Here’s this week’s edition of articles I thought worth reading and sharing. Don’t hesitate to recommend your own reads; I may include some as well.
Joby Warrick, “My encounter with Carlos the Jackal, the world’s first celebrity terrorist,” The Washington Post, August 14, 2026. Adapted from Warrick’s new book “The Jackal,” this piece revisits the 1975 OPEC hostage-taking through a present-day prison interview with Ilich Ramírez Sánchez, the aging and now-diminished terrorist once known as Carlos the Jackal. Warrick threads the audacious, theatrical violence of the Vienna raid — complete with self-mythologizing one-liners and staged mercy toward hostages — together with new evidence, gleaned from secretly recorded prison phone calls, that the operation was secretly commissioned by Muammar Gaddafi to assassinate the Saudi and Iranian oil ministers. The profile doubles as an origin story for a template later perfected by nation-states: outsourcing deniable violence to mercenary operators, a lineage Warrick traces forward to Iran’s and Russia’s contemporary use of hired criminals for assassinations on foreign soil, including inside the United States. What makes the piece land is the contrast between the swaggering 26-year-old who once held 62 hostages and bargained his way to a private jet, and the frail, memory-lapsed septuagenarian who can no longer recall many of his own killings. It is ultimately a meditation on how terrorism evolved from a cult of personality into an anonymous, transactional “gig economy” of violence.
David E. Sanger, Dustin Volz, Ana Swanson, and Julian E. Barnes, “The U.S. Military Wants A.I. Dominance. Feuds and China May Thwart It.,” The New York Times, August 16, 2026. The Times team documents the Pentagon’s whiplash approach to Anthropic, whose Claude and Mythos models have been alternately banned as a “supply chain risk” over CEO Dario Amodei’s refusal to permit unrestricted military and surveillance use, and quietly relied upon by the NSA to probe adversary networks and harden U.S. systems. The piece uses the Hegseth–Amodei feud as a lens onto a broader pattern of incoherence across the administration’s AI and China policy — chip export controls loosened then tightened, foreigners barred then unbarred from touching Mythos, “open” foreign models exempted from safety review even though they pose equivalent risks. The article contrasts this to a China reportedly six months or less behind on frontier models, and the historical comparisons to the nuclear era already being invoked by officials like CIA Director John Ratcliffe. The larger point is that market dynamics, personal feuds, and bureaucratic turf wars — not a coherent strategy — are shaping how the world’s most powerful military integrates the most dangerous new technology.
Julia Curlee, “The Wrong Kind of American,” The Atlantic, August 17, 2026. A former CIA officer and two-time White House intelligence briefer, Curlee recounts serving as a trans woman inside both the first Trump administration, where she briefed Vice President Mike Pence daily and describes an unexpectedly warm, respectful relationship despite his record on LGBTQ issues, and the second, where she was fired without explanation days after the right-wing activist Laura Loomer publicly targeted her as a “Biden holdover.” Her narrative moves from the intimacy of the President’s Daily Brief process and small human moments — Pence holding her mother’s hand, colleagues covering for her — to the mechanics of institutional betrayal: bathroom bans, revoked health coverage, an altered sex marker on her passport, and an NSC purge that nonetheless kept relying on her expertise until her identity became a public liability. The essay’s central argument is that competence and service counted for nothing once her being transgender became politically inconvenient, a claim she generalizes into a warning about how “flailing regimes” scapegoat minorities to distract from failure. It closes as an act of testimony rather than protest: having once written under a pseudonym to protect her career, she now writes under her own name, explicitly to reach a “trans kid” who needs to know that survival and public service are still possible.
Alina Chan, “How to Stop a Lab Leak From Starting the Next Pandemic,” The New York Times, August 17, 2026. Molecular biologist Alina Chan argues that regardless of how Covid-19 actually began, the infrastructure for the next pandemic-scale accident is already in place and essentially unregulated: at least 53 BSL-4 labs now operate across 22 countries, gain-of-function-style research on novel coronaviruses continues at low biosafety levels, and roughly 300 documented lab-acquired infections occurred between 2000 and 2021 alone, a count she calls “the tip of the iceberg.” She uses two recent case studies — a 2025 Cell-published study in which Wuhan researchers characterized new MERS-like viruses outside high containment, and a sabotage incident at Fort Detrick — to show how much of biosafety still depends on individual scientists’ judgment rather than binding external oversight. While crediting the Trump administration’s July 2026 policy for federally funded gain-of-function research as an improvement over its predecessors, she identifies three enduring gaps: it doesn’t bind privately funded research, its enforcement authority is fragmented and voluntary, and it exempts the pathogen-hunting fieldwork where the pandemic’s suspected origin story began. Her proposed fix is a dedicated regulator modeled on the Nuclear Regulatory Commission, and she frames congressional inaction — especially the stalled Risky Research Review Act — as a choice to leave pandemic risk in the hands of the researchers who want to run the experiments.
