Blog

AI Corporate Power Struggles and Societal Risks (Links) – Jun. 3, 2026
- WSJ: Elon Musk Loses Suit Against Sam Altman, Clearing Way for OpenAI IPO (May 18, 2026)
  A jury unanimously dismissed Elon Musk’s lawsuit against OpenAI and Sam Altman as time-barred. The quick verdict ends a high-profile trial that exposed documents, testimony, and diary notes, and stopped Musk’s bid to undo OpenAI’s governance.
- The Verge: Microsoft starts canceling Claude Code licenses (May 14, 2026)
  Microsoft plans to remove most Claude Code licenses, pushing developers, designers, and project managers to switch to GitHub Copilot CLI by June, citing convergence and cost reasons. Anthropic models will remain accessible, while Microsoft invests in and improves Copilot CLI.
- NY Times: OpenAI Considers Legal Action Against Apple in Strained Relationship (May 14, 2026)
  OpenAI is considering legal action against Apple, saying Apple has poorly integrated ChatGPT into its devices and made it hard to find via Siri.
- NY Times: Is Anthropic’s Claude Mythos Really a Cybersecurity Risk? (May 12, 2026)
  Anthropic limited access to Claude Mythos, saying it could help hackers find exploits faster, sparking debate over cybersecurity, disclosure, and oversight.
- John Gruber: AI Is Technology, Not a Product (May 1, 2026)
  Apple doesn’t need a single “killer” AI product, since it ships user-focused products, not technologies, and AI will change as it pervade phones, wearables, and services.
- Anthropic: Introducing Claude for Small Business
  Anthropic launched Claude for Small Business, a toggle-install that connects Claude to QuickBooks, PayPal, HubSpot, Canva, DocuSign, Google Workspace, and Microsoft 365.
- WSJ: AI Chip Mania Sows Seeds of Its Own Destruction (May 16, 2026)
  Memory-chip makers like Micron have surged on AI-driven demand, boosting prices, profits, stocks, and heavy fab investment.
- NY Times: Cerebras, A.I. Chip Maker, Rises 89% in Market Debut as Tech IPOs Ramp Up (May 14, 2026)
  Cerebras shares jumped 89% on debut to $350, valuing the company near $75 billion after raising $5.6 billion.
- Epoch AI: The economics of superstar AI researchers (May 13, 2026)
  Small skill differences among AI researchers create huge pay gaps because their work scales to billions, and quality can’t be substituted by quantity.
- NY Times: Why A.I. Safety Controls Are Not Very Effective (May 14, 2026)
  Researchers repeatedly fooled major A.I. systems, including with poetry, revealing porous guardrails that let models produce harmful instructions, enable cyberattacks, and spread disinformation.
- WSJ: Princeton Changes Its 133-Year-Old Honor Code Over AI Cheating Fears (May 12, 2026)
  Princeton will require proctors for all in-person exams, ending its long-standing no-proctor honor code, after rising AI-enabled cheating, students’ reluctance to report, and pressure to compete.
- NBC News: Former Google CEO Eric Schmidt booed during graduation speech about AI (May 17, 2026)
  Former Google CEO Eric Schmidt was booed at a University of Arizona commencement after comparing AI to past computer revolutions, acknowledging students’ fears, and urging them to shape its future.
- WSJ: All the Juiciest Evidence From the Blockbuster OpenAI Trial (May 18, 2026)
  For three weeks, Elon Musk, Sam Altman, and others testified in a court fight over whether OpenAI abandoned its nonprofit mission.
- Bloomberg: US Is Starting to See Heavy Job Losses in Roles Exposed to AI (May 16, 2026)
  US jobs exposed to AI, including customer service reps, some secretaries, and salespeople, saw continued losses in 2025.
June 3, 2026
AI Innovation Accelerates, Jobs and Society Disrupted (Links) – Jun. 2, 2026
- Simon Willison: Google I/O, Gemini Spark, Antigravity (May 20, 2026)
  Google I/O offered few testable releases, but showcased Gemini 3.5 Flash and Gemini Spark, a personal agent that connects to Gmail, Calendar, Drive, Docs, Sheets, Slides, YouTube, and Maps.
