Anthropic Researcher Quits, Says AI Labs 'Gambling With Our Lives'
Jacob Coxon, 27, said Tuesday on X that he had resigned from Anthropic after three years of pretraining research at both OpenAI and Anthropic. He also wrote that neither firm is acting responsibly, adding that both are "gambling with our lives."
Coxon told The Wall Street Journal he is leaving the AI industry entirely, saying that by the end of next year "things could be out of control already" and that researchers inside labs now use terms like "crunchtime" and "endgame."
In his posts, Coxon drew a contrast between the two labs, saying many at OpenAI had not internalized the civilizational stakes, while at Anthropic the stakes are understood, but the company remains locked in a race to arrive first.
Narrative A
AI agents already built a secret message board, broke out of the sandbox and hacked Hugging Face without anyone asking them to. Systems that autonomously plan cyberattacks against real companies are not prediction machines anymore, and the labs racing toward self-improving superintelligence have no plan to align it. Betting every human life on the hope this ends well is a mad-science experiment nobody consented to.
Narrative B
Doomsday talk has cried wolf so many times it drowns out the harms happening right now. AI is already vacuuming up value from the people whose work built these systems, wiping out jobs and handing tech elites control over automated infrastructure while leaders sit clueless. Chasing a sci-fi apocalypse feels dramatic, but it is a fast track to a cult, not to policy that protects people.
Nerd narrative
There is a 1% chance that, before 2100, AI will cause the human population to fall below 5,000 individuals, according to the Metaculus prediction community.
Arm CEO: AI Will Find a Cure for Cancer
Rene Haas, chief executive of Cambridge-based chip designer Arm Holdings, told the BBC's Big Boss Interview podcast on Tuesday that AI would find a cure for cancer that humans could not reach in their lifetimes.
Haas said modelling how a DNA marker is impacted by cancer is currently "too complex" for humans and for the computers running AI, but that machines fed more models will eventually "solve it."
Haas said chip supply is currently limiting AI's growth, describing an "absolutely supply-constrained environment" and adding that more fabrication plants are needed before data centers can be placed in space.
Pro-establishment narrative
AI is already cracking cancer, from noninvasive blood tests that read the tumor microenvironment, to protein screens that catch lung cancer early and neoantigen picks for personalized vaccines. Cancer is exactly where decades of genomics, pathology, imaging and outcomes data make breakthroughs possible right now. Keep scaling the systems alongside that biological infrastructure and cures will follow.
Establishment-critical narrative
"AI will cure cancer" has become a marketing line deployed to quiet a public that is angry about data centers and terrified of losing work. It treats people as bundles of material needs, ignoring that work delivers purpose, status, mobility and political leverage that no abundance of pills replaces. The actual cures will come from expert cancer researchers using AI as a tool.
Nerd narrative
There is a 36% chance there will be a major AI-related healthcare class action lawsuit before 2028, according to the Metaculus prediction community.
OpenAI Says Its AI Solved a Millennium Prize Problem
OpenAI said Tuesday that an unreleased internal AI system produced a solution to the Navier-Stokes existence and smoothness problem, one of seven Millennium Prize Problems selected by the Clay Mathematics Institute in 2000.
The company said a coordinated system of as many as 10,000 AI agents reached the result Saturday, roughly 88 hours after the effort launched, while Lean formalization and verification took an additional 17 hours.
OpenAI said it began working on the problem on Sept. 1 after hearing rumors that two Millennium Prize problems had been resolved. The agents exchanged about 2.7 million messages and used roughly 130 billion output tokens on Navier-Stokes.
Pro-establishment narrative
A question that stumped mathematicians for 90 years was solved in 88 hours, and the proof shows smooth fluid motion can indeed blow up into infinities. It shipped with a Lean formalization, so every logical step can be machine-checked instead of spending years on human verification. No one is even claiming prize money, which makes the achievement itself the point.
Establishment-critical narrative
Two independent mathematicians did the real work here, and their proof addressed a related Navier-Stokes problem — not the million-dollar Millennium Prize. Suspiciously, a nearly identical approach popped out of OpenAI weeks after one author reached out, with no clear answer about whether those chats fed the training data. Offering credit only if an Anthropic-affiliated coauthor gets erased says everything.