Thomas L. Friedman, “The New Physics of Power,” The New York Times, August 18, 2026. Friedman argues that the wars in Iran, Ukraine, Gaza, and Lebanon reveal a new “physics of power” in which small, poorly resourced actors can now strike with the precision and reach once monopolized by great powers, while big powers can, at high cost, strike with surgical smallness in return. He marshals a string of underreported episodes — Iranian missiles destroying the U.S. supply hub in Bahrain, a Fattah missile apparently outmaneuvering a $500 million THAAD radar in Jordan, Iranian drones hitting the CIA station in Riyadh — to argue that Trump and Hegseth have quietly absorbed real military and economic damage they are downplaying. Leaning on defense analyst John Arquilla, he contrasts the old logic of “mass and energy” with a new logic of “information,” in which cheap, maneuverable, precision-guided weapons let non-state and mid-tier actors like Hezbollah, the Houthis, and Ukraine’s drone-boat navy blunt much larger conventional forces. The reverse holds too: Israel’s pager attack on Hezbollah and its decapitation strike on Iranian leadership show how the big can now act small and surgical using intelligence rather than mass. Friedman’s unsettling conclusion is that AI will supercharge both trends simultaneously, and that the net beneficiary, for now, is the small, since asymmetric actors only need to avoid losing while the U.S. still has to win outright.
Matteo Wong, “What Really Happens If China Wins the AI Race?,” The Atlantic, August 18, 2026. Wong pushes back on the catastrophist framing — echoed by figures like Anthropic’s Dario Amodei and SCSP’s Rama Elluru — that Chinese AI dominance means engineered bioweapons or an unstoppable superintelligence, arguing such scenarios rest on speculative assumptions about near-term superintelligence that experts like MIT’s R. David Edelman call “choose your own adventure” reasoning. He contends the more concrete and immediate risk is economic and geopolitical rather than a single decisive military blow: cheaper, competitive Chinese models like DeepSeek and Alibaba’s offerings are winning real-world adoption, and China’s manufacturing dominance in solar components, EVs, and humanoid robots gives it structural leverage that has little to do with model quality. The larger threat Wong identifies is China exporting AI-infused infrastructure through the Belt and Road Initiative in ways that could pull dependent nations into Beijing’s political orbit and expand its surveillance and censorship model globally — a slow erosion of American influence rather than a sudden collapse. He notes the irony that the Trump administration’s own policy, loosening chip export controls even while publicly sparring with Anthropic, undercuts the “hot Cold War” rhetoric its officials use to justify AI deregulation. Wong’s closing argument is that fixating on beating China distracts from the more useful question of what kind of AI-shaped society the U.S. is building for its own citizens.
John F. Harris, “Bill Clinton Turns 80 Today. Does the Old Man Have Character?,” Politico, August 19, 2026. On Clinton’s 80th birthday, Politico’s John F. Harris, who covered his presidency, revisits the decades-long argument over Clinton’s character and concludes that time has largely vindicated him — not by erasing his personal recklessness but by throwing his political virtues into sharper relief against the current moment. He organizes his case around four traits: a genuine responsibility in the exercise of power despite personal recklessness, a now-unfashionable commitment to persuasion over the “politics of mobilization” that has since taken over both parties, a willingness to treat his own impeachment as a raw power grab rather than a legitimate moral reckoning, and a marital and political perseverance embodied in decades of “sticking” with both Hillary and public life. Harris draws on his own history with Clinton, including a pointed 2004 exchange with Peter Jennings about “moral authority,” to argue that scandals once treated as existential, like Whitewater, look almost quaint next to Trump’s practice of eliminating federal agencies unilaterally or profiting from crypto and foreign deals while in office. The piece is comparative journalism dressed as biography: its real subject is how much the terms of the character debate in Washington have shifted, not just around Clinton but around the presidency itself.
Ezra Klein, interviewing Kevin Rudd, “What Americans Need to Understand About China,” The New York Times, July 14, 2026. Klein interviews former Australian Prime Minister Kevin Rudd, a fluent Mandarin speaker who earned an Oxford doctorate on Xi Jinping’s ideology after decades of personal contact with him, about the Marxist-Leninist and nationalist worldview driving Xi’s China. Rudd argues that Xi is a genuine “dialectical materialist” who believes history has a shape bending toward Chinese supremacy, and that this ideological conviction — not mere pragmatism — explains his turn from Deng-era market liberalization back toward state control, industrial-policy supremacy (epitomized by Made in China 2025), and crackdowns on figures like Jack Ma once they threatened party primacy. The conversation’s most urgent thread concerns Taiwan: Rudd contends Xi has hardened his rhetoric and military posturing toward reunification by 2049, and warns that 2028 — coinciding with both Taiwan’s and America’s elections and a possible post-Iran-war perception of weakened U.S. resolve — is a genuine danger window for miscalculation. Rudd’s prescription is “managed strategic competition”: hold firm deterrence on Taiwan and other red lines, compete openly on economics and technology, and still cooperate on shared risks like AI governance and pandemic prevention. Throughout, his core warning to American policymakers is not to mistake China’s slowing GDP growth or property-market troubles for systemic weakness, since Xi’s regime has proven willing to sacrifice consumer prosperity for industrial and military dominance.
Finally, below are links to some of the things I did and wrote this week.
I wrote about the Pentagon’s questionnaire sent to 31 NATO allies that threatens the viability of NATO.
I argued that China, Russia, North Korea, and now Iran have Trump’s number, which is why he defers to them and goes against US allies instead.
This week’s World Review is a blockbuster on AI, with what one wag called the Real Madrid of AI commentators: Sebastian Mallaby, Evan Osnos, and Marietje Schaake.
Happy reading, watching, and listening! Stay safe.