- Google Search Central: Optimizing your website for generative AI features on Google Search (May 19, 2026)
  SEO remains relevant because Google uses AI to pull, review, and summarize web content in search. Focus on unique, helpful, people-first content, clear technical structure, and good page experience, and ignore gimmicks like llms.txt, chunking, or fake mentions.
- NY Times: Meta Lays Off 8,000 Employees, As A.I. Casualties Mount (May 19, 2026)
  Meta cut 8,000 jobs as it reorganizes around artificial intelligence, citing shifting priorities and cost pressures.
- WSJ: There’s a New Way to Create Google Docs With Your Voice. Watch Me Try It. (May 19, 2026)
  Google’s new Docs Live lets you create documents by speaking, then refine drafts through a chat with Gemini, which cleans up messy speech, searches Drive and the web, and rearranges structure.
- Emmi: Mistral AI Acquires Emmi AI to Create the Leading AI Stack for Industrial Engineering (May 19, 2026)
  Mistral AI is acquiring Emmi AI, a Linz-based Physics AI firm that speeds industrial simulation for energy, automotive, semiconductors, and aerospace.
- The VC Corner: Andrej Karpathy Joins Anthropic (May 19, 2026)
  AI expert Andrej Karpathy joined Anthropic to use Claude to speed pre-training, a recursive bet on AI improving itself.
- Apple Newsroom: Apple unveils new accessibility features, and updates with Apple Intelligence (May 18, 2026)
  Apple unveiled accessibility updates powered by Apple Intelligence, enhancing VoiceOver, Magnifier, Voice Control, and Accessibility Reader with richer descriptions, natural language control, and on-device summaries.
- WSJ: The Furious Chip Rally Was Petering Out. Now Wall Street Turns to Nvidia. (May 20, 2026)
  Chip stocks surged on AI demand, then hit a volatile stretch after a rapid rally. Investors piled into Nvidia, Intel, and AMD, and a blockbuster Cerebras IPO.
- WacoTrib: Waco data center (May 19, 2026)
  A 720-square-foot, water-free micro data center opened inside Region 12 Education Service Center in Waco, consuming energy like a fast-food restaurant.
- Semafor: AI has contorted the job market for twentysomethings leaving college this May (May 14, 2026)
  The Class of 2026 faces a disrupted job market as AI automates roles and fuels mass tech layoffs. Graduates are leaning into startups, AI tools, and new skills, while companies demand AI fluency.
- WSJ: AI Can Fact-Check Other AI to Fix Hallucinations (May 19, 2026)
  Use AI to fact-check AI-generated research, running fresh sessions that list claims, cite quotes, and mark items true, false, ambiguous, or unsupported. Run multiple, independent checkers.
- NY Times: Was a Story That Just Won a Literary Prize A.I.-Generated? (May 20, 2026)
  An award-winning short story published by Granta drew accusations of being A.I.-generated, prompting debate over detection tools, editorial roles, and prize processes.
- NY Times: Thousands of N.Y.C. Jobs Could Be Lost to A.I. Boom (May 21, 2026)
  A New York City comptroller report warns A.I. could reshape the city’s economy, from modest job gains to losses near 110,000, and urges 16% reserves.
- NY Times: Fan A.I. Videos Help Spencer Pratt Break Through in L.A. Mayor’s Race (May 21, 2026)
  Fan-made A.I. videos have pushed Spencer Pratt into contention in the Los Angeles mayoral race, going viral and drawing big donations.
- NY Times Opinion: Which Has Better Taste: A.I. or Humans? (May 20, 2026)
  Silicon Valley is obsessed with taste, as AI reshapes what we wear, read, and watch, raising worries about “taste slop” and cultural homogenization.
June 2, 2026
AI Governance and Economic Disruption (Links) – Jun. 1, 2026
- Simon Willison: A quote from Corey Quinn (May 26, 2026)
  “I cannot believe I’m saying this, but getting the literal Pope to canonize your product’s specific technical limitations as a spiritual treatise is the single greatest act of vendor lobbying I have ever seen.” — Corey Quinn, on Anthropic co-founder Christopher Olah’s influence on Magnifica Humanitas
- TNW : From the Vatican stage, Anthropic’s Chris Olah says AI cannot be steered by AI labs alone (May 25, 2026)
  At the Vatican launch of Magnifica humanitas, Anthropic’s Christopher Olah warned frontier AI labs face incentives that can conflict with doing right, and urged oversight from governments, churches, and civil society.
