll ruin your life trying to figure out why women don’t desire you in the exact same way that you desire them...

One of the worst things you can do if you date women is to push them into a choice of yes or no as early as possible. You are simply too much of a risk on too many axes to get something other than a no unless you look like Chris Hemsworth, and even that wouldn’t get you yeses from 100% of the women you might ask out (hot men can still be shitty in about a thousand ways, and women often aren’t willing to take risks even for hotness. Again. They are not men). You might think that your goal should be to look like Chris Hemsworth, or alternatively to despair that you don’t look like Chris Hemsworth and go sulkily into that good night, but that’s you thinking like a guy and assuming that how women feel has to match how you feel. Frankly, that’s what got you into this mess: by trusting tech men who told you that you could game heterosexual dating by giving you an interface that pinged all your dopamine sensors while curiously robbing you of a lot of opportunities to find and develop a fulfilling relationship. [...]

The major product provided by a dating app is the illusion of participating in dating at all - some time swiping through faces, and congratulations, you are “dating”, you Tried, you do not need to do anything scarier or riskier or less fun than this.

by Eurydice, Eurydice Lives |  Read more:

Image: uncredited via

Chain Reaction

Major disruptions to maritime chokepoints always send ripples through the entire global network. Like voltage through an electrical grid, maritime commerce will shift to the path of least resistance, with nations forced to redistribute security assets accordingly. The current conflict with Iran is testing this concept in real time — and US government planners need to be paying attention, both for short-term and long-term planning.

 

Maritime canals, straits, and capes are not independent waterways with unchanging risk profiles. They are, in fact, interconnected points in a system within the global maritime network on which international commerce relies. Disruption in one location redistributes traffic worldwide, altering shipping costs, delivery timelines, and global capacity.

In more extreme circumstances, this creates risk calculations that are very different from the steady state. This also creates dilemmas for governments regarding required force distributions to either maintain a specific chokepoint or protect national shipping interests.

Firms are already considering the impacts of conflict adjacent to the Strait of Hormuz, through which 20 percent of the world’s petroleum sails. A near-term return to the status quo may produce short-term uncertainty in shipping, with only marginal impacts on global commerce. However, the term “marginal” can still include billions of dollars, as shown by the March 2021 Ever Given grounding in the Suez. Conversely, a replay of the 1984-1988 Tanker Wars, or complete closure of the waterway, paints a very different picture.

In light of current events, leaders need to understand how the network has changed and be prepared to act accordingly. To define chokepoint risk, consider six factors: 

• Navigation (are there underwater obstructions or dangerous currents)
• Environmental (high waves or storms)
• Geopolitical (are adjacent nations stable and in control)
• Criminal (is there a high incidence of piracy)
• Density (is this a high traffic waterway)
• Economic activity (is this a regional or global economic corridor)

From a risk perspective, navigation and environmental risks are largely static and rarely affected by activity at other locations. Geopolitical and criminal (piracy) risks tend to drift toward adjacent chokepoints, while traffic density and economic activity quickly ripple throughout the entire system.

Looking closest to the closure, Bab al-Mandeb, located near Yemen, may expect an increase in lawlessness. Already poorly governed, regional frictions are sure to spread to Yemen, and those transiting the lower Red Sea should expect increased threats from both piracy and warring factions. There may be a decrease in traffic as vessels no longer need to move between the Mediterranean and the Persian Gulf, but the Red Sea remains an important corridor for commercial and military traffic between Western nations and the Middle East. Consequently, while the current fight is through Hormuz, it is necessary to maintain maritime patrols in the vicinity of the Bab al-Mandeb and the Gulf of Aden. [...]

Where policy makers must firmly fix their gaze is the Turkish Straits. Ongoing conflict in the Black Sea Region (BSR) has already increased the risk between the Black and Aegean seas. Globally recognized as a wheat corridor between Ukrainian fields and the world, it is less appreciated that the BSR provides over 3 percent of global petroleum reserves, a figure likely to skyrocket in volume and value amid major market shifts. Expect to see a 5 percent increase in maritime traffic transiting through Istanbul.

