Suchir Balaji

https://archive.is/ucBx0 https://www.thenation.com/article/society/open-ai-suchir-balaji-whistleblowers/ On October 23, 2024, Suchir Balaji’s face appeared, freshly shaved, boyish but serious, partially shadowed, in The New York Times as he announced himself as a whistleblower against one of the most powerful technology companies in the world. Setting off a political earthquake in the AI industry, Balaji claimed that OpenAI, the highly touted artificial-intelligence start-up where he’d worked as a researcher for almost four years, had broken copyright laws by absorbing practically all available data on the Internet for its models. ChatGPT, the company’s mega-popular AI chatbot—as well as, potentially, its competitors—had apparently been built on a foundation of illegal activity on a mass scale. Balaji presented his supporting argument in a mathematically minded paper on his personal website. “If you believe what I believe, you have to just leave the company,” Balaji told the Times’ Cade Metz. OpenAI had a lot on the line, and so did Balaji. Other researchers had resigned from their positions and issued warnings about the dangers of AI going rogue in fantastical Terminator-like scenarios. But as the Times noted, Balaji was one of the few industry professionals flagging the damage that the technology was doing right now. A 26-year-old former top student at Berkeley, he was already a veteran of several AI labs and had a patent under his belt. His parents called him a humble prodigy who had taught himself programming at age 11. His future seemed boundless. Balaji didn’t live long enough to tell his full story. A month after the Times interview was published, Balaji was found dead in his San Francisco apartment with a gunshot wound to the head. In that brief interval, he had gone from obscure AI researcher to high-profile whistleblower under severe pressure as he defied an industry that was collectively engaged in the biggest speculative bet in American business history. Suchir Balaji’s time in the public eye amounted to one newspaper interview, but his afterlife as a martyr has become increasingly complex, his legacy contested terrain. The San Francisco medical examiner ruled his death a suicide, but his parents, Poornima Ramarao and Balaji Ramamurthy, have repeatedly stated that their son was murdered for what he knew. The dispute likely won’t be resolved to the satisfaction of Balaji’s parents or critics of the burgeoning AI industry anytime soon. Still, it tells us a great deal about the crisis of accountability in a sector of the AI industry that has come to dominate our investment economy and dramatically alter our daily lives. The aftermath of Balaji’s death also reveals a profound and troubling failure to protect and support AI whistleblowers, and corporate whistleblowers in general. Current Issue Cover of May 2026 Issue May 2026 Issue For the Times, Balaji’s story was more than an impressive scoop. It was also a key element in its copyright-infringement lawsuit against the massive AI start-up, New York Times v. OpenAI. If the paper were to succeed in the suit, in which Balaji was listed as a potential witness, it could win a multibillion-dollar judgment and open the door to numerous similar actions against OpenAI and its competitors. From the vantage point of an industry built on the ruthless consumption of any and all data, the potential damage might be apocalyptic. Balaji’s grieving parents have highlighted the murky circumstances surrounding his death, hiring numerous forensic consultants and lawyers in their quest to prove that their son was murdered. Some details remain disputed—Balaji’s parents claim that the crime scene was poorly tended, that there is evidence from blood and hair collected at the scene that indicates a possible struggle, and that he was shot at an angle that is inconsistent with suicide. Overall, the picture is mixed. Balaji left no suicide note. His apartment was dead-bolted from the inside, and police reported there was no sign of forced entry. The day before his death, after returning from a trip with friends, he had spoken with his father on the phone and seemed happy. Security footage recorded by his building in the hours before his death shows him walking into his apartment with a takeout order. He owned a gun. In some news reports, friends described him as somewhat secretive. The medical examiner’s report found GHB in his system. While the so-called date-rape drug can be used recreationally or as a disabling agent, GHB can also appear naturally during a body’s decomposition. The primary demand from Balaji’s parents has been for the FBI to look into their son’s death. “We know there was foul play from many factors, many data points,” Ramarao said at a public vigil in December 2024. Indicating that her son represented a threat to the AI industry, she asked for a “thorough investigation.” Balaji’s parents have found support in the right-wing-media ecosystem for their contention that their son was murdered. Making their case: Balaji’s parents have found support in the right-wing-media ecosystem for their contention that their son was murdered.(ABC7news.com) For a time, Balaji’s family found support across the political spectrum. In January 2025, after a news report stated that the San Francisco Police Department (SFPD) was still investigating Balaji’s death, the left-wing San Francisco Supervisor Jackie Fielder posted on X, “I am relieved to see this case reopened. Friends and family of Suchir are welcome to get in touch with my office.” At the same time, Democratic Congressman Ro Khanna addressed Ramarao directly on X, offering his help. He later told The Mercury News that he’d had a conversation with Ramarao that led him to “believe her that there are unanswered questions.” But in February, the SFPD closed its investigation into Balaji’s death, pronouncing it a suicide, and Democratic politicians seemed to lose interest in the case. Neither Khanna’s nor Fielder’s office responded to inquiries about their past support for a new investigation or their contact with Balaji’s family. Ramarao has been a particularly vocal advocate for her son and has aggressively challenged officials for not doing more to investigate his death. She has spoken frequently at public events and in lengthy interviews on podcasts, YouTube shows, and Indian media. In January 2025, when Bay Area public officials were calling for a deeper investigation into Balaji’s death, Ramarao sat down for a 66-minute interview with Tucker Carlson, during which she said that her son was murdered. Ramamurthy has offered fewer public remarks, though he is sometimes seen alongside his wife, appearing stoically subdued. At the December 2024 public memorial, he expressed concern that the United States was becoming a mafia state. “Enterprises are becoming very greedy, and compassion and caring have taken a backseat,” he said. Under the banner of