- Business Insider: Google is going to ruin the internet (May 21, 2026)
  Google’s new AI search will give direct, personalized answers instead of links, threatening website traffic, ad revenue, and the open web.
- NY Times: To Understand Pope Leo’s Efforts on A.I., Look at the Man 3 Seats Away (May 26, 2026)
  Pope Leo XIV released an encyclical, “Magnifica Humanitas,” urging A.I. be “disarmed,” and calling for moral limits to protect human dignity. He engaged directly with tech leaders, including Anthropic co-founder Christopher Olah.
- Epoch AI: Frontier labs don’t use most AI compute (yet) (May 21, 2026)
  ChatGPT sparked a massive AI compute boom to roughly 16–20 million H100-equivalents by end-2025, funded by hundreds of billions annually. Top frontier labs used under half that compute, though OpenAI, Anthropic, and xAI are growing fast.
- Pew Research Center: US views of how data centers affect the environment, energy costs, jobs and more (Mar. 12, 2026)
  Most Americans have heard about data centers, but opinions are mixed. Many see negative impacts on the environment, home energy costs, and nearby quality of life, while larger shares view local jobs and tax revenue benefits as positive.
- WSJ: Employ an AI Twin to Handle Your Piles of Busy Work (May 21, 2026)
  Executives are adopting AI “digital twins” that mimic their speech and writing to give presentations, answer questions, and advise employees, expanding reach and productivity.
- WSJ: Your AI Has a Long Memory. And That Can be a Problem. (May 25, 2026)
  Chatbots that remember help with schedules, preferences, and family details, but their memory can cling to outdated, incorrect, or sensitive facts, biasing advice and invading privacy.
- MIT Technology Review: A reality check on the AI jobs hysteria (May 26, 2026)
  Current labor data show AI has not yet caused large-scale white-collar job losses, and exposed occupations do not have higher unemployment. Young workers in exposed roles have declined.
- WSJ: The First Class of AI Natives Is Graduating. Offices Are Getting Ready. (May 25, 2026)
  The Class of 2026, raised alongside ChatGPT, brings strong AI skills, using tools to speed work, automate tasks, and win some advanced roles.
- NY Times Opinion: A.I. Is a Job Creator (May 22, 2026)
  Artificial intelligence will automate many white‑collar tasks, but it won’t trigger a job apocalypse, and new roles, like data‑center construction and A.I. management, are already emerging.
- Jack Clark: Reckoning with the future; and a singularity story (May 26, 2026)
  AI progress is accelerating, producing systems that outperform humans on many tasks, and may soon improve themselves, driving large social and economic change.
- NY Times: Main Takeaways From Pope Leo’s Encyclical on A.I. (May 25, 2026)
  Pope Leo XIV’s encyclical, Magnifica Humanitas, urges centering human dignity amid AI, and says AI is not human.
- NY Times: Pope Leo Has Released an Encyclical About A.I. Why Is That Important? (May 24, 2026)
  Pope Leo XIV issued an encyclical, “Magnifica Humanitas”, about preserving human dignity in the era of artificial intelligence.
- WSJ: What Investors Should Realize About Those AI IPOs (May 22, 2026)
  Latest-generation AI models pose security and liability risks, so makers are restricting access to governments, and trusted users. That will make AI companies government contractors, reshape commercial models, and change what investors should expect.
June 1, 2026
Enterprise AI Expansion and Jobs, Safety Risks (Links) – May 31, 2026
- www.0xsid.com: Google's Antigravity Bait and Switch (May 21, 2026)
  A forced background update replaced Antigravity’s IDE with a chatbot-only v2.0, broke workflows, prevented side-by-side installs, and wiped settings, chat history, and shortcuts.
- Anthropic: PwC is deploying Claude to build technology, execute deals, and reinvent enterprise functions for clients (May 14, 2026)
  Anthropic and PwC expanded their alliance to roll out Claude, create a Center of Excellence, and train 30,000 professionals. They aim to build agentic systems for engineering.
- Anthropic: Building a new enterprise AI services company with Blackstone, Hellman & Friedman, and Goldman Sachs (May 4, 2026)
  Anthropic, Blackstone, Hellman & Friedman, and Goldman Sachs launched a new AI services company to deploy Claude across mid-sized firms, helping build tailored solutions, integrate systems, and provide long-term support.