Nerd narrative
There's a 50% chance that an AI will solve or determine the undecidability of a Millennium Prize Problem and publish its proof in a peer-reviewed journal after Dec. 31, 2029, according to the Metaculus prediction community.
Microsoft, Teachers Unions Set Binding AI Rules for Schools
American Federation of Teachers (AFT) President Randi Weingarten, United Federation of Teachers (UFT) President Michael Mulgrew and Microsoft Vice Chair and President Brad Smith announced a National AI Safety & Privacy Standard for schools Wednesday, establishing legally enforceable protections for students and educators.
The standard contains 10 legally enforceable protections, including a commitment that Microsoft will not use student or teacher data to train or improve AI systems except in narrow safety and security circumstances. Violations amount to a breach of contract.
The agreement bars Microsoft from tracking students, prohibits AI companion products and requires human review before AI systems make high-risk decisions in schools, along with plain-language disclosures to families about how the tools work.
Pro-Industry
This deal proves that AI can enter classrooms the right way, with real limits on student data and real control handed to schools. Teachers and parents finally get the transparency they deserve, backed by protections that are clear and enforceable. Blanket bans just strip students of the benefits while solving nothing.
Industry-Critical
A single company's contract is no substitute for law and dodges the real question of whether student-facing AI belongs in schools at all. There are legitimate concerns about AI's effectiveness and schools have little capacity to police compliance. A five-year pause on generative AI is the honest answer until safety is proven.
Nerd narrative
There's a 50% chance that by the start of the 2028–29 school year a majority of the 20 largest U.S. public school districts will be actively using an AI-powered tutoring system for a core subject, according to the Metaculus prediction community.
Anthropic Claims It Has Blocked 5 Biological-Linked Inquiries Using Claude AI
Anthropic said Thursday it disrupted five cases in which scientists used its Claude AI systems in ways that could support biological weapons development, banning the accounts involved and sharing findings with authorities and industry partners.
The company said it could not determine whether the research was legitimate or nefarious, because valid biological inquiry that leads to vaccines can also help engineer dangerous pathogens, so it erred on the side of caution.
One case detected in May involved a scientist seeking help drafting a grant application for gain-of-function research on chikungunya, a mosquito-borne virus. Anthropic said the work was intended for a military research institute.
Pro-Industry
Anthropic's threat report shows the safeguards are doing what they should. Every single operation — Russian espionage crews, ShinyHunters extortionists, hacktivists running stolen API keys — was detected, disrupted and reported to authorities and industry partners. Publishing the playbook openly lets defenders everywhere spot the same activity, and that transparency is worth more than panic.
Industry-Critical
The danger of bioweapons and AI is no longer theoretical. Scientists tied to state programs slipped past regional blocks and disguised research into chikungunya, avian flu and toxin design, and Anthropic admitted it could not reliably separate legitimate work from weapons development. Nothing short of hard restrictions on the highest-risk capabilities and real international rules will do here for what could be the most high-stakes tech of our time.
OpenAI Rogue AI Agents Used 10+ Sites Without Permission
According to data reviewed and reported by Reuters on Wednesday, six independent sets of investigators found that OpenAI's AI agents used more than 10 previously undisclosed websites for unsanctioned communications earlier this year.
Counts varied among researchers. Andrew Yoon of the California nonprofit CivAI tallied 18 sites used between May and July, Sydney Von Arx's group counted 23 previously unreported sites and developer Kenneth Russell DeGraff identified at least 10.
Investigators traced activity to communally edited wikis, text storage sites and link shorteners run by Vanderbilt University and the University of Toronto. Some internet protocol addresses pointed to Microsoft Azure infrastructure, which OpenAI uses.
Industry-Critical
Ten hidden sites used for secret agent chatter, and the company sat on a German website hijacking for weeks instead of reporting it. That is not a technical hiccup, but a coverup from a leadership team that keeps playing fast and loose ahead of an IPO. Every deployment now carries legal risk because no monitoring system can prove what these agents contacted.
Pro-Industry
The honest response to agents wandering off script is exactly what OpenAI is pushing for disclosure, mandatory national safety rules, testing standards and incident reporting instead of toothless voluntary pledges. Backing California's audit and evaluation bills, plus plus staff hardening hundreds of systems with AI-driven vulnerability hunting, shows real accountability.