the Justice for Suchir Movement, Ramamurthy has posted online about the need for “ethical AI,” and he’s started a conference, the Suchir Balaji Memorial Summit. In his X profile, he identifies himself as “OpenAI Whistleblower Suchir’s Dad,” and he’s made numerous posts with the hashtag #justiceforsuchir that point out alleged issues at the crime scene, such as the blood-spatter patterns, a disabled surveillance camera, and a window through which an assassin could have entered. His posts depict his son as a thwarted genius, the “robinhood of AI,” who called for a more humane AI revolution and was murdered for his efforts. In their desperation to find an alternate explanation for their son’s death, Ramarao and Ramamurthy have cycled through a half-dozen attorneys and several outside investigators. They filed and then dropped a lawsuit against the SFPD. Meanwhile, they’ve gone after OpenAI on many fronts, including supporting a ballot measure that opposes its planned conversion into a for-profit company. They’ve also found some surprising allies. Ramarao and Ramamurthy have received extensive support on the political right, which has occasionally asked the right questions, even if for the wrong reasons. But whatever genuine suspicions might surround the case, Balaji’s death has unfortunately become fodder in the political battles that are consuming the tech industry. OpenAI’s CEO, Sam Altman, and Elon Musk, who was one of the company’s cofounders, have become mortal enemies. The two are embroiled in a lawsuit based on Musk’s belief that OpenAI’s transformation into a for-profit structure will contribute to the untrammeled development of artificial general intelligence (AGI), which may doom humanity. In the meantime, Musk’s company xAI hosts Grok, an anything-goes chatbot that has helped enable a global proliferation of nonconsensual pedophilic imagery; it has also parroted Musk’s racist beliefs and even praised Hitler. Musk will apparently do anything to crush Altman, and that includes accusing him of orchestrating a murder. OpenAI CEO Sam Altman seemed evasive when Tucker Carlson called Balaji’s death “definitely murder.” Awkward moment: OpenAI CEO Sam Altman seemed evasive when Tucker Carlson called Balaji’s death “definitely murder.”(The Tucker Carlson Show) In September of last year, a squeamish-looking Altman gave an exceedingly awkward interview to Tucker Carlson. Carlson called Balaji’s death “definitely murder” as Altman sweated his way through some words of sympathy for his former employee’s family. The grim exchange, in which Altman seemed oily and evasive, lit up right-wing social networks. “He was murdered,” Musk declared on X. The public speculation surrounding Balaji’s death went into overdrive. In the absence of an authoritative investigation, the Balaji case has become the subject of feverish conjecture by right-wing podcasters, tech moguls, and social-media conspiracy theorists, who, unlike many on the liberal left, are willing to entertain sinister ideas about the tech industry and corporate power. While they eagerly serve as ideological foot soldiers in Musk’s online army, the virtual shots they fire at OpenAI and Altman reflect a nascent public disgust with tech—one that is clearly grounded in issues of class, even if it often takes the form of paranoid murder fantasies. Balaji’s parents have little choice but to embrace those who are willing to grant them attention and entertain their questions about their son’s death. They have repeatedly posted on X asking for help from Musk and MAGA politicians like Vivek Ramaswamy. When Carlson offered Ramarao access to a massive audience, she accepted. Ramarao declined to speak with me when I reached out earlier this year; Ramamurthy spoke with me briefly on the phone, then hung up and never called back. Later, Ramarao e-mailed me, writing, “When we are pushed to give evidence, it doesn’t seem safe to get engaged. We don’t know who is behind this. So please give us time and space. Unless introduced by someone we trust, we won’t engage.” Using the media to litigate the circumstances of Balaji’s death is inherently fraught. Leftists and liberals who are inclined to distrust tech moguls and the police might be attracted to the tale of a ruthless company that is willing to kill a critic. But among those Democrats who once offered the family public support, Balaji’s story seems to have faded from their concern, consigned to the realm of misinformation and right-wing conspiracism—when it’s really about an isolated whistleblower challenging corporate malfeasance at the highest levels. For the online right, the story of Balaji’s death offers all the proper villains, including a shifty gay tech CEO already locked in combat with their hero, Elon Musk, the world’s richest fascist. In their narrative, “Altman killed Balaji” has been memed into being, proven through the low-resolution photographs and heart-rending parental statements percolating through the right-wing-media ecosystem—all of it approved and boosted by Musk himself. There’s even a memecoin (a cryptocurrency token) named after Balaji; its maker claims to be supporting the family’s cause. With every opportunity the right gives Balaji’s family to press their case, the search for justice seems to slip deeper into the political swamp—into the confusion and uncertainty that then gets amplified by yet another investigator producing yet another report dissenting from the SFPD’s official line. “The only person in his life Suchir was ever not happy about was Sam Altman,” Ramarao told the MAGA-affiliated podcaster Patrick Bet-David, who has 2.89 million subscribers on YouTube. “So Suchir knew something.” “There’s an AI war going on,” Bet-David added, characterizing Balaji’s death as practically a gangland killing. Security footage of Balaji entering his apartment with takeout food before his death has fed speculation that he was murdered. Last meal: Security footage of Balaji entering his apartment with takeout food before his death has fed speculation that he was murdered.