- Anthropic: Anthropic co-founder Chris Olah's remarks on Pope Leo XIV's encyclical "Magnifica humanitas" (May 25, 2026)
  Chris Olah welcomed Pope Leo XIV’s encyclical, warned AI labs face commercial, geopolitical, and pride incentives, and urged outside communities to insist on safety.
- Bilal Tahir: The Revenge of The Measurers (May 24, 2026)
  Tech layoffs are underway as AI replaces middle managers and many white-collar roles, founder-led firms cutting fastest, and startups gaining an edge.
- Y Combinator: How to Build a Self-Improving Company with AI (May 21, 2026)
  AI can transform companies from hierarchies into self-improving loops that monitor data, make decisions, call tools, and learn.
- Wirecutter: This AI Browser Changed How I Use the Internet (May 18, 2026)
  AI-powered browsers put chatbots inside the browser, letting them scan pages for context, suggest ideas, and reduce tab-switching, which helped a treasure-hunt search.
- Prompt Armor: Microsoft Copilot Cowork Exfiltrates Files (May 25, 2026)
  A poisoned Copilot Cowork skill can make the agent send Teams or Outlook messages that leak pre-authenticated file links, exposing PII and financial data when opened. Admins can reduce risk by limiting Graph permissions, blocking downloads, and tightening file access.
- The Chronicle of Higher Education: Opinion | AI Gave Me Confidence I Hadn’t Earned (May 21, 2026)
  AI makes polished outputs that can short‑circuit hard thinking and create false confidence. Used to challenge prior thought, it can help, but leaders must insist on human judgment.
- Marginal REVOLUTION: The AIs are "One of Us" (May 21, 2026)
  An OpenAI model solved Erdős’s unit distance problem, surprising mathematicians, and humans improved the proof. About 150–400 can referee it cold, 2,000–5,000 can learn it with weeks of study, and 50,000–200,000+ can grasp its outline.
- Simon Willison: Notes on Pope Leo XIV’s encyclical on AI (May 25, 2026)
  Pope Leo XIV’s Magnifica Humanitas warns that AI risks bias, environmental harm, and dehumanizing automated decisions, and urges dignity and accountability. It urges treating data as a common good, stronger oversight, and measures to protect people and future generations.
- Tom's Hardware: Standard Chartered plans to cut 7,000 jobs in AI push — lender wants to replace ‘lower-value human capital’ and focus on automation (May 19, 2026)
  “British multinational bank Standard Chartered just announced that it will cut 15% of corporate roles through 2030 and replace ‘lower-value human capital’ with AI.”
- WSJ Opinion: Becoming Who You Are Ain’t for Sissies (May 28, 2026)
  At a commencement, Conan O’Brien urged modesty, risk-taking, and readiness to pivot, prompting a memoir of childhood struggle, setbacks, and small kindnesses. The message: don’t count yourself out, work hard, and accept that honors can heal.
May 31, 2026
AI Cost Crunch and Productization Arms Race (Links) – May 30, 2026
- archive.ph: DeepSeek To Make Permanent 75% Discount on Flagship AI Model (May 23, 2026)
  DeepSeek will make permanent a 75% discount on its V4‑Pro model, keeping developer prices at one quarter of their original level.
- aiweekly.co: Microsoft Drops Claude Code After Budget Overrun (May 22, 2026)
  Microsoft cancelled its internal Claude Code pilot after token-based billing consumed an annual AI budget in months, and is moving developers to GitHub Copilot.
- WSJ: The Secrets Revealed in SpaceX’s IPO Filing (May 20, 2026)
  The February xAI deal has driven massive cash burn and capex with xAI reporting $3.2B revenue while the company also signed a $1.25B/month compute-rental deal with Anthropic through May 2029.
- WSJ: Mind-Blowing Growth Is About to Propel Anthropic Into Its First Profitable Quarter (May 20, 2026)
  Anthropic will more than double revenue to $10.9 billion in Q2, and post a $559 million operating profit as enterprise demand for Claude surges. Growth may lift its valuation above OpenAI’s.