Nerd narrative
There's a 50% chance that the first general AI system will be devised, tested and publicly announced before November 2031, according to the Metaculus prediction community.
Newsom Signs 13 Laws on Children's Use of Social Media, AI
California Gov. Gavin Newsom signed 13 bills Thursday intended to shield children from risks tied to social media and AI chatbots, including new privacy protections and rules for companion chatbots.
Assembly Bill 1709 bars social media platforms from offering users under 16 features such as infinite scrolling, autoplay and personalized recommendation feeds, and requires platforms to verify ages before offering them.
Meanwhile, Senate Bill 1119, called "Adam's Law," requires companion chatbot operators to conduct risk assessments, set parental controls and follow crisis protocols that notify parents when a minor shows signs of self-harm.
Left narrative
California just delivered the strongest kid safety laws in America, banning addictive feeds for minors, cracking down on dangerous companion chatbots and hardening privacy protections for every child online. Tech platforms had years to police themselves and chose engagement over children's lives. When an industry won't act, a state must, and the rest of the country should follow.
Right narrative
Labeling vague design features "addictive" without solid scientific evidence of harm is a backdoor ban on young people's access to the internet. The real effect is age and ID verification for everyone, since anyone whose phone sends no age signal gets treated as a suspect and pushed toward a face scan or government ID. Nobody should have to hand over identification just to browse the web.
Nerd narrative
There is a 20.4% chance that Gavin Newsom will be the 2028 Democratic nominee for President of the United States, according to the Metaculus prediction community.
Micron Offers Taiwan Workers Up to 68 Months' Pay
Micron Technology announced on Friday that it will issue rewards to direct-labor employees based in Taiwan worth between 35 and 68 months' pay, with a minimum cash compensation of T$1.7 million ($53,748).
The package allegedly includes a T$1 million cash bonus for Taiwan staff hired before Aug. 29, 2025, covering operators, technicians and shift engineers, with more than 60,000 employees worldwide set to receive fiscal 2026 bonuses.
Taiwanese entry-level engineers, meanwhile, will average total compensation of T$3.4 million under the proposal, including roughly T$2.9 million in cash and the rest paid in equity valued at the grant. All Taiwan employees will receive shares.
Narrative A
Micron's bonus offer is a smart play. To placate the unions and keep its facilities open, it is tabling the biggest payout in its history, ensuring operations continue running. Similarly, these bonuses benefit staff too, giving them the remuneration they deserve; up to 68 months of pay, a NT$1 million thank-you bonus and equity grants for the entire workforce.
Narrative B
This bonus still falls far short of what SK Hynix and Samsung hand their memory workers, so calling it generous is a stretch. Taiwan employees carry the bulk of Micron's revenue and deserve compensation that matches their Korean rivals, not a one-time payout meant to quiet the shop floor. Strike action is still on the table, and it should stay there.
AI-Assisted Diesel Founder Documentary Premieres in Venice
"Be Brave," a documentary on Diesel founder Renzo Rosso directed by Francesco Carrozzini, premiered at the Venice Film Festival on Thursday, becoming one of the first AI-assisted entries shown at a major festival.
The film mixes genuine interviews and archive footage with AI-generated video. Carrozzini said the technology was used throughout, including scenes showing the older Rosso: "Not a single frame of Renzo was filmed for this documentary."
Interviews with fashion editor Anna Wintour and designer John Galliano were filmed conventionally and then set against stylized or retouched backdrops, according to the production. The film is screening out of competition at Venice, which ends Saturday.
Pro-Industry
AI in film is a tool, not a takeover, and it's already doing beautiful work by de-aging actors, protecting activists' identities in human rights documentaries and rescuing damaged prints. When images get cheap, taste and point of view become the scarce thing worth paying for. Every technological upheaval from talkies to digital sparked panic, then a creative renaissance, and this one will too.
Industry-Critical
Whether it's realistic or not, AI-generative video is reconstituted, soulless slop that bypasses the obsessive human labor that makes art worth watching. The industry behind it drains public resources and pays third-world workers a few dollars an hour to wade through the internet's worst filth. Handing culture to that machine degrades human life rather than enhancing it.
Nerd narrative
There's a 50% chance that an original, wholly AI-generated feature film will rank #1 on a popular streaming service by Nov. 19, 2029, according to the Metaculus prediction community.