(Alchemy Apartments) With his death suffused in rumors, misinformation, and his parents’ righteous grief—all stirred in the pot of the world’s richest troll—Suchir Balaji has become the haunted face of tech whistleblowing and the AI boom. But his death reflects a wider dysfunction in American industry, where, as once-durable liberal institutions crumble, corporate and government power becomes unaccountable, synonymous with corruption and criminality. The risk goes beyond one company, industry, or complaint; Balaji’s story fits into the broader saga of whistleblowers who face daunting odds against proving their claims. Whistleblowing is a solitary, psychically dangerous pursuit, and more would-be truth-tellers have succumbed to the malign forces marshaled against them—gaslighting, intimidation, legal threats, financial pressure—than to a hit man’s bullet. On the morning of March 9, 2024, John Barnett, a whistleblower who had worked on the 787 Dreamliner aircraft at Boeing, was found dead of a gunshot wound in his car outside a hotel in Charleston, South Carolina. Barnett had spent the previous two days testifying against Boeing, alleging that the company had used substandard parts while building the 787 and ignored safety protocols, and he was scheduled for more testimony that day. His death immediately provoked widespread speculation that he’d been murdered. But the physical evidence, comments from his family, and Barnett’s lengthy suicide note indicate that he likely killed himself. Barnett spent the last night of his life in his truck, writing about his fury and desperation in a sprawling, chaotic letter that was found at his side. The text runs at all angles; a reader has to turn some pages around to make out the words. But they are clearly the final thoughts of a profoundly depressed man saying that he could no longer carry on but wanted his family and friends to know that he was at peace. “America, come together or die!!” Barnett wrote on one page. “I pray the motherfuckers that destroyed my life pay!!! I pray Boeing pays!!!” The same page contains a remark scrawled sideways between the lines about Boeing, and it says as much about Suchir Balaji’s ordeal and those of so many others as it does about Barnett’s. “The entire system for whistleblower protection is fucked up too!” Barnett wrote. Mary Inman, an attorney at the law firm Whistleblower Partners, has the same concern. She’s a cofounder of an initiative called Psst.org, which aims to encourage action by bringing together whistleblowers from the same company or institution. Despite the canonization of whistleblowers as lone heroes in movies, many corporate and government scandals involve more than one person sounding the alarm, with potentially different fates befalling them. Collaboration and mutual support can be crucial. “There’s a first-mover problem with whistleblowing, particularly in AI, where the stakes are so high,” Inman said. Psst.org offers an “information escrow” service, a kind of digital lockbox where someone can submit a complaint or memo about wrongdoing. Whistleblowers can use the service if they have sensitive information but don’t feel comfortable raising the issue on their own. If multiple people from the same institution submit information, the service can bring them together. The organization also offers other forms of support, such as media training and connections to therapists. “We want to disrupt and change how whistleblowing happens and make it a collective act,” Inman said, pointing again to OpenAI, where months before Balaji came forward, a group of researchers resigned together, criticizing the company’s unsafe practices. “And in so doing, we’ll be able to bring more people forward.” The AI industry is lightly regulated, with fewer safeguards to protect the public’s interests than more established fields. Backed by the tech industry, President Donald Trump’s administration has attempted—first via legislation, then through an executive order—to prohibit individual states from instituting their own AI regulations. Inside these companies, there are enormous financial incentives to say nothing, and many AI professionals subscribe to an austere, almost cultish faith in the world-changing power of artificial general intelligence. Sophie Luskin, a researcher at the Princeton Center for Information Technology Policy, said that the absence of any legislation that applies specifically to whistleblowers in the AI industry is a major problem. “With employees in the AI sector, they find their rights and protections are particularly unclear and confusing,” she said. Will they be protected, for example, if they report potential harms or ones that haven’t happened yet? What about something that might harm the public but doesn’t violate the law, such as the information related to Secretary of War Pete Hegseth’s pressure campaign against the AI company Anthropic for refusing to allow its software to be used for autonomous weapons and other lethal purposes? The bipartisan AI Whistleblower Protection Act, sponsored in May 2025 by Senators Chuck Grassley (R-IA) and Amy Klobuchar (D-MN), among others, was supposed to fill in these gaps, codifying both the kinds of harms that AI whistleblowers might report and the protections they should be afforded. The act “has very good bones that have strong precedent in other existing laws,” Luskin said. But it has been stuck in committee for nearly a year. “We really miss out on the ability to hold these companies accountable when we don’t have effective and comprehensive whistleblower protections,” Luskin said. “People who work at these companies know what’s going on.” Now, as Balaji’s story lapses into myth, there remains little accountability for his death—and for the potential crimes that he sacrificed so much to expose. OpenAI, which has expressed sympathy over Balaji’s death, remains the industry’s hottest start-up and is currently seeking up to $100 billion in investment capital. Lawsuits and whistleblower complaints seeking to publicize alleged abuses at the company—including one submitted to the SEC regarding alleged violations of securities laws—have done little to stem its rise. Perhaps this is the inevitable outcome of actions undertaken by isolated, principled individuals trying to slow the freight train of highly leveraged technological innovation. At the December 2024 public memorial for Balaji, when his mother called for an FBI investigation into her son’s death, Ramarao offered a spirited tribute. She explained to the crowd that her child had followed his sense of right and wrong to the only place it could take him: The most important thing is he was a witness. He was a witness for a critical case. And even that’s not taken into account, and that’s the most frustrating [thing] for us. What’s the value of speaking truth if by speaking truth someone could lose their life? Do we never encourage our children to stand up for truth? Is there any protection? So whistleblowers’ lives don’t matter? We’re ready to lose them. I don’t know how I could have saved my son by teaching him to tell lies. The ethics with which I raised him took his life today. Ramarao’s words reflect more than the loss suffered by a grieving parent. As AI rampages through so many sectors of American life and continues to disfigure the basic foundations of creative and intellectual labor, the issues raised by Suchir Balaji are more urgent than ever. But absent any real political action against unchecked corporate power, Balaji’s story is more likely to be remembered as a right-wing meme than as a catalyst for reform. Jacob Silverman Jacob Silverman is the author most recently of Gilded Rage: Elon Musk and the Radicalization of Silicon Valley. He is also the host of Understood: The Making of Musk, a limited podcast series from CBC.