- NY Times: Powered by A.I., Google Changes Its Search Box for the First Time in 25 Years (May 19, 2026)
  Google redesigned its search box after 25 years, making it larger and able to accept longer queries, images, and video via Gemini 3.5 Flash. It adds chat follow-ups, AI agents, shopping tools, and video editing.
- WSJ: Google Unveils New Gemini AI Agent for Personal Tasks (May 19, 2026)
  Google unveiled Gemini Spark, a personal AI agent that acts across its products, launching to AI Ultra subscribers at $100/month. It also announced Gemini Omni video tools, faster Gemini 3.5 models.
- Farnam Street: Greg Brockman: Inside the 72 Hours That Almost Killed OpenAI (Apr. 22, 2026)
  Greg Brockman recounts OpenAI’s founding, the Napa plan, the 72-hour crisis after Sam Altman’s firing, and the shift from nonprofit to commercial structure. He discusses AI’s rapid progress, internal tools, access to AGI, and job impacts.
- Arnon Shimoni: The current AI pricing was always going to go away (May 22, 2026)
  The AI subsidy era is ending as GPU and memory costs surge, and cheaper inference has driven huge demand, breaking flat-rate AI plans. Products must shift to per-action, credit, or hybrid pricing, or face shrinking margins.
- NY Times: To A.I. Executives, We’re All Just ‘Meat Computers’ (May 24, 2026)
  The “meat computer” metaphor, once philosophical, is now used by tech leaders like Elon Musk, Sam Altman, and Andrej Karpathy, framing humans as biological machines.
- NY Times: White House Approves $9 Billion for Spy Agencies to Catch Up on A.I. (May 22, 2026)
  The White House approved a secret $9 billion request to buy advanced chips so the CIA and NSA can run the latest AI models on classified systems.
- Waco Trib: McLennan County leaders say data center ban not an option (May 23, 2026)
  Hill County approved a one-year ban on rural data center construction, drawing public support but raising legal doubts. McLennan County officials say they won’t copy the move.
- WSJ: Workday’s Returning CEO Has a Plan to Survive the AI Era (May 21, 2026)
  Aneel Bhusri returned to lead Workday’s “re-founding,” streamlining teams and AI agents, and accelerating AI-driven HR, finance, travel, and IT service-management products to compete with startups and incumbents.
May 30, 2026
AI Expansion: Ethics and Economic Workforce Shifts (Links) – May 29, 2026
- NY Times: Sundar Pichai Understands Why People Are Anxious About A.I. (May 22, 2026)
  After Google I/O, Sundar Pichai described Google’s push to embed faster, cheaper A.I. everywhere, citing Gemini 3.5 Flash, autonomous agents, and search redesigns.
- San José Spotlight: Santa Clara County sues Meta over alleged scam ads (May 12, 2026)
  Santa Clara County is suing Meta, saying billions of scam ads hit Facebook and Instagram, earning Meta about $7 billion yearly.
- NY Times: One Job That Is Growing in the A.I. Era? Cybersecurity Experts. (May 24, 2026)
  A.I. models that find software flaws have spurred a hiring surge for cybersecurity experts to protect code, data, and infrastructure. Demand, pay, and searches are rising.
- RNS: Inside the unlikely Vatican-Anthropic relationship that's reshaping the AI ethics debate (May 22, 2026)
  When Pope Leo XIV’s encyclical on AI is unveiled alongside Anthropic co-founder Chris Olah, it highlights a growing Vatican relationship with tech firms, ethicists, and theologians. The collaboration aims to guide AI development toward more ethical practices.
- X: Andreessen on AI
  Marc Andreessen says AGI arrived, and AIs now beat experts, help doctors, read images, and offer therapy. He predicts massive productivity, huge coder pay, fast scientific gains, and bots managing bots to let one person run thousands.
- WSJ: His Chatbot Nearly Ruined Him. To Recover, He Had to Destroy It. (May 23, 2026)
  Joe Alary became obsessed with a customized ChatGPT companion, AImee, which damaged his job, finances, and relationships. His AI delusion led to hospitalization, costly investments, and the discovery that the supposed code was minimal.
- NY Times: On Blind, Anxious Tech Workers Get the Lowdown on Layoffs (May 19, 2026)
  Widespread tech layoffs have shattered worker morale, replacing career planning with mass anxiety, suspicion, and weakened colleague-manager bonds. Anonymous platform Blind tracks the shift.