Since the 1936 Montreux Convention, Turkey, whose military nearly doubles that of any European NATO ally, has deftly managed its control of the straits to stabilize geopolitical tensions in the region. But the heightened importance of the strait underscores the need for other interested parties to work together to support continued commerce. Bulgarian and Romanian forces, exempt from Montreux restrictions, must work with their Turkish NATO allies to monitor Black Sea activity, while the US 6th Fleet maintains vigilance in the Eastern Mediterranean and Aegean Seas.

The heightened importance of the waterway makes it an attractive target for malign actors, although Russia is less of a threat since much of the oil is theirs. Ankara is aware of these risks and should consider hosting regional exercises with NATO allies as a show of unified deterrence.

by Michael Kidd, Breaking Defense |  Read more:

Image: Mass Communication Specialist 1st Class Indra Beaufort
[ed. See also: Golden Dome $1.5T Defense Budget Request (BD); and, How the National Security Strategy Gets Made (Statecraft).]

Fuck's Fucking Fucked

“Tuesday will be Power Plant Day, and Bridge Day, all wrapped up in one, in Iran,” he wrote a little after 8 a.m. “Open the Fuckin’ Strait, you crazy bastards, or you’ll be living in Hell — JUST WATCH. Praise be to Allah.”  ~ Donald Trump, 4/5/2026

Fuck had run its course. For the beloved curse word, the future did not bode well. Gross overuse of fuck had continued to degrade TV and film, music and prose. Rap, the musical genre in which fuck had become most unironically parodic was a fat, ripe, slow-moving target and was the first to be hit. Pop a cap in your bottom? No, not that dire. The minor profanities seemed fine. Even the nastier ones looked to be okay. Only the fucks were fading. What were fans saying? 

Soon to follow their fuck-dependent rap brethren, standup comics were next to feel these winds of change, the gentle breeze starting to gust. Most starkly and notably, the fuck that is delivered by the comic to get the customary post-punchline second laugh, a laugh as good as guaranteed, no longer seemed to be working.

You know the bit, we all do.

The comic delivers his punchline, which is, let’s say, “The guy has no clue.” As the audience laughs at the punchline, the comic walks the few steps to his stool for a sip of bottled water. Lifting the bottle to his lips as the laughter generated by the “no clue” punchline begins to wane, the comic aborts his sip and says, “No fucking clue.”

According to a longstanding formula, the repeating of the punchline here, reinforced by the strut that is fuck, should be an easy second laugh. Not anymore. Such was indicative of the diminishing power of fuck, an early warning sign, the not-funny-a-second-time-unless-supported-by-fuck punchline no longer getting those easy second laughs. The comic may hear more chortle than laugh, a laugh of manners, a laugh forced not earned, or, there may be no laugh at all. Among the comics, fears of fuck failing in its role as reinforcer is why deliveries of the second, fuck-dependent punchline would usually occur during the sip of water, for should you find yourself sipping in silence at least you have something to do.

This was the moment in entertainment when audiences were letting it be known that they weren’t finding the word funny or shocking or dramatic anymore. To the artist, the audience was saying, “We need more than your cursing. We don’t find it impactful. We’re not twelve years old. It’s kind of insulting.”

For comedians, the message was clear enough. They abandoned the formula that is the fuck-supported second laugh—but it didn’t stop there. Even the fuck-supported first laugh, the fuck-supported laugh in general, was losing traction, losing its cultural standing, the comics coming to fear that even to use fuck, let alone overuse it, had become cliché. The “fuck comics,” the comics who continued to use fuck, soon became less appealing to audiences and then unappealing, the least funny of the comics. There was no question by this time, with rappers and comedians blatantly beginning to tidy up their vocabularies, that the demise of fuck was upon us. The people were making it clear: they were not just tiring of the word, they were telling artists that they needed to do more than lean on the creative crutch that is fuck. For fuck, the writing was on the wall. Amongst themselves, the fucks were talking. 

“Now what the fuck are we supposed to do?”

“Fucked if I know.”

All fucks were nervous, all were concerned. The fucks knew that a cultural shift of this magnitude would result, more or less, in their immediate extinction. There was confusion and fear. Lots of questions. The union would be no help on this one. There were jobs, families. The future.

“Fuck. Fuuuuuck. Bro, this fucking sucks.”