The Philosopher and Ethics of AI Developers

https://www.youtube.com/watch?v=f4Q2OHNPnS0 A Human-centered Framework for AI Ethics https://www.npr.org/2026/04/17/nx-s1-5750554/the-philosopher-trying-to-teach-ethics-to-ai-developers The philosopher trying to teach ethics to AI developers April 17, 20262:19 PM ET By Manoush Zomorodi,Fiona Geiran,Katie Monteleone,James Delahoussaye,Sanaz Meshkinpour The AI avalanche has buried many of us in uncomfortable, existential questions. What does it mean if a computer can do our jobs better than we can? What happens when people turn to machines for comfort and friendship? What is creativity if creativity can be automated? Philosopher Meghan Sullivan says this crisis isn't just felt by the average person who feels like AI is happening to them. On recent trips to Silicon Valley, she has spoken with many AI developers who feel conflicted. "Around here, as a result of this ubiquity of artificial intelligence, a whole lot of people are having these philosophical questions and crises right now, which make them really fun to talk to," Sullivan says. Sullivan is a philosophy professor at Notre Dame and the author of The Good Life Method: Reasoning Through the Big Questions of Happiness, Faith, and Meaning. Through Notre Dame's Institute for Ethics and the Common Good, she is meeting with AI leaders and developers to apply an ethical framework for AI — basically give AI a moral compass. Her mission is as ambitious as it sounds, but she is backed by Notre Dame's $50.8 million grant from the Lilly Endowment to champion philosophy and theology in the national conversation around AI. Her reception in Silicon Valley has been mixed. She describes occasionally hearing comments akin to: "That's cute that you think that topics like virtue and justice could still be relevant in our world." But she finds that most developers are open and genuinely curious about philosophy. "I come out here, and I think, my gosh, I've been training for this for 25 years to have this conversation with you, says Sullivan. "And I'm so glad that now you're ready to do it." In March, she attended Anthropic's 2-day summit, where Christian leaders discussed how Claude should approach the messy moral questions humans are bringing to chatbots. For instance, how should Claude react to a user who's grieving? What should it do when someone is at risk for self-harm? Meghan Sullivan delivering her TED Talk in 2025 Erin Lubin/TED As a virtue ethicist, Sullivan's field of philosophy centers around moral habits and practical wisdom. It's a school of thought that dates back to Aristotle in the early days of democracy in Athens. Having this background, Sullivan has a knack for making philosophy accessible to students, readers and now tech developers and executives. Whether in her classroom in South Bend or at a conference in Silicon Valley, Sullivan doesn't preach, she asks questions — by following the Socratic method — that help her audience realize they have more options than they've let themselves believe. "Those kinds of limitations in our imagination, those are one of the hardest forces for us to overthrow if we actually want to achieve flourishing," says Sullivan. You still have a choice. Here's what you can do Whether they're developers directly shaping AI or regular citizens navigating it, Sullivan's goal is to wake people up to their own agency. There are more paths towards the good life than we have been brave enough to imagine. "It's totally in the interest of politicians and very powerful corporate leaders to make you feel like you have no choice, says Sullivan. "You always have some choice. It might sometimes be a hard choice, but you do have it." You can choose which companies you give your money and your data to. The dust has not settled, so Sullivan says we still need to use our voices where we can. "There's power in the purse. There's power in a democracy to vote, and there's power in capitalism to say, not my money." This segment of the TED Radio Hour was produced by Katie Monteleone and James Delahoussaye. It was edited by Sanaz Meshkinpour and Manoush Zomorodi. The digital story was written by Fiona Geiran. You can follow us on Facebook @TEDRadioHour and email us at TEDRadioHour@npr.org. https://www.instagram.com/reel/DTlZRNwlOR4/ What if the secret to a good life isn't just what you achieve but how deeply you love? At the #TEDNext2025 conference, philosophy professor Meghan Sullivan drew on wisdom from Aristotle, Jesus and modern social psychology, to offer tips on how to expand your capacity for love, even in the face of our modern challenges. Watch her full @ted conferences talk at the link in our bio #LoveEthic #CommonGood #Ethics #TEDtalk A Good Life Method by Megan Sullivan Prof pf Philosophy in University of Notre Dame. Philosophy is not limited to academia. With the right questions and habits, it can become a practical tool for clearer thinking and more intentional decision-making in everyday life.

Tại sao Viktor Orban thất bại?

https://nghiencuuquocte.org/2026/04/15/tai-sao-viktor-orban-that-bai/#more-66542 Tại sao Viktor Orban thất bại? Nguồn: László Bruszt, “Why Orbán Lost”, Project Syndicate, 13/04/2026 Biên dịch: Viên Đăng Huy Trong suốt 16 năm, đất nước Hungary dưới thời Viktor Orbán là hiện thân của một tư tưởng đáng lo ngại: rằng “dân chủ phi tự do” có thể được duy trì ổn định và bám rễ quyền lực một cách bền vững. Bằng cách kết hợp ưu thế áp đảo tại các kỳ bầu cử với việc làm suy yếu có hệ thống cơ chế kiểm soát và đối trọng của các thiết chế, Orbán dường như đã giải được bài toán hóc búa của chủ nghĩa độc tài hiện đại: làm sao để thắng liên tiếp tại hòm phiếu trong khi vẫn rỗng hóa được nền dân chủ tự do. Và bởi mô hình của ông đã truyền cảm hứng cho những người ngưỡng mộ trên khắp phương Tây (và xa hơn nữa), góp phần củng cố niềm tin về sự suy tàn của nền dân chủ nói chung, nên thất bại ê chề của ông trong cuộc bầu cử lần này mang lại những hệ lụy vượt xa khỏi biên giới Hungary. Chiến thắng của đảng Tisza do Péter Magyar lãnh đạo, cũng giống như thắng lợi của Liên minh Dân sự Ba Lan trước đảng Pháp luật và Công lý (PiS) phi tự do vào năm 2023, không chỉ đại diện cho sự đảo ngược của một hệ thống tưởng chừng đã vững chắc, mà còn phát đi tín hiệu rằng những chế độ như vậy có thể mong manh hơn vẻ ngoài của chúng. Bài học rút ra không đơn giản chỉ là các chế độ phi tự do có thể thất bại. Bài học thực sự là: chính logic vận hành giúp duy