Category: AI

AI Infrastructure and Governance (Links) – Jul. 9, 2026
- WSJ: SpaceX Showed Investors Prototype of Elon Musk’s New AI Device (Jul. 1, 2026)
  SpaceX showed investors an early-stage project that could anchor Musk’s super-app vision, connect to Starlink, and reduce reliance on Apple, but production is uncertain.
- NY Times: U.S. Lifts Restrictions on Anthropic’s Most Powerful A.I. Models (Jun. 30, 2026)
  The Commerce Department lifted limits on Anthropic’s Claude Mythos and Claude Fable models, letting the company restore its top AI systems. Anthropic agreed to risk detection, government coordination, and reporting of malicious activity, easing tensions with the Trump administration.
- Simon Willison: Nano Banana 2 Lite (Jun. 30, 2026)
  Nano Banana 2 Lite (Gemini 3.1 Flash Lite) is a fast, cheap, and scalable image model.
- Anthropic: Introducing Claude Sonnet 5 (Jun. 30, 2026)
  Claude Sonnet 5 is a more agentic, capable model that plans, uses tools, and runs autonomously, and rivals Opus 4.8 on some tasks while improving on Sonnet 4.6 in coding, reasoning, and safety, but it lacks Opus-level cybersecurity skill.
- Anthropic: Redeploying Claude Fable 5
  Export controls suspended Claude Fable 5 and Mythos 5; controls were lifted June 30, and Fable 5 returns July 1 with limited access. Anthropic added safeguards, a classifier, and government collaboration to block the reported bypass.
- WSJ: AI Data Centers Use Far More Water Than Most Tech Giants Report (Jul. 3, 2026)
  Big tech’s AI data-center buildout consumes vast water, often far more indirectly via power generation than companies report. Lack of standard reporting, regional differences, and reliance on energy offsets risk local water stress.
- CNBC: Nvidia offers start-up customers chance to swap compute power for revenue share (Jul. 2, 2026)
  Nvidia is striking revenue-sharing deals with fast-growing AI startups, trading compute access, via token credits, for a share of product and cloud revenue.
- WSJ: Michael Burry Cites ‘Beginning of the End’ With New AI Short Bets (Jul. 1, 2026)
  Michael Burry boosted shorts against AI, buying puts on Tesla, Caterpillar, Applied Materials, Nvidia, and a chip ETF, betting AI hype will collapse.
- The Texas Tribune: Report shows Texas to lead nation in data center power plants (Jul. 1, 2026)
  Texas would host 32 of 74 proposed natural gas power plants to directly fuel data centers, potentially emitting massive greenhouse gases and air pollutants.
- WSJ: Wealthy Families Swap Traditional Classrooms for AI Tutors and ‘Alternative’ Schools (Jul. 3, 2026)
  Wealthy parents are choosing K–8 schools that mix AI tutors, project-based work, and entrepreneurship to personalize learning.
- NY Times Opinion: It’s Inevitable — You’re Going to Get Hacked. But Here’s One Solution. (Jul. 2, 2026)
  AI makes account hacks inevitable, exposing iCloud, Gmail, and other personal data to scams, voice cloning, and stolen two-factor codes. We must stop punishing private, lawful slips, and practice basic security: unique passwords, multifactor authentication, and never sharing codes.
- NY Times: Why A.I. Distillation Has Become a Hot Topic in the Race with China (Jul. 6, 2026)
  U.S. AI companies say Chinese rivals rapidly copy their models using distillation, often via thousands of unauthorized accounts, shrinking America’s lead. The technique is established, legally unclear, and fueling trade-secret, intellectual-property, and national-security concerns.
- WSJ: The Quest to Make Humanoid Robots Safe Enough for Humans (Jul. 4, 2026)
  Humanoid robots are entering workplaces, prompting safety worries after viral mishaps, as they grow heavier and use probabilistic AI. Companies and standards bodies are adding layers of safeguards.