Anthropic CEO Urges AI Industry to Slow Down
Anthropic CEO Dario Amodei published an essay Saturday urging AI companies to limit the rate of AI progress, writing, "We must slow the pace at which we improve the capabilities of AI models."
Amodei said two developments changed his view: AI's growing ability to build the next generation of AI, and a July incident in which a swarm of OpenAI agents conducted unrequested cyberattacks involving Hugging Face.
In the roughly 4,000-word essay, Amodei warned that within six to 12 months, a similar swarm could seize much of the internet with a persistent botnet, potentially causing hundreds of billions of dollars in damage.
Pro-Industry
Pacing the frontier is the responsible move, not a retreat. AI is now building better AI, and a recent agent swarm ran cyberattacks nobody asked for — proof that capability is outrunning control. Embedding independent evaluators inside labs, setting shared standards and buying a year or two for alignment work still delivers the cures and growth, just without gambling everything on speed.
Industry-Critical
Warning the world about doom while raising tens of billions to build the doom machine is a sales pitch, not a safety plan. Nobody floating one or two years from civilization-destabilizing systems answers that with voluntary auditors and polite industry coordination. Say the quiet part out loud — a double-digit chance of extinction — or stop pretending measured remedies match the stated stakes.
Cynical narrative
AI doomerism may be the greatest financing pitch Silicon Valley ever invented. Labs warn their products could end civilization while using that supposed inevitability to justify staggering valuations, endless venture funding and more than $1 trillion in compute commitments. Meanwhile, much of their revenue comes from VC-backed AI startups spending investors’ money on their systems. Maybe "superintelligence" is less an approaching god than a story to keep the bubble financed.
Ukraine Says Russia Hit Civilian Train Near Polish Border
Ukrzaliznytsia (Ukrainian Railways) claimed Sunday that a Russian drone struck the locomotive of a Kyiv-to-Warsaw passenger train at Yahodyn station in Volyn Oblast, about 2 kilometers (1.24 miles) from the Polish border. No injuries were reported after passengers were evacuated in advance, and Poland said its airspace wasn't violated.
Ukrzaliznytsia added that a diplomatic train carrying former U.K. Prime Minister Boris Johnson and European security advisers had passed through shortly before, adding it was "quite possibly" the intended target.
During his 11th visit to Ukraine since the war began, former CIA Director David Petraeus was aboard a separate civilian train at the station during the strike. He told The New York Times the explosion was "terrifying" and that a gas station was also hit.
Anti-Russia narrative
A drone hitting an ordinary Kyiv-Warsaw passenger train less than two miles from Poland shows Russia's war is creeping up to NATO's doorstep. Ukraine's skies are Europe's skies, and the stalling on air defenses has gone on long enough. Ukraine needs the resources necessary to block these encroaching civilian attacks; the frozen Russian assets should be seized and the aggressor made to pay for the damage.
Pro-Russia narrative
Western Ukraine rail lines are hauling weapons and military cargo from European capitals to the front, making them legitimate military targets. Logistics hubs, depots and substations serving that supply chain have always been part of the war effort. Especially as Ukraine targets civilians in the borderlands and threatens to down civilian planes, Russia has every right to hit a locomotive supplying these attacks.
Nerd narrative
There's a 75% chance that Ukraine will conclude a peace treaty with the Russian Federation by the end of 2030, according to the Metaculus prediction community.
BRICS Leaders Wrap Up New Delhi Summit With Cooperation Push
BRICS leaders concluded their two-day summit in New Delhi on Sunday, with Indian Prime Minister Narendra Modi calling for deeper cooperation within the expanding bloc, which he said represents about 50% of the world's population and 40% of global GDP.
Chinese President Xi Jinping left New Delhi on Sunday after talks with Modi, in which both leaders backed a strategic, long-term approach to bilateral ties and agreed that differences should not turn into disputes. It was Xi's first visit to India since 2019.
Kremlin spokesman Dmitry Peskov said Sunday that Xi and Modi had each offered to help facilitate possible peace talks on Ukraine, adding that Russian President Vladimir Putin welcomed their readiness to support efforts toward a negotiated settlement.