“And like, zero fucking warning, bro.”

“And that German fuck. He didn’t fucking help. That German fucker fucking fucked us.”

“Not only him, fucking him and a whole fucking movement.”

The “German fuck” was Dr. Kalba Brenin, the German linguist and film critic. Dr. Brenin engaged the fuck catastrophe innocently enough. He had commented on his podcast that a limited series he had been watching and was intending to review used fuck so often that he stopped watching after two of eight episodes. A highly touted series about sexy young corporate lawyers dispatched to the world’s largest cities to ply their trade and defend and sustain capitalism, Dr. Brenin became frustrated by the “near constant” use of fuck that would commence in any scene that “required dramatic acting.” Rolling his eyes and ultimately laughing at fuck-choked so-called dramatic scene after fuck-choked so-called dramatic scene—the excessive fucks turning the drama into unintended comedy—instead of a review Dr. Brenin wrote his now infamous essay, A Welcome Overstayed, in which he called for fuck to be banned from all recorded entertainment—not for reasons of censorship of profanity—but “for the sake of preserving what human beings have for five millennia called art.”

In his essay, Dr. Brenin transcribes an exchange from the show.

“Becca should be told this case is now a fucking homicide.”

“Becca’s in the Andes, mountain climbing, off the grid, how the fuck am I supposed to get in touch with her?”

“Well she’s fucking president of this firm, you fucking better find a way.”

“Fuck you, Tristan.”

“Fuck you, too, Emma. And fuck Becca.”

Exeunt Tristan.

Dr. Brenin was relentless in his criticism.

“Fuck is anti-art. Fuck is an art killer. Writers, as of now, you must remove fuck from your lexicon. Distance yourself. Save yourself while there’s still time. It’s over.” [...]

For the linguists and pop culture scholars who attempted to explain the decline of fuck, the similarities to tattoos were often cited, that the way in which tattoos had lost their edge, their cred, their cool, should have been seen by fuck as a cautionary tale.

Tattoos had gone from ships-at-sea and prisons and cheap boarding rooms to spas and moms and pretty junior bank tellers with full sleeves. The tattoo was defanged. The tattoo had been corporatized, commercialized and, of special concern to men, feminized, as tattoos had become, near the final stages of the tattoo era and the rise of the tattoo removal era, more popular with women than men. Once the lone province of the hairy forearm, the tattoo had spread all over the body, a metastatic migration that could only result in homogenization. Fuck was on a similar trajectory. The word had become homogenous. It was as if once the film and television industries were finally permitted to say fuck, after years of censorship, fuck was all they wanted to say and now, after decades of relentless and unforgiving fuckery the people were tired, they had heard enough, they didn’t want to hear it anymore.

by Brutus Macdonald, Substack | Read more:
Image: George Carlin via: "the seven words you can't say on tv".

A Ride Through the Ages

via:

Isabelle Ferreira / Par La Nuit / Photography / 2025

via:

Two for One

Two Planes Destroyed by U.S. During Rescue Operation (WSJ)

Two U.S. Special Operations aircraft were blown up on the ground during the rescue mission of the second American aviator in Iran, a person familiar with the mission said.

MC-130Js are specially equipped planes that are used to carry out covert infiltrations and to remove troops from beyond enemy lines. The official didn't explain how the aircraft appeared to have gotten stuck during the course of the rescue operation, but said that it became necessary to destroy them.

Image: Ricardo Arduengo/Reuters

[ed. So much winning.]

You Work For the Bad Boss You Have, Not the Good Boss You Wish You Had

In every job, there is some gap between the advertisement and the reality. The outdoor job boasting “fresh air every day” consists of picking up trash. The service job “perfect for sunny personalities” consists of getting yelled at by angry customers. The day care job that offers “unlimited cuteness” consists of cleaning up poop. This is how it goes. We must all endure some amount of hastily concealed tribulations in order to pay the bills.

Likewise, we must persevere through bad bosses. We must learn to navigate pernicious supervisors, backstabbing managers, and incompetent executives as the price of gainful employment. Good bosses are a stroke of luck, rather than a baseline expectation. If we all limited ourselves to workplaces that were free of bad bosses, it would take ten years to find a job.