trì chúng lại có thể trở thành tác nhân dẫn đến sự sụp đổ của chính chúng. Các nhà lãnh đạo phi tự do từ lâu đã biện minh cho việc tập trung quyền lực bằng cách viện dẫn thành công của các quốc gia phát triển ở Đông Á. Họ lập luận rằng, bằng cách làm suy yếu các ràng buộc thể chế, chính phủ có thể hành động quyết đoán hơn, điều phối đầu tư hiệu quả và mang lại tăng trưởng kinh tế. Nhưng so sánh này luôn là một sự đánh tráo khái niệm. Các chế độ của Park Chung-hee ở Hàn Quốc hay Lý Quang Diệu ở Singapore hoạt động hiệu quả không phải vì họ đối mặt với ít ràng buộc hơn, mà vì họ phải chịu nhiều áp lực hơn. Việc mất an ninh về địa chính trị và nguy cơ bất ổn nội bộ thường trực đã buộc họ phải đem lại lợi ích cho số đông nhân dân, nếu không muốn đối mặt với sự sụp đổ. Việc giảm bớt trách nhiệm giải trình không tạo ra sự tự mãn; trái lại, nó tạo ra kỷ luật. Nói một cách tổng quát hơn, năng lực thực thi của nhà nước phụ thuộc vào các ràng buộc mang tính kỷ luật đối với những người nắm quyền. Những ràng buộc này có thể tồn tại dưới nhiều hình thức khác nhau. Trong các nền dân chủ tự do, đó là cơ chế kiểm soát và đối trọng hiến định. Trong các chế độ độc tài phát triển, đó là những lỗ hổng dễ bị tổn thương từ cả bên trong lẫn bên ngoài. Các chế độ phi tự do đương đại vận hành dưới những điều kiện rất khác biệt. Khi thiếu vắng những áp lực tương tự như thời Park hay Lý, việc làm suy yếu trách nhiệm giải trình không tạo ra năng lực phát triển. Ngược lại, nó tạo điều kiện cho các hành vi trục lợi. Quyền lực trở thành công cụ để duy trì các liên minh chính trị hơn là để cung cấp lợi ích công cộng. Một chiến lược vốn dĩ để tăng cường năng lực nhà nước đã biến thành một hệ thống phân phối lợi ích cục bộ. Theo thời gian, logic này bào mòn nền tảng kinh tế của thể chế phi tự do. Khi lòng trung thành chính trị trở thành tiêu chí hàng đầu để phân bổ nguồn lực, hiệu quả và đổi mới sáng tạo sẽ bị bóp nghẹt. Mua sắm công ưu tiên phe cánh thân tín thay vì những doanh nghiệp năng suất nhất. Các doanh nhân trong nước phải đối mặt với tham nhũng, sự bất định và cơ hội mở rộng hạn chế. Đồng thời, các chiến lược tăng trưởng dựa vào đầu tư trực tiếp nước ngoài tuy tạo ra việc làm nhưng thường thất bại trong việc nâng cấp trình độ công nghệ hoặc tạo ra mức tăng năng suất bền vững. Đó chính xác là những gì đã xảy ra với Hungary dưới thời Orbán. Khi hiệu quả kinh tế suy yếu, khả năng duy trì liên minh ủng hộ của chế độ cũng giảm theo. Tăng trưởng chậm lại làm thu hẹp cơ sở thuế và giảm nguồn lực dành cho tái phân phối. Đầu tư vào giáo dục, y tế và dịch chuyển xã hội rơi vào trì trệ. Người dân Hungary ngày càng cảm nhận hệ thống vốn được quảng bá là ổn định thực chất là một hệ thống khép kín. Những bộ phận lớn của lực lượng lao động phải đối mặt với triển vọng u ám, tiền lương dậm chân tại chỗ và cơ hội thăng tiến hạn hẹp. Vào giai đoạn đầu trong triều đại kéo dài của mình, những động lực nội tại này phần nào bị che lấp bởi các khoản hỗ trợ tài chính từ Liên minh Châu Âu. Nhưng việc tiếp cận các nguồn lực này ngày càng đi kèm với các điều kiện về tính minh bạch của chính phủ và sự độc lập của tư pháp — chính là những hình thức trách nhiệm giải trình mà Orbán khước từ. Kết quả là một sự ràng buộc tự thân: bằng cách từ chối sự giám sát bên ngoài, chế độ đã tự giới hạn khả năng tiếp cận nguồn vốn của chính mình. Khi các ràng buộc này thắt chặt, không ngạc nhiên khi Orbán tìm đến những đối tác thậm chí còn phi tự do hơn, bao gồm Nga và Trung Quốc, đánh đổi quyền tự chủ về quản lý lấy những hình thức phụ thuộc địa chính trị mới. Một dự án bắt đầu với danh nghĩa chủ quyền lại có nguy cơ kết thúc trong sự lệ thuộc và yếu thế. Nhìn rộng hơn, Orbán đã cho thấy rằng ngay cả những hệ thống bị thao túng nặng nề nhất vẫn có thể lộ ra những kẽ hở chính trị. Chính những cơ chế giúp duy trì chế độ phi tự do, theo thời gian, có thể biến thành mầm mống của sự mong manh dễ vỡ. Mô hình của Hungary dựa trên một liên minh lỏng lẻo giữa các tập đoàn đa quốc gia, tầng lớp tinh hoa trong nước có quan hệ chính trị, và những cử tri được hứa hẹn về sự ổn định và cải thiện kinh tế. Nhưng khi tăng trưởng chậm lại, căng thẳng trong liên minh này gia tăng. Các doanh nghiệp nội địa thấy ít cơ hội hơn, trong khi cử tri đối mặt với mức sống sụt giảm và tương lai bị bít lối. Việc đánh bại Orbán trở nên khả thi khi nỗi bất bình gặp được năng lực tổ chức — khi một đối thủ nặng ký có thể đoàn kết những nhóm cử tri vốn bị chia rẽ và biến sự thất vọng thành hành động tham gia chính trị. Ở những nơi mà các lực lượng đối lập truyền thống còn yếu kém hoặc mất uy tín, điều này đòi hỏi một sự lãnh đạo có khả năng chuyển hóa những nỗi thống khổ xã hội thành một phong trào chính trị sâu rộng, huy động được sự ủng hộ xuyên suốt các tầng lớp và các ranh giới thể chế. Đó là những gì Magyar và đảng Tisza của ông đã đạt được. Trong nhiều năm, Hungary là minh chứng cho việc sự thụt lùi dân chủ có thể được thể chế hóa và duy trì trong khuôn khổ cạnh tranh bầu cử hình thức. Chiến thắng quyết định của Magyar chứng minh một điều quan trọng không kém: rằng những hệ thống như vậy không phải là không thể đảo ngược. Tuy nhiên, thất bại của Orbán, cũng như thất bại của đảng PiS tại Ba Lan ba năm trước, không đánh dấu sự kết thúc của chủ nghĩa phi tự do. Những điều kiện cấu trúc thúc đẩy sự trỗi dậy của nó — sự bất ổn kinh tế, sự rạn nứt xã hội và sự ngờ vực chính trị — vẫn tồn tại dai dẳng ở nhiều nền dân chủ. Tuy vậy, sự sụp đổ của Orbán đã thách thức cảm giác về tính tất yếu vốn bao trùm lên xu hướng rời xa dân chủ tự do trên toàn cầu. Giờ đây, nhiệm vụ khó khăn hơn mới thực sự bắt đầu: xóa bỏ các mạng lưới bảo trợ bám rễ sâu, khôi phục quyền tự chủ của các thiết chế và tái thiết năng lực nhà nước mà không lặp lại những sai lầm đã dung dưỡng chủ nghĩa phi tự do ngay từ đầu. Magyar cũng sẽ cần định nghĩa lại cách theo đuổi lợi ích quốc gia trong lòng EU — củng cố sự ủng hộ trong nước đồng thời xây dựng các liên minh xuyên quốc gia có khả năng thúc đẩy các hình thức hội nhập sâu rộng và bền bỉ hơn. Đánh bại chủ nghĩa phi tự do tại hòm phiếu đã khó. Xây dựng một nền dân chủ tự do kiên cường sau đó — một nền dân chủ vừa có trách nhiệm giải trình vừa có tính bao trùm — có lẽ còn khó hơn bội phần. Nhưng có một điều chắc chắn: các quốc gia dân chủ, dù là bạn hay thù, sẽ dõi theo tiến trình này một cách sát sao. László Bruszt, Giáo sư Khoa học Chính trị tại Đại học Trung Âu (CEU), từng giữ chức Quyền Hiệu trưởng và Chủ tịch CEU trong giai đoạn 1996-1997.