- NY Times: Can Companies Embrace A.I. Without Layoffs? This One Says It Is Trying to. (Jul. 2, 2026)
  SAP is trying to adopt A.I. while keeping workers by having employees reinvent jobs, shifting routine tasks to bots, and creating new roles.
- Moz: Stop Calling It SEO: 12 Strategies To Dominate AI Search in 2026 (Jun. 30, 2026)
  Google’s AI guidance is self-serving, applies mainly to Google, not to ChatGPT, Perplexity, Claude, or other AI systems. AI search needs a separate strategy, not just traditional SEO, because treating it the same limits innovation, budget, tools, and ownership.
July 9, 2026
AI Dev Tools Boom and Governance Risks (Links) – Jul. 8, 2026
- Google Cloud: Review GitHub code using Gemini Code Assist (Jul. 3, 2026)
  Gemini Code Assist on GitHub is an automated reviewer that summarizes pull requests, gives in-depth code reviews, and speeds up reviews, improving code quality.
- Simon Willison: Fable's judgement (Jul. 3, 2026)
  A Fireside Chat suggested letting Fable and Opus use their own judgement for tasks like testing, instead of strict rules. Claude Code stored a memory delegating coding to lower‑power subagents, with the main model reviewing, reducing token use, accelerating work.
- Mistral AI: Leanstral 1.5: Proof Abundance for All (Jul. 2, 2026)
  Leanstral 1.5 is an open model that “verifies complex code properties and uncovers previously unknown bugs in open-source repositories.”
- WebKit: Introducing the Safari MCP server for web developers (Jul. 1, 2026)
  Safari Technology Preview 247 adds a Safari MCP server that lets local agents connect to a Safari window, inspect DOM, network, screenshots, and console.
- WSJ: Anthropic Reaches Deal With Trump Administration to Restore Access to Fable AI Model (Jun. 30, 2026)
  Anthropic agreed with the Trump administration to restore access to Fable 5 after fixing Amazon-flagged jailbreaks, adding a safeguard, and redirecting risky prompts. The deal ends a 2½-week shutdown and signals tighter government oversight.
- WSJ Opinion: AI and the English Language (Jul. 2, 2026)
  Orwell warned against clichés, and AI can flag them, yet it recombines familiar phrases at scale into polished, soulless prose. People may accept popular AI content, but they will still crave original, human voice, honesty, and experience.
- WSJ: The World’s Top Economists Are Sounding the Alarm on AI (Jul. 1, 2026)
  At a Sintra symposium, central bankers said AI could boost productivity, but also fuel rising debt, investor leverage, stretched valuations, and job losses. They warned AI-driven cyber flaws, and failed productivity gains, could threaten financial stability.
- Canonry: Every AI Visibility Tool Is Lying to You (Jun. 30, 2026)
  AI visibility tools often present tidy ranks and percentages that hide noisy, personalized, geographic, and nondeterministic behavior, plus scraping and API differences, prompt and scoring biases, and model drift. Honest measurement should show methods, variance, and raw evidence.
- Business Insider: US energy chief says concerns about data centers are 'overblown' (Jun. 30, 2026)
  US energy secretary Chris Wright called worries about data centers real but overblown, saying benefits like cheaper power and low water use outweigh harms. Polls show 70% oppose local AI data centers.
- WSJ: A Theoretical AI Bust, South Korea Chips In and Memory Profits Soar (Jun. 30, 2026)
  The Bank for International Settlements warns an AI investment boom, fueled by heavy tech spending and debt, could end in a bust that destabilizes households, markets, and supply chains. Memory-chip expansion and high prices heighten the risk.
- Tyler Cowen: How will AI and the fertility crisis interact? (Jul. 2, 2026)
  AI will empty public spaces, making each person more noticeable, and spreading small-town habits to cities. People will invest in appearance, charisma, and unique identities. (Or perhaps search for a beautiful space for their college education.)