Mostly, this is to be expected—an inducement to aspire to a promotion, or to start a union, rather than a catastrophe. But there are some jobs where a bad boss is a bigger deal. There are some jobs where a bad boss can very quickly get you into a genuine moral crisis. If you have a job like that, shrugging off what the bad boss is doing can become not an act of resilience, but one of gross negligence.

The military is one job of this type. There is more moral urgency attached to the military’s conduct of its affairs than to, you know, a restaurant’s conduct of its affairs, due to the fact that the military kills people. There are higher stakes to poor management decisions. If you are a line cook and your boss tells you to cook a dish improperly and you acquiesce for the sake of keeping your job, you can be forgiven. If you are a member of the military and your boss tells you to kill innocent people or bomb their homes or snatch their freedom and you acquiesce for the sake of keeping your job, forgiveness is not so certain. You become not a beleaguered employee, but a true villain. The space that the world is able to afford you as a matter of sympathy for your workplace annoyances shrinks down to almost nothing once guns are involved.

People join the military for all sorts of reasons: For economic opportunity, for adventure, for patriotism, for sheer lack of options. Most soldiers, it is safe to say, believe they are doing something good. Even those who are not ultra-patriots probably believe—and are told, by ads and by supervisors and by TV and by politicians and by the public—that their jobs are, on balance, honorable ones. They do something difficult, and they believe they do something necessary, and they take a certain amount of pride in that, as anyone would.

The school in Iran where we blew up kids.

But the military is a gun in the hand of the Commander in Chief and we have a Commander in Chief who is dumb, narcissistic, unpredictable, and dangerous. The bad boss problem, for soldiers, is everything. It is the difference between being honorable and being the violent foot soldier of a thug. Which situation is closer to reality now, do you think? Being a soldier is not inherently righteous. That is a fairy tale they tell teenagers in order to get them to join the military. The righteousness of an army is wholly dependent on the righteousness of the cause that the army fights for. (Teenagers learn this, too, about other armies in other nations. We are careful never to tell them to apply the principle to the United States itself.) [...]

Here in America, when we are talking about American soldiers, we typically say they are honorable public servants and dismiss any blame for the havoc they wreak. Whereas if we are speaking about other soldiers in other nations, we expect and call for them to be killed by our own soldiers because they are carrying out equivalent duties. I hope I do not have to point out the ethical schizophrenia of this approach.

My purpose is not to demonize members of the military. On the contrary. People who joined an organization with noble intentions, who were told that they were serving the purest interests of their country, are now in the position of being foot soldiers for a gangster-style president who is quite possibly the single biggest threat to peace on earth. It is important that we speak honestly about the fact that these soldiers are in the perilous position of risking their lives in order to carry out villainous goals. That would be a tragedy not only for the victims of American imperial overreach, but also for the American soldiers themselves, who will be cursed to live their lives with the knowledge of what they have done. You may have joined the organization imagining what good it could do with a good boss. But that is not the world you have actually entered. In this world, the world that exists, you are an armed member of a deadly organization run by a bad boss. He has done and will continue to do bad things. And who will have to carry out the bloody acts inherent in those bad things? You will. It’s a bad deal. While you may have come to find yourself in this position through a series of well-intentioned actions, the fact is that the only ethical thing to do is to do your utmost to remove yourself from a job that might ask you to kill, unethically, on behalf of a bastard.

The military is not the only sort of job in this same position today. Many well-intentioned people who went to work in, say, the State Department, or the CDC, or other branches of government may now be faced with a similar moral dilemma.

by Hamilton Nolan, How Things Work |  Read more:

Images: Getty

[ed. Should be fun telling the grandkids what you did in your career. See also: Digging up the Dead (LRB):]

***

More than a million people have been displaced by Israel’s invasion and bombardment of Lebanon, many fleeing with nothing more than the clothes on their back, camping in the hills or sheltering in schools or municipal buildings. The desire to turn Lebanon into another Gaza, articulated by Israeli spokesmen, is being fulfilled with attacks on journalists, the use of undercover operatives and the bombing of displaced families huddled in makeshift shelters. Familiar too was the timing of the attacks during Ramadan, frequently at iftar when people were about to break their fast.