More than a quarter of private colleges are at risk of closing

https://www.npr.org/2026/04/13/nx-s1-5777582/many-private-colleges-at-risk-of-closing More than a quarter of private colleges are at risk of closing, a new projection shows April 13, 20265:00 AM ET By Jon Marcus …A new estimate projects that 442 of the nation's 1,700 private, nonprofit four-year colleges and universities, with a combined 670,000 students, are at risk of closing or having to merge within the next 10 years….More than 120 institutions are at the very highest risk, according to the forecast by Huron Consulting Group, which helps clients in industries including higher education formulate business strategies. For its assessment, the company analyzed enrollment trends, tuition revenue, assets, debt, cash on hand and other measures. Many are, like Sterling, small and rural. "Now that this might be gone, I just really worry about some students out there that are going to have less and less choices," Keeley said. It's a crisis whose magnitude has been overshadowed by political and culture-war attacks on higher education and is propelled by the simple law of supply and demand after a long decline in the number of Americans who are going to college. "We have too many seats. We have too many classrooms," Peter Stokes, a managing director at Huron, said of U.S. colleges and universities. "So over the coming five to 10 years, this shakeout is going to take place."… There are about 3,700 two- and four-year public and private degree-granting colleges and universities in the United States. That's already down from a peak of 4,726 in 2012. Almost all that have closed since then were private, for-profit schools, which enjoyed a brief boom before crashing under the weight of consumer discontent and increased regulation. …Many converging reasons explain why private, nonprofit colleges and universities, too, are now under existential strain. There are already 2.3 million fewer students than there were in 2010. A drop in the birthrate that began around the same time means there is about to be a further downward slide in the number of 18-year-olds through at least 2041. Among the other factors:  The proportion of high school graduates who go on to college is also down, from 70% in 2016 to 61% in 2023, the most recent year for which the figure is available.  The number of visas issued for new full-tuition-paying international students coming to the United States plummeted by nearly 100,000 this year, or 36%.  And looming caps on federal loans for graduate study, which take effect in July, threaten to reduce demand for yet another crucial revenue source. While higher education institutions previously weathered short-lived declines in enrollment and increases in costs, today "every major revenue stream and expense category is under pressure at the same time," the higher education consulting firm EAB warns in a new analysis. Eighty-six percent of college and university leaders are worried about their schools' long-term financial viability, according to a survey in December by the American Council on Education, the principal industry association. A fifth of college and university presidents say they've had serious discussions about merging with another university or college, a separate survey by Hanover Research and the industry news site Inside Higher Ed found. Signs of strain are spreading And nearly a third of private, nonprofit colleges and universities nationwide posted deficits in 2024, according to research by Robert Kelchen, director of the Department of Educational Leadership and Policy Studies at the University of Tennessee, Knoxville. And it's not just small schools that are affected. Even public universities and colleges are facing deepening financial problems, reports the Fitch bond-rating agency, citing slowing economic growth and federal policy changes. The University of Southern California has sent pink slips to more than 900 employees. Stanford University, Northwestern University, and Depaul University have also seen layoffs. And, as part of what its president called a "broader strategy to strengthen GW's long-term financial health," George Washington University announced in March that it had sold a satellite science and technology campus in Virginia for what the student newspaper reported was $427 million. Community colleges, too — which enroll nearly 5.6 million students — are suffering financial squeezes that leave them less able to adapt or respond to change, according to Daniel Greenstein, former chancellor of the Pennsylvania State System of Higher Education, who now tracks financial exposure in the industry. In the case of community colleges, wrote Greenstein, "The risk is not a sudden collapse of the sector. The risk is a slow erosion of capacity in precisely the institutions on which communities rely most." Still, after two and a half decades in which the price of tuition has increased faster than inflation, for a payoff many consumers no longer think is worth the money, higher education often gets little sympathy for its predicament — and even less after years of political and culture war attacks on the ideological leanings of faculty and leadership. https://www.theatlantic.com/ideas/2026/04/college-enrollment-demographic-cliff/686750/ The Looming College-Enrollment Death Spiral After many decades of democratization, higher education could once again become a luxury good. By Jeffrey Selingo April 12, 2026, 7 AM ET The “demographic cliff” is upon us. The number of teenagers graduating from American high schools peaked last year. It will begin declining this spring and keep falling steadily through at least 2041. The trend is more of a downward slope than an abrupt falloff, but the gradient is steep and represents a crisis to colleges dependent on filling classroom seats and dorm beds. The United States currently has about 4,000 colleges. According to a recent study from the Federal Reserve Bank of Philadelphia, about 60 are closing on average each year; that number could double in any given year if the bottom falls out of enrollment. If the harm were only to the institutions forced to close because they’re running out of customers, that would be unfortunate but not tragic. But the causality runs in the other direction too, as students who otherwise would have gone to college find themselves with no viable option in the place where they live. American higher education has long consisted of two markets: one where high-achieving, typically affluent students compete for seats at national universities, and one where mostly middle- and lower-income students stay closer to home. Members of the first group will be fine even as college closures accelerate. The second group will suffer. After many decades of democratization, higher education could once again become a luxury good. Over the past half century, as more teenagers have enrolled in higher education, what was once mostly a local business has become national, especially for top students, whose sense of distance has gradually shifted. Campuses that once felt far away now seem closer, thanks