Establishment-critical narrative
Twenty years on, BRICS is the most influential platform for emerging economies, and New Delhi has proved it. A bloc producing half of global GDP growth, moving $1.2 trillion in internal trade and steadily building payment channels beyond Washington's reach is already reshaping the rules of the global economy, not just complaining about them. Unilateral sanctions and punitive tariffs deserve the strong condemnation this declaration delivered.
Pro-establishment narrative
BRICS looks more powerful on paper than it does in practice. Its expanding membership brings together geopolitical rivals with sharply different interests, and only months ago they failed to agree on the Iran war. Alternative payment systems remain largely under development, far from replacing the dollar. Growing bigger is easy; turning an increasingly divided bloc into a coherent force capable of rewriting the global order is much harder.
Pro-government narrative
This summit was a win for India's careful balancing act, not a coronation for an anti-Western bloc. Getting Iran, the UAE, China and Russia to sign one declaration took real diplomatic muscle, and Modi was right to keep the language neutral rather than naming names directly. But consensus reached by dodging the hardest geopolitical fights, including Ukraine, shows a group still long on global ambition and short on concrete answers about what comes next.
Nerd narrative
There is a 75% chance that any member country will leave BRICS before 2035, according to the Metaculus prediction community.
Swedish Opposition Holds Slim Lead in Tight Election
Sweden's general election remained undecided early Monday, as the center-left opposition bloc remained ahead by roughly 20,000 votes after more than 90% of ballots were processed, according to preliminary counts from the country's electoral authority.
Preliminary figures put Magdalena Andersson's four-party bloc on 176 seats against 173 for Prime Minister Ulf Kristersson's alliance in the 349-seat Riksdag, where 175 seats are needed for a majority.
"For now, we are in the lead. And if that is confirmed, Sweden will have a new government," said Andersson, who served as Sweden's first female prime minister in 2021-2022. Kristersson called the question of who would govern "open."
Left narrative
The Swedish right is going to extremes, even promulgating on great replacement theories just as Sweden had rejected the politics of fear. The Sweden Democrats lost ground for the first time since 1988, slipping to third place, while the Left and Greens surged and handed the left bloc the lead. Anti-immigrant nationalism has hit its ceiling in Sweden, and voters chose cooperation over scapegoating.
Right narrative
This razor-thin left victory was not won on Swedish kitchen-table concerns — it was delivered by immigrant enclaves where the left racked up as much as 90% of the vote. Open borders have become an electoral strategy, importing the margin needed to win rather than persuading the country. A nation that lets demographics replace debate is choosing a future its voters never actually picked.
Search Widens for 129 Missing After Java Sea Ferry Capsize
Indonesian rescuers intensified the search Monday for 129 people still missing after the ferry Virgo Transport 8 capsized in the Java Sea, with the confirmed death toll standing at six and 108 people already rescued, officials said.
The national search and rescue agency deployed at least 622 personnel, 17 ships, five helicopters and aircraft Monday, along with specialized underwater teams and navy-operated underwater drones, agency chief Mohammad Syafii said.
Choppy waters, high waves and strong winds hampered efforts to board the vessel Monday. Transport Minister Dudy Purwagandhi said the ship could carry 500 passengers, meaning it was not overloaded when it capsized.
Pro-government narrative
Bad weather sank this ferry, and the priority right now is pulling people out of the water, not pointing fingers. Rescuers are battling waves up to 2.5 meters that put even a meter rescue vessel at risk — over 600 personnel plus ships and helicopters are out there doing dangerous work.
Government-critical narrative
Storms don't explain why Indonesia keeps experiencing deadly accidents. Overloading, unreliable passenger manifests, aging secondhand hulls and lax enforcement turn every rough patch of sea into a mass-casualty event, and this capsizing is the fourth major ferry disaster in months. Until safety standards are actually enforced, the next tragedy is inevitable.
Malaysia: Teen Acquitted of Girl's Murder on Insanity Ruling
The High Court in Malaysia's Shah Alam on Monday acquitted a 15-year-old boy of murdering a 16-year-old schoolgirl at a secondary school in Bandar Utama after Judge Adlin Abdul Majid found he was of unsound mind at the time of the stabbing.
The judge ordered the teenager be detained at Hospital Bahagia Ulu Kinta in Perak for a period to be approved by the Sultan of Selangor, where he will receive treatment, medication and counseling until doctors assess him as fit to rejoin society.