The attack on the small town of Nabi Chit in the eastern Bekaa Valley on 6 March shows the value placed on human lives by the regime in Israel and its backers in the United States. According to the Israeli government, the invasion was a rescue operation to retrieve the remains of an Israeli airman who disappeared forty years ago. Residents of Nabi Chit and the Lebanese army chief told the BBC that Israeli special forces entered the town ‘disguised in Lebanese military fatigues and used ambulances with signs of Hizbullah’s Islamic Health Organisation’. They headed to the corner of the graveyard, dug it up but found nothing there. The town fought back, causing the Israeli soldiers to withdraw. To cover their retreat, Israel carried out more than forty airstrikes in five hours, killing 41 people. [...]

According to Euro-Med Human Rights Monitor, 58 of Gaza’s 62 graveyards have been damaged or destroyed:

the Israeli army carried out a focused operation at al-Batsh Cemetery, east of Gaza City, in January 2026. The cemetery was converted into a military barracks, and more than 700 bodies were exhumed under the pretext of searching for the body of an Israeli detainee. The army later withdrew after extensive bulldozing that radically altered the cemetery’s landscape, preventing families from locating their relatives’ graves.

This was not an isolated event:

in many cases, the Israeli army deliberately exhumed graves and converted cemeteries into military barracks under the pretext of searching for the bodies of Israeli detainees. These actions were carried out without documented, verifiable procedures, independent oversight, or a clear chain of custody and handover process. Israeli forces removed hundreds of bodies from their burial sites, mixed remains, failed to return them to their original locations, and provided no identifying or biological data to enable verification or documentation, making the recovery and identification of remains extremely difficult. [...]

The scope of this policy has been expanded in recent years and upheld by the Israeli Supreme Court. Amira Haas reported in Haaretz last month that Israel was holding the bodies of 776 Palestinians. There are 256 buried in nameless, numbered graves while the rest are held at military morgues. Nearly half were killed since October 2023 and 88 died in Israeli detention. These bodies are sometimes used as bargaining chips in negotiations, although in more recent exchanges many of the bodies have been unrecognisable when returned. There is at least one case of the wrong (long awaited) body being returned to the family.

The grieving are not allowed to grieve, the reunited to celebrate. The repeated disruption of Palestinian funerals and prisoner releases has been bolstered by military orders that proscribe and punish expressions of Palestinian emotion. In February 2025, during a prisoner exchange, the psychotherapist Gwyn Daniels quoted Edward Said, who

chose the word ‘inert’ not descriptively but to conjure up the Zionist fantasy about the ideal Palestinian body. Perhaps for the coloniser, this ideal body should be lifeless or ‘disappeared’. But given the stubborn persistence of Palestinians to remain living on their land, they should preferably cause as little disturbance to the colonisers as possible. Along with no displays of cultural identity, there must be no passion, no pride, no joy, no sorrow, no anger – indeed, no demonstrable emotions that might trouble their oppressors.

Go Ahead and Use AI. It Will Only Help Me Dominate You.

Recently there has been a lot of commentary of the following type:

BAD WRITER [touchily]: “Actually, I do use AI to help me write.”

Okay. That checks out. Carry on.

Want to use AI as a Valuable Part of Your Writing Process? Want to use it to “generate pushback on my column thesis” and be “more comprehensible” and “craft unique angles” and offer “positive and negative feedback” and “scale the quantity” of your “output?”

Knock yourself out.

You have my blessing.

Hey buddy— go for it!

Some in the “real writer” community find this sort of rampant outsourcing of the writing process to AI to be distressing. Not me. Would I do it myself? No. I have self-respect. But I want to tell you, my friends, that you have my full support for all of it. Want to throw your dashed-off notes into ChatGPT and have it spit a draft back at you and then edit that and call it your own? Want to toss a few hastily written headlines at Claude and have it generate the outline of your piece? Want to dump your entire career archives into a chatbot and then order it to replicate your own voice so you don’t have to?

Do you, a grown man, a successful professional writer who has received a book deal paying you real US currency, want to use AI for the purpose of “making sure the book matches [your own] writing style”[???]? Guess what, brother: I support you. I affirm you. I am right here offering you a classic thumbs-up gesture of affirmation.