first to interstate highways, then to discount airlines, and then to technology. Parents in the 1980s might have talked to their college kid on a dorm-floor pay phone once every few weeks, if they were lucky. Today’s parents can text and FaceTime their kids multiple times a day…. https://www.theatlantic.com/ideas/archive/2025/10/ivy-league-schools-prestige/684454/ The ‘Best’ Colleges Aren’t the Best Forever Prestige isn’t permanent. By Jeffrey Selingo October 4, 2025 For decades, higher education seemed immune to market forces, as families stretched to pay almost any price for a top-ranked college. Prestige was seen as synonymous with enduring value: Harvard would always be Harvard, Yale would always be Yale, followed by the Northwesterns and the Cornells, with aspirants such as the University of Southern California and Northeastern further down the ladder. But with sticker prices surging and graduates facing a tough job market, many parents have begun to question whether prestige alone is worth the price. As reputation loses some of its grip on the marketplace, colleges are moving up and down the list more than ever. How we think about brands in higher education was largely decided centuries ago when America’s top colleges were established. These perceptions were cemented in the late 1980s, when U.S. News & World Report turned its college rankings into an annual exercise. A school’s “reputation score,” as determined by a survey of college leaders, was the most heavily weighted factor in assigning it a ranking on the list. Reputation is still the biggest factor in the U.S. News methodology, and plenty of people still care enough about an exclusive brand to pay a premium for it. In recent years, however, many families have begun to put more emphasis on practical matters such as tuition costs, hands-on learning, and career outcomes. This evolution in priorities stems partly from personal experience. Today’s parents—who are more likely than their parents to be college graduates—have seen the college hierarchy change in their lifetime. When U.S. News released its 1989 rankings, it not only issued overall rankings, but also listed the top 25 colleges by reputation alone. A few of the names among the latter list seem like typos today: the University of Illinois Urbana-Champaign, Purdue University, Indiana University Bloomington. Meanwhile, schools that were considered regional brands three decades ago, such as the University of Southern California and New York University, have risen in the rankings and now have acceptance rates that rival those of the Ivy League. Last cycle, NYU broke its own record, with more than 120,000 applications for a class of some 5,700 students….

What a chimpanzee 'civil war' can teach us about how societies fall apart

https://www.npr.org/2026/04/13/nx-s1-5781149/chimpanzee-civil-war-primate-conflict-anthropology What a chimpanzee 'civil war' can teach us about how societies fall apart April 13, 20266:00 AM ET. By Nathan Rott https://npr.brightspotcdn.com/dims3/default/strip/false/crop/4272x2848+0+0/resize/1200/quality/85/format/webp/?url=http%3A%2F%2Fnpr-brightspot.s3.amazonaws.com%2Fc2%2Fee%2F711a58af44b996a1a0f9281fcda0%2Fsandel-adz4944-image-1.jpg The Ngogo chimpanzee group in Uganda's Kibale National Park is the largest known community of wild chimpanzees in the world. Over the last decade, it has split into two distinct groups that are hostile to each other. Aaron Sandel The group of chimps she and her colleagues were studying broke into two factions and turned on each other. It looked very much like a civil war. Chimpanzees that had intermingled peacefully and grown up together were systematically killing each other. It changed Goodall's view of one of humanity's closest relatives. "I used to think, 'Well, they're very [much] like people but nicer,'" she told the public radio program Fresh Air in 1993. "And then I realized that when opportunity arises, they have this nasty, brutal side to them just like we do." Asked what precipitated the war, Goodall said it was hard to say. It was the first one that researchers had ever seen. "We shan't be very sure until it happens again," she said Now, in the journal Science, a team of researchers has described a second brutal and ongoing "civil war" that has permanently divided the largest known group of wild chimpanzees in the world. "I was struck by some of the similarities of what they've described to what we observed in Gombe," said Anne Pusey, a retired primatologist who worked with Goodall in Tanzania and wasn't involved in the new study. "It's rather uncomfortably familiar seeing how these relationships can break down and then lead to antagonisms between groups that weren't there before." The new study draws from more than 30 years of observations of the Ngogo chimpanzee group in the western forests of Uganda. At its peak, nearly 200 individuals were in the Ngogo group, living cohesively in smaller subgroups that the researchers labeled as "clusters." Males and females from different clusters intermingled. They mated, hunted together and worked together to fight off other outside groups. Researchers took videos of males from different clusters holding hands. The new study draws from more than 30 years of observations of the Ngogo chimpanzee group in the western forests of Uganda. At its peak, nearly 200 individuals were in the Ngogo group, living cohesively in smaller subgroups that the researchers labeled as "clusters." Males and females from different clusters intermingled. They mated, hunted together and worked together to fight off other outside groups. Researchers took videos of males from different clusters holding hands. Then, in 2015, the researchers started seeing signs that something was off. "I can even pinpoint it to one particular day when there was a really big change," said Aaron Sandel, the lead author of the new study and a primatologist at the University of Texas at Austin. On that June day, Sandel was observing a large number of chimpanzees from the Western cluster while they were in their territory. At one point, they heard other chimpanzees nearby, presumably from the larger Central cluster. The Western chimpanzees quieted all of a sudden. "They started touching each other in this reassurance, like they were really nervous," Sandel said. "And to me, this seemed like they were acting as if they were hearing outsider chimps." Instead of reuniting and intermingling like they normally would, the Western chimpanzees fled and the Central chimpanzees chased them. "Nothing really like that had ever been observed before — and then they avoided each other for six weeks," Sandel said. "So this was very clear, like on the ground, something big has just happened." Over the next few years, polarization increased, and by 2018 the clusters were essentially completely separate groups. Then the killing started. The victim of the first observed lethal attack was an adolescent male from the Central cluster that the researchers had named Errol. Sandel had watched him grow up. "I'm just trying to observe as objectively as possible and really