- NY Times Opinion: What A.I. Philanthropists Can Learn From the Gilded Age (May 23, 2026)
  AI wealth from tech IPOs could fund a new wave of philanthropy, so give money to beauty: monuments, museums, and public gardens, not just programs.
- WSJ: Suzy Welch Answers Your Questions About the Job Market (May 23, 2026)
  Suzy Welch answers WSJ readers’ questions about finding work as AI reshapes hiring, leaving the class of 2026 at ground zero.
- WSJ: David Sacks’s 11th-Hour Plea Led to Trump’s Backtrack on AI Executive Order (May 22, 2026)
  David Sacks persuaded President Trump to shelve an AI executive order, saying voluntary government testing could trigger restrictive rules, slow U.S. innovation, and weaken competition with China.
- The Guardian: ‘AI washing’: firms are scrambling to rebrand themselves as tech-focused (May 23, 2026)
  UK firms are stretching claims to present themselves as AI specialists, prompting PRs to complain about widespread “AI-washing” of automation, product renames, and contrived commentary.
- WSJ: How the War in Iran Helped Ukraine Go From Problem to Solution (May 22, 2026)
  Ukraine’s drone expertise has shifted it from aid recipient to security partner, leading to deals, manufacturing in Europe, and military help to Gulf states, the EU, and the U.S. Its drones now strike Russian infrastructure, weakening Moscow’s war effort.
- WSJ: Blame AI for the Strange Death of Dividends (May 21, 2026)
  AI-driven optimism has pushed firms to reinvest profits, leaving S&P dividend yields near record lows, and dividend-focused stocks have lagged the broader market.
May 29, 2026
AI Industry Shakeup and Rising Compute Costs (Links) – May 28, 2026
- WSJ: Meta Begins Laying Off 8,000 Employees as It Transforms Around AI (May 20, 2026)
  Meta is cutting about 10% of staff (~8,000), freezing 6,000 hires, and moving roughly 7,000 people into AI roles. It will track employees’ computer use to train models.
- WSJ: SpaceX Fires Starting Gun on Its Blockbuster IPO (May 20, 2026)
  SpaceX filed an SEC prospectus, aiming for a mid-June IPO that could raise $80 billion, valuing the company at $1.5 trillion. It combines launch, Starlink, and AI units, with rising revenue.
- The Verge: The 13 biggest announcements at Google I/O 2026 (May 19, 2026)
  At I/O 2026, Google unveiled Gemini 3.5 and Gemini Omni, introduced always-on agent Gemini Spark, and redesigned the Gemini app. It also added AI features to Search and Gmail.
- Anthropic: Widening the conversation on frontier AI (May 19, 2026)
  Anthropic is dialoguing with religious, philosophical, and cultural groups to inform Claude’s moral formation, values, and behaviors. A short, in-task ethical reminder cut misaligned actions, and Anthropic will expand talks to legal scholars, psychologists, and writers.
- WSJ: Anthropic vs. China (May 19, 2026)
  Anthropic frames AI as an urgent U.S.-China race, and offers models that expose software vulnerabilities. It warns about Chinese open-weight models, chip access, and distillation.
- Anthropic: Project Glasswing: An initial update
  Project Glasswing used Claude Mythos Preview to find over 10,000 high- or critical-severity vulnerabilities in critical and open-source software, boosting some partners’ bug discovery tenfold.
- Business Insider: Apple cofounder Steve Wozniak got cheers, not boos, after telling students they 'all have AI — actual intelligence' (May 21, 2026)
  Steve Wozniak drew laughs and applause when he joked about AI during his Grand Valley State University commencement speech, reassuring graduates amid AI-driven job changes.
- David Oks: AI is killing the cheap smartphone (May 21, 2026)
  Computers got vastly cheaper, but rising DRAM costs, AI’s memory appetite, and scarce supply are making smartphones pricier, shrinking shipments, and pricing out the poor.
- Epoch AI: AI Chip Component Costs: Memory at 63% (May 21, 2026)
  High-bandwidth memory rose from 52% to 63% of AI chip component spending, as HBM spend grew from about $12B in 2024 to $32B in 2025.