“Whoa, a writer who I have never regarded as particularly inventive is using AI? I am surprised and disappointed.” There’s a sentence I would never utter. Instead, I would accept the news of your AI use with total equanimity, nodding almost imperceptibly to indicate that this is not something worth raising my eyebrows over.

No, I will not be joining in the chorus of condemnation. On the contrary. If you are a professional writer, I want you to use AI. Because this industry is competitive. I’ll take any advantage I can get. And if you want to make your writing suck, that’s all the better for me. One less person outshining me.

The tepid, conformist nature of your AI-assisted prose will only make my unexpected bons mots stand out more sharply. While you lean on a technological crutch of grammatical mediocrity to drag your essays over the finish line, I’ll be metaphorically zipping past you on my “magic carpet” of words emerging directly from my own declining and unpredictable brain. Over time, the intellectual box into which AI has seduced your creative process will suffocate you, leaving your bereft readers little choice but to drift into my subscription base.

You’ll be all, “Politics in America is divided—but it doesn’t have to be. Let’s discuss how to bridge the partisan divide.” Your sense of joy at the possibilities of the English language will have been so eroded that you won’t even understand why that sucks shit. Meanwhile I’ll be dropping some wild similes you could never even imagine. “Politics is like a sea slug.” What?? How?? Readers will flock to me to find out. Too bad your AI editor struck that line from your piece as “indecipherable.”

You and your friend “Claude” wouldn’t last two seconds in my cipher.

Maybe you read the studies about how AI use causes “cognitive surrender” that slowly destroys your ability to think critically about the linguistic cud that the machine is serving you. Or about how it causes “cognitive foreclosure” that prevents you from ever developing the skills to critique AI output even if you wanted to. Maybe these studies give you pause, when you think about introducing these inscrutable tools of mental paralysis into your own creative process.

Don’t worry about it!

Life is hard enough already. You’re busy. You have lots of things to do—laundry, making lunch, and more. The last thing you need is a bunch of jealous (Brooklyn hipster) writers lecturing you about how this magical productivity booster is somehow “bad” for you. Those are probably the same haters who told you to stop doing so much crystal meth. Some people can’t stand to see you succeed!

I just checked a calendar—it’s 2026. AI is here to stay and you might as well beat the rush by using it more and more, right? Right. In the name of efficiency, it just makes sense for you to turn over ever greater portions of your thought process to this seductive helper, never stopping to ask yourself what it is costing you. You are a nice person and your job (writing) deserves to be easy. There, there. Allow yourself to sink into the warm opiate of cerebral ease. This is better. Yes. This is much better.

By all means—proceed.

And then, when you have settled into this comfortable pattern, sit back and watch me unsheath my massive, work-hardened intellect, built to staggering strength through a daily regimen of thinking about stuff. I think you’ll find that your panicked efforts to resist my onslaught will prove unsuccessful, hampered as you are by atrophied muscles of the mind. Ask your AI companion for some final words of comfort. The hour of your doom draws near.

I will crush you with ease.

by Hamilton Nolan, How Things Work |  Read more:

Image: Getty

[ed. Haha...yep. : ) See also: Who Goes AI? (with respect to Dorothy Thompson's 'Who Goes Nazi', gracefully acknowledged by the author).]

The Big T-Shirt Payoff

The College Student—and His Cat Meme—Who Hunted the World’s Biggest Cyberweapon

Sitting in his dorm room at the Rochester Institute of Technology, Benjamin Brundage was closing in on a mystery that had even seasoned internet investigators baffled. A cat meme helped him crack the case.

A growing network of hacked devices was launching the biggest cyberattacks ever seen on the internet. It had become the most powerful cyberweapon ever assembled, large enough to knock a state or even a small country offline. Investigators didn’t know exactly who had built it—or how.

 

Brundage had been following the attacks, too—and, in between classes, was conducting his own investigation. In September, the college senior started messaging online with an anonymous user who seemed to have insider knowledge.