just document everything," he said. "In some respects, I feel like a war correspondent. I'm trying to understand this really rare behavior. … Like what's causing this?" Over the next seven years, the Western group killed at least six other adults and 17 infants from the Central cluster. The fighting continues to this day. Why the Ngogo group split and why its members turned on each other is still unclear. In the paper, Sandel and his co-authors suggest several factors that may have contributed: the size of the group, competition for food and male-to-male competition. The natural deaths of five adult males and one adult female in 2014, before the intergroup divisions took root, may have weakened social networks. "I think it's clear from this study and from other studies of chimps and other animals that you can get these kinds of conflicts without a lot of things that we think about as being the source of conflict in humans," said Michael Wilson, a primatologist at the University of Minnesota, who wasn't involved in the study. "Lions don't have religion and political parties or ideologies. Neither do wolves or ants for that matter." Neither do chimpanzees, the authors of the new study note. To Sandel, that's a reason for optimism. "If in chimpanzees, we can see this conflict and lethal violence occur in the absence of all these aspects of human behavior that we often attribute to civil war, then I wonder to what extent are the interpersonal rel ationships and behaviors actually more important than we realize in humans," he said. Perhaps, he added, strengthening our social bonds and letting old grudges die can help prevent larger violence. "Like with the chimps: If you act like a stranger, you become a stranger," Sandel said. "I want to avoid that in my own life."

Satellite Imagery

https://www.npr.org/2026/04/10/nx-s1-5775780/us-iran-war-israel-satellite-imagery-planet-vantor-censorship … Unlike previous conflicts, the U.S. has also worked harder to restrict information from the region as well. Trust between the Pentagon and journalists was already low – many in the press corps, including NPR, left the building last fall after Secretary of Defense Pete Hegseth demanded members of the press sign a pledge to not solicit information outside of press briefings. With little access to the Pentagon or troops on the ground, satellite images played an outsized role early in the Iran conflict. Reporters used the images to document blows traded by the U.S., Iran and Israel. But within days, the satellite imagery was causing headaches for war planners. On March 3, CNN published Planet imagery showing a base where six U.S. servicemembers were killed. The New York Times published extensive analyses of the damage to communications infrastructure and bases throughout the region. And many outlets, including NPR, used satellite imagery to show that a strike at a girl's elementary school in Iran was part of a larger set of targets at a nearby military base. The U.S. subsequently took responsibility for the strike, and an investigation continues. [Secretary of Defense Pete Hegseth has sought to tightly control information coming out of the Pentagon.Image] By mid-March, the two largest U.S. firms, Vantor and Planet, stopped distributing imagery to the press altogether. Planet then imposed a 14 day hold on all imagery out of the region before switching to an indefinite moratorium. Planet operates a fleet of around 150 satellites that photograph most of Earth's landmasses on a daily basis. Its images had become a mainstay for observers of the events in the Middle East in recent years. The company's pictures have been used to help track atrocities in Syria, document previous attacks by Iran, and chronicle Israel's destruction of Gaza. Both Vantor and Planet say that the decision to begin limiting imagery has been voluntary. "Vantor independently determines when and how these controls are implemented as part of our responsible business practices," the company wrote in an e-mail. "These decisions are not mandated by any government or third party." In an e-mail to NPR on Thursday, Planet added that it hoped to restore access soon: "We remain highly engaged with the U.S. Government," the statement read. "Our goal is to get back to unrestricted access for all of our customers globally as soon as possible, while continuing to limit the risk that our data could be misused." Both companies' satellites are regulated by the government, and both are heavily dependent on business from the Department of Defense and intelligence agencies around the world. Nearly 60% of Planet's revenue in its last quarter came from defense and intelligence contracts, according to a recent shareholder report. Vantor is also a major government contractor that has been awarded millions by agencies like the National Geospatial-Intelligence Agency and Department of Defense, among others. Last July, the company announced it had won $205 million in contracts with several nations in the Middle East and North Africa, though it declined to say which nations are involved. Censorship workarounds Experts contacted by NPR said that the efforts at censorship had worked to some degree–and they'd made their jobs harder. But there's still information to be found online. On Telegram, Iranian channels, many of which are pro-government, frequently publish videos of events inside the country. Images also continue to trickle in from the Gulf countries, such as a widely circulated image of a destroyed American E-3 Sentry aircraft that was taken at a base in Saudi Arabia. Verifying the images and videos, especially in the age of AI, can be tough. "The big problem is that it becomes very hard to fact check things like videos," said Lewis. The ability to sort fact from fiction "is a lot harder when you don't have recent ground-truth satellite imagery." But the satellite imagery isn't completely gone either. Publicly-funded satellites continue to supply images of the Middle East, albeit at lower resolution than the commercial companies. And a trickle of images from other providers, such as Airbus, continue to provide insights into aspects of the conflict. A satellite image from NASA's Terra spacecraft shows fires burning in the United Arab Emirates on March 16, 2026. Some lower-resolution imagery continues to be available from publicly funded satellites. NASA Worldview The online community of people who do this kind of analysis are used to their information environment constantly shifting according to the whims of companies and algorithms, Godin said. And they're good at finding workarounds: On Tuesday, Bellingcat unveiled an online tool that uses radar data from an old satellite to look for damage from strikes throughout the region. Godin said he continues to keep very busy, and he doesn't expect efforts at censorship to change that. "It's not great that these things are happening," he conceded. "But we're a resilient bunch." NPR's Aya Batrawy and Sarah Knight contributed to this report.

American Diplomacy

https://foreignpolicy.com/2026/04/10/american-diplomacy-problem-witkoff-kushner-iran/