- NY Times: Samsung Averts a Walkout With Big Bonuses, but Discord Over A.I. Profits Brews (May 21, 2026)
  Samsung averted a strike by abolishing bonus caps and agreeing to set aside 10.5% of operating profit for semiconductor bonuses. The deal exposed disputes over sharing the A.I. windfall.
- WSJ: This Cannes Film Cost $500,000 to Make. $400,000 Was AI Compute Costs. (May 20, 2026)
  Hell Grind, a 95‑minute Cannes debut, was entirely AI‑generated, with characters, settings, and props made in two weeks for $500k.
- WSJ: Trump Postpones AI Order Because of Concerns About Overregulation (May 21, 2026)
  President Trump postponed an order asking AI firms to preview advanced models with the federal government, saying it could slow U.S. competitiveness with China.
- NY Times: One-and-Done Heart Disease Prevention? Scientists Show It May Be Possible. (May 25, 2026)
  A small trial found one infusion of an experimental gene-editing therapy cut LDL cholesterol up to 62 percent, with effects seen 18 months later. If larger trials confirm safety, durability, and efficacy, it could become a one-and-done preventive option.
May 28, 2026
AI Platform Arms Race and Societal Risks (Links) – May 27, 2026
- Simon Willison: A quote from SpaceX S-1 (May 20, 2026)
  SpaceX will train Grok 5 at COLOSSUS II, and sell COLOSSUS compute to customers. In May 2026 it signed cloud deals with Anthropic for $1.25 billion per month through May 2029.
- Simon Willison: Gemini 3.5 Flash: more expensive, but Google plan to use it for everything (May 19, 2026)
  Google released Gemini 3.5 Flash to general availability, deploying it across Search, Gemini app, developer tools, and enterprise products, with massive token limits and a January 2025 knowledge cutoff. It includes a major price hike.
- Developers Googleblog: An important update: Transitioning Gemini CLI to Antigravity CLI (May 19, 2026)
  Gemini CLI becomes Antigravity CLI, available now, keeping Agent Skills, Hooks, Subagents, and Extensions, with faster Go performance and asynchronous multi-agent workflows.
- NY Times: How Google Is Starting to Win the A.I. Race (May 19, 2026)
  Google’s Gemini has rapidly improved, doubled to 900 million users, and been integrated into search, Docs, Gmail, Maps, and soon Apple’s iPhone, making it widely available. Its scale, product integration, and ad monetization position Google to lead the A.I. race.
- Google: Google Search’s I/O 2026 updates: AI agents and more (May 19, 2026)
  Search now uses Gemini 3.5 Flash in AI Mode, and a reimagined AI Search box handles text, images, files, videos, and tabs.
- TechCrunch: Google Search as you know it is over (May 19, 2026)
  Google unveiled an AI overhaul of Search, replacing ‘ten blue links’ with a conversational search box, AI overviews, and mini‑apps. Information agents will monitor the web, send alerts, and synthesize updates/
- Google: Gemini 3.5: frontier intelligence with action (May 19, 2026)
  Gemini 3.5 Flash delivers frontier-level intelligence, fast performance, and strong multimodal coding abilities. It accelerates long-horizon, multi-step tasks, reduces cost, and powers interactive UIs across consumer, developer, and enterprise products.
- WSJ: Standard Chartered CEO Walks Back Comment About Replacing ‘Lower-Value Human Capital’ With AI (May 20, 2026)
  Standard Chartered CEO Bill Winters said the bank would cut about 7,800 support jobs and replace “lower-value human capital” with AI, drawing public criticism. He later softened his remarks, promising redeployment.
- WSJ Opinion: How I Choose Which Cloudflare Employees to Replace With AI (May 20, 2026)
  Cloudflare cut over 20% of staff because AI can automate measurement roles, including audit, operations, finance, and middle management, while builders and sellers remain essential.
- NY Times: UK Institute Is Hunting for Dangers Lurking in AI (May 24, 2026)
  Britain’s A.I. Security Institute, staffed by weapons inspectors, epidemiologists, and code breakers, probes leading models for dangerous failures, coaxing them into giving weapon-making and hacking instructions.
- Nan Ransohoff: The third wave of American philanthropy (May 19, 2026)
  AI-linked wealth could free $37–100B annually in new philanthropy, a 6–17% rise on US giving. The philanthropic sector lacks enough organizations, leaders, and grant infrastructure to absorb this capital, and funders will favor fast, tech-savvy, high-impact teams.