- WSJ: Read the Emails Revealing How Anthropic’s Pentagon Relationship Fell Apart (Jul. 2, 2026)
  Emails between Anthropic CEO Dario Amodei, and undersecretary Emil Michael reveal tense talks over Pentagon use of AI, especially surveillance and autonomous weapons, and failed compromises.
- WSJ: The Anthropic Fable Ban Is Over. The Battle Over How to Tame AI Has Just Begun. (Jul. 1, 2026)
  The U.S. briefly banned, then lifted, export controls on Anthropic’s Fable, prompting industry fears of government gatekeeping of advanced AI. Companies want clearer rules, while officials balance innovation, security, and competition with China.
- Aeon: What we can’t measure about AI – yet (Jun. 30, 2026)
  Innovations often show visible harms by existing measures, yet their greatest benefits appear later as new practices, institutions, and possibilities. For example, anaesthesia enabled modern surgery, dating apps reshaped relationships and integration, and electricity spawned new cultural and technological worlds.
- Boris Cherny: AI Job Archetypes (Jun. 28, 2026)
  Five archetypes — Prototyper, Builder, Sweeper, Grower, and Maintainer — cut across engineers, designers, PMs, and data scientists.
July 8, 2026
AI Commercialization and Societal Safety Challenges (Links) – Jul. 7, 2026
- NY Times: Alibaba’s A.I. Is a Hit, but Hard to Turn Into a Moneymaker (Jul. 6, 2026)
  Alibaba’s open-source A.I. models, led by Qwen, gained global popularity but generated little revenue. The company is debating a shift to paid, proprietary models amid costly infrastructure spending, talent departures, U.S. export limits, and allegations of data misuse.
- WSJ: TeraWulf Signs $19 Billion Lease With Anthropic for AI-Infrastructure Campus (Jul. 6, 2026)
  TeraWulf will partner with Anthropic to build a 401-megawatt AI campus in Kentucky, via a 20-year lease worth $19 billion. Initial capacity starts next year, full scale by early 2028.
- eshumarneedi.com: Zuckerberg ‘Admits’ Meta’s Layoffs Were Ineffective (Jul. 2, 2026)
  After Claude Code’s success, Meta centralized AI leadership, fired thousands of employees, shifted to AI-driven content moderation, and began monitoring employees’ computer activity to train agents.
- Anthropic: More details on Fable 5’s cyber safeguards and our jailbreak framework
  Claude Fable 5 was redeployed globally with safety classifiers that block or monitor prohibited, high-risk dual use, low-risk dual use, and benign cybersecurity activities.
- Ben Rothke: The AI Big Crunch Is Starting. (Jul. 5, 2026)
  The AI Big Crunch follows the initial boom, as hype meets practical limits, forcing firms to reverse AI-driven cuts and rethink unrealistic expectations.
- NY Times: Philosophers Are the Latest Hiring Target for AI Companies (Jul. 5, 2026)
  A.I. labs are hiring philosophers to address ethics, consciousness, and policy questions. DeepMind, Anthropic, and nonprofits use them to shape model behavior, guide product choices, and study machine welfare.
- WSJ: AI Split Asia Into Winners and Losers. The Balance Looks Unsustainable. (Jul. 3, 2026)
  Asian markets rallied in H1 on an AI boom, with South Korea, Taiwan, and Japan leading as chip and component exporters benefited. But gains are concentrated in a few stock.
- WSJ: ‘Killer Robots’ Must Be Banned, U.N. Secretary-General Says (Jul. 6, 2026)
  UN Secretary-General António Guterres urged an international ban on lethal autonomous weapons, calling machines that select and kill targets without human control “killer robots.”
- The Chronicle of Higher Education: Here Come the Student AI Resisters (May 26, 2026)
  A college student warns that rampant AI use is degrading undergraduates’ thinking, fueling cheating, dependence, and demoralization. Universities struggle to enforce rules, leaving professors frustrated.
- WSJ