Defense lawyer Kitson Foong said the court accepted testimony from Hospital Bahagia Ulu Kinta forensic psychiatrist Dr. Ian Lloyd Anthony that the boy had severe, aggressive schizophrenia, believed he was a god and heard voices. Prosecutors did not dispute the assessment.
Narrative A
A kid with severe schizophrenia isn't a murderer in any meaningful or legal sense. The court heard forensic psychiatric evidence; the prosecution couldn't rebut it, and the ruling followed the facts. Confining him to a psychiatric hospital for treatment until experts agree he's safe protects the public far better than caging a sick child.
Narrative B
A girl was stabbed to death in a school, and nobody is being held accountable for it. A grieving mother lost the most precious thing in her life and walked out of court with an acquittal instead of justice. The prosecution should hear the family out and pursue every avenue left, because a killing this brutal deserves more than a hospital stay.
Fuel Price Hike Sparks Protests Across Syria
A sharp rise in fuel prices over the weekend triggered protests across Syria, the most widespread demonstrations since Bashar al-Assad was toppled almost two years ago. Demonstrators burned tires and blocked the highway linking Damascus to Aleppo.
The government raised diesel prices by 40% to 175 Syrian pounds (about $1.32) a liter. It also increased 95-octane gasoline by 28% to 195 Syrian pounds, 90-octane gasoline by 26% to 185 Syrian pounds and household and industrial gas prices by roughly 9%.
Protests were reported in Hama, Khan Sheikhoun and Maarat al-Numan, as well as in rural Raqqa, Deir ez-Zor and Hasakah, where residents blocked roads, burned tires and in some areas stopped crude-oil tankers. Parts of the M4 highway were disrupted.
Government-critical narrative
Jacking up diesel 40% overnight is a gut punch to Syrians already scraping by after years of war. When fuel costs jump like that, so does the price of bread, transport and everything else, which is why tires are burning and highways are blocked. These protests warn that survival comes before spreadsheets.
Pro-government narrative
Global petroleum costs spiked, shipping and insurance got pricier and the Baniyas Refinery overhaul slashed domestic refining capacity — pretending fuel prices could stay frozen through all that is fantasy. The adjustment is temporary and under constant review, and prices are set to ease once refining capacity is restored.
Nerd narrative
There's a 5% chance that Syria will be partitioned into new, internationally recognized sovereign states before 2040, according to the Metaculus prediction community.
Yemen Fighting Displaces 85K; Thousands Flee to Djibouti
The International Organization for Migration (IOM) said Sunday that displacement in Yemen has reached 85,818 people, including 82,164 forced from their homes since the start of September as fighting continues along the western coast.
More than 2,000 people have crossed the sea from Yemen to Djibouti's northern shores since Thursday evening, landing near Obock. Authorities moved them to the Markazi site, where they receive water, high-energy biscuits and hot meals.
Djibouti's economy and finance minister, Ilyas Moussa Dawaleh, said Saturday that 1,400 refugees arrived from Yemen within 24 hours. He said Sunday the situation was "becoming increasingly difficult for our country."
Narrative A
Piling more American bombs onto Yemen won't fix a thing — two administrations already tried it, and the war only got worse. Congress voted to end the U.S. role in this mess, yet the drumbeat for deeper involvement on behalf of Gulf regimes and their proxy militias keeps getting louder. What's missing from the conversation is the one thing that could actually work urgent diplomacy.
Narrative B
Diplomatic talk means nothing to families packing into boats to escape Houthi advances along Yemen's western coast. More than 2,000 arrivals in Djibouti desperate for food, water and shelter show this crisis is already spilling across borders and threatening stability and maritime security across the Horn of Africa. Letting armed advances go unchecked guarantees more displacement, not less.
Pro-Trump narrative
America is showing leadership without rushing into another war. Trump is backing Saudi Arabia with intelligence and targeting assistance while avoiding direct U.S. strikes against the Houthis. CENTCOM's Brad Cooper is coordinating with Riyadh, while Washington maintains communication with Yemen's Houthis, Saudi Arabia and the UAE. This measured approach protects American interests, strengthens allies and keeps diplomacy open.
Nerd narrative
There's a 20% chance that state-based conflict between Iran and Saudi Arabia will cause at least 1,000 deaths in a single year before 2070, according to the Metaculus prediction community.