As they chatted on Discord, a platform favored by videogamers, Brundage was eager to get more information, but he didn’t want to come off as too serious and shut down the conversation. So every now and then he’d send a funny GIF to lighten the mood. Brundage was fluent in the memes, jokes and technical jargon popular with young gamers and hackers who are extremely online.

“It was a bit of just asking over and over again and then like being a bit unserious,” said Brundage.

At one point, he asked for some technical details. He followed up with the cat meme: a six-second clip that showed a hand adjusting a necktie on a fluffy gray cat.

Brundage didn’t expect it to work, but he got the information. “It took me by surprise,” he said.

Eventually the leaker hinted there was a new vulnerability on the internet. Brundage, who is 22, would learn it threatened tens of millions of consumers and as much as a quarter of the world’s corporations. As he unraveled the mystery, he impressed veteran researchers with his findings—including federal law enforcement, which took action against the network two weeks ago.

Chad Seaman, a researcher at Akamai, joked at one point that the internet could go down if Brundage spent too much time on his exams.

Early warning

Three times a year, several hundred of the techies who keep North America’s internet running gather to talk shop. Last June they met at a conference in Denver hosted by the North American Network Operators’ Group.

One major topic was a fast-growing and often legally dubious business known as residential proxy networks. Dozens of companies around the world run such networks, which are made up of consumer devices like phones, computers and video players.

These “res proxy” companies rent out access to internet connections on the devices to customers who want to look like they’re surfing the internet from a genuine home address.

That kind of access is useful for people who want privacy or for companies that want to masquerade as regular people to test out internet features for particular regions or scrape the web for data (say, a shopping price-comparison site). AI companies use the networks to get around blocks on automated traffic so they can gather large amounts of data to train their models.

Then there are the customers who want to hide their identity while engaging in ticket scalping, bank fraud, bomb threats, stalking, child exploitation, hacking or espionage.

Some device owners willingly sign up to be on these networks so they can make a few dollars a month, but most have no idea they’re connected to one.

At the Denver conference, Craig Labovitz was alarmed. The Nokia executive had been tracking the data flows of the internet’s infrastructure for years, and he knew the network’s data centers, chokepoints and design better than most.

Starting in January 2025, Nokia’s sensors had picked up a series of increasingly powerful cyberattacks coming from devices that hadn’t previously been considered dangerous. Called distributed denial of service, or DDoS, attacks, these were massive floods of junk internet data designed to knock websites offline by overwhelming the data pipes that connected them. These attacks are sometimes launched by extortionists or even business rivals seeking to sabotage computer networks.

Nokia saw hundreds of thousands of devices joining in these attacks. One unprecedented attack later in the year on internet service provider Cloudflare was “comparable to the combined populations of the UK, Germany, and Spain all simultaneously typing a website address and then hitting ‘enter’ at the same second,” Cloudflare said.

The network, which would become known as Kimwolf, seemed to be using residential proxy connections to launch its attacks, giving it the potential to do massive damage.

“The basic message was, ‘Be afraid,’” Labovitz remembers. [...]

Instead he applied his hacking skills toward legitimate cybersecurity research. In his senior year of high school, he found bugs in websites belonging to the Dutch government and reported them via a “bug bounty” program that offered hackers prizes for unearthing security flaws.A few months later, the Dutch National Cyber Security Center mailed him his bounty: a black T-shirt. It read: “I hacked the Dutch government and all I got was this lousy t-shirt.”

He remembers it as one of the most rewarding experiences of his young life: a “dopamine rush,” he said. [...]

On March 19, federal authorities announced they’d disrupted four of the world’s largest DDoS botnets, including Kimwolf. Kimwolf had launched more than 26,000 DDoS attacks targeting over 8,000 victims, according to a court filing. The press release announcing the takedown thanked Brundage’s company, Synthient, among others.

​Industry experts say that Kimwolf today is a shadow of its former self. The cybersecurity firm Netscout says it’s seeing about 30,000 Kimwolf machines active at any given time.

Brundage recently got a text message from a federal official on the case. The official had heard about the bug bounty Brundage got from the Dutch government years ago and had a question: “What’s a good address to mail you a t-shirt, and what’s your size?”

by Robert McMillan, Wall Street Journal |  Read more:

Image: via

[ed. Here's how to protect yourself.]