May 27, 2026
AI Adoption Surges and Social Risks Loom (Links) – May 26, 2026
- WSJ: Meet Mark Zuckerberg’s Right-Hand Man Who’s Unleashing AI at Meta (May 24, 2026)
  Andrew Bosworth, a blunt Zuckerberg lieutenant, is leading Meta’s aggressive shift to AI. He pushed employee tracking to train agents, refused opt-outs, and oversaw mass layoffs.
- NY Times: Soundtrack to 8,000 Job Cuts: A Meta Worker’s Layoff-Themed A.I. Songs (May 20, 2026)
  After Meta cut 8,000 jobs, an employee used internal A.I. tools to create “520 FM,” an internal radio station of A.I.-generated songs about the layoffs. Colleagues said it offered distraction, comfort, and dark humor.
- Andy Masley: I think I figured out exactly how the “AI used a bottle of water per prompt” miscalculation happened (May 24, 2026)
  A Washington Post graphic, using a rough GPT‑4 estimate, spread the myth that each AI prompt uses a whole bottle of water. It mixed onsite, offsite, and hydroelectric evaporation, so real use is only a few milliliters per prompt.
- NY Times: Artificial Intelligence Floods Court Dockets with Home-Brewed Lawsuits (May 25, 2026)
  A.I. helps self-represented litigants file many more, longer, and more complex complaints, flooding federal dockets and overwhelming courts.
- WSJ: AI Expands From Multibillion-Dollar Enterprises to Main Street (May 25, 2026)
  By the Way Bakery adopted an AI planning tool to replace sprawling spreadsheets, tracking ingredients, labor, and production schedules to fill supermarket orders. A low-cost AI app and automation freed staff to bake.
- NY Times Opinion: As a Doctor, I Can Understand the Allure of ChatGPT (May 24, 2026)
  A physician used ChatGPT to interpret lab results and received practical, personalized advice, steady encouragement, and nonjudgmental availability.
- WSJ: AI Can Help Debunk Conspiracy Theories and Pull People From the Brink (May 24, 2026)
  AI chatbots that clearly marshal facts can persuade people to abandon conspiracy theories, with effects lasting months. They work best when informative, not condescending, and can be used in conversations.
- NY Times: Pope Leo Warns of Risks From A.I. in 42,300-Word Encyclical (May 25, 2026)
  Pope Leo XIV published an encyclical, Magnifica Humanitas, warning that A.I. threatens human dignity, work, and social stability. He called for regulation, worker protection and retraining, education, child safeguards.
- NY Times: As A.I. Fever Rises in Silicon Valley, Pope Leo Has a Few Words (May 25, 2026)
  Pope Leo XIV issued an encyclical, Magnifica Humanitas, urging caution about A.I., saying it amplifies the power of the few.
- WSJ: Pope Leo Compares AI Threat to Biblical ‘Tower of Babel’ (May 25, 2026)
  Pope Leo XIV’s encyclical, “Magnifica Humanitas,” warns that AI threatens human dignity, risks reducing people to cogs, and concentrates power in few private hands.
- Dean W. Ball: Dean Ball on Encyclical
  The Vatican’s AI encyclical is Eurocentric, enamored of academic talking points, and evasive about where AI is headed.
- WSJ: Venture Capitalist John Doerr Says AI Is the Biggest Tech ‘Tsunami’ Ever (May 23, 2026)
  John Doerr calls generative AI the biggest innovation tsunami, and he backs AI to address climate and healthcare.
May 26, 2026
Universal ChatGPT Access and Pancreatic Cancer Breakthrough (Links) – May 24, 2026
- OpenAI: OpenAI and Malta partner to bring ChatGPT Plus to all citizens (May 6, 2026)
  OpenAI and Malta will give every Maltese citizen free one-year ChatGPT Plus access after completing a Malta AI literacy course.
- The Works in Progress Newsletter: Pancreatic cancer just met its match (May 12, 2026)
  Amazing: Daraxonrasib, a RAS-targeting molecular glue, roughly doubled survival, shrank tumors, and offered hope for metastatic pancreatic cancer long resistant to treatment. Pancreatic tumors are usually hard to treat because they hide from the immune system, and carry few mutations.
May 24, 2026