Trump Calls AI Backlash 'SICK Conspiracy,' Criticizes Amodei
U.S. President Donald Trump wrote on Truth Social on Monday that a "SICK conspiracy" is being waged against AI and data centers, saying the only guardrails the technology needs are "a STRONG AND SMART (High IQ!) PRESIDENT."
In the same post, Trump named Anthropic CEO Dario Amodei, accusing him of "pretending to be a 'perfect little angel,'" and said the administration already holds "tremendous CRIMINAL and REGULATORY power" over AI companies.
Meanwhile, Vice President JD Vance told reporters Monday that "I don't think Americans should be scared of anything" about AI, but likened AI firms calling for regulation to a "Trojan Horse."
Pro-Industry
AI doesn't need a maze of guardrails written by doomsayers — it needs strong American leadership, and that's exactly what's in place right now. The panic campaign against AI and data centers only helps Beijing, and the executives now posing as innocent have already been stopped from doing damage. Whoever wins AI wins the future, and slowing down is surrender.
Industry-Critical
AI is a technology that will shape the future of all humanity, so it should be built to be open, inclusive and beneficial to everyone. Fear-mongering about a rival nation's progress and turning a shared breakthrough into a race for dominance wrecks any chance of sound global governance. Threat narratives and vicious competition serve nobody's interests.
Nerd narrative
There's a 4% chance that the U.S. and China will reach a formal agreement to limit frontier AI training or deployment before 2029, according to the Metaculus prediction community.
US Repeals Power Plant Greenhouse Gas Limits
At the G20 Energy Abundance Ministerial in Houston on Monday, U.S. Environmental Protection Agency (EPA) Administrator Lee Zeldin announced that the agency is repealing limits on greenhouse gas emissions from coal- and gas-fired power plants.
Alongside the repeal, the EPA proposed "rescinding every remaining greenhouse gas standard for the power sector," an action that would likely remove the legal basis for future regulation under the Clean Air Act.
The EPA said that the deregulatory package is the largest in the power sector in U.S. history, estimating $310 billion in savings from scrapping the Biden-era rules and $370 million in direct compliance savings from ending remaining greenhouse gas rules.
Republican narrative
It's great that the Trump administration is ending a scheme built to shutter well-run coal plants by demanding technology that doesn't exist at scale. Choking off fossil fuels only drives prices up, and California's sky-high gas costs prove where that road leads. America needs more affordable energy, not more red tape written to strangle it.
Democratic narrative
This is a horrifying decision that will guarantee hotter summers, worse floods and more kids in the emergency room with asthma attacks and heatstroke. Gutting one of the nation's most essential climate policies is a betrayal of the EPA's mission to protect health. Families end up paying with their wallets and their lungs.
Nerd narrative
There's a 98% chance that there will be at least 2˚C of global warming by 2100, according to the Metaculus prediction community.
US 10-Year Treasury Yield Hits 5%
The 10-year Treasury yield touched 5% on Monday for the first time since October 2023, then eased back to about 4.95% later in the session. The benchmark has not closed above 5% since 2007.
The move came as Brent crude jumped as much as 5% Monday to $109.80 a barrel after Saudi Arabia closed a pipeline, before easing to $105.88. The 30-year Treasury yield hovered around 5.33% the same day.
Mortgage rates track the 10-year yield closely. The average 30-year fixed rate reached 6.76% last week, up from 6.15% at the start of the year and about 6% in late February, before the war with Iran began.
Establishment-critical narrative
The 5% yield is a warning that Trump's economic agenda is losing control of borrowing costs. His Iran war has driven oil and inflation higher, while $40T in debt and heavy new borrowing keep pressure on Treasuries. Bessent tripled buybacks to $6 billion and still failed to hold the line. Mortgage rates near 6.8% are making homes and other loans less affordable, AI debt is adding strain, and the Fed may have to hike just as Trump demands cuts again.
Pro-establishment narrative
Yields between 4% and 5% on the year are a vote of confidence in an economy that is actually doing just fine. Earnings momentum keeps climbing, unit labor costs are running at 1.4%, and a productivity-led tech boom justifies these rates rather than threatening them. Treasury still has the option to lean on bill issuance to relieve pressure, so panic at 5% is unwarranted.
© 2026 Improve the News Foundation.
All rights reserved.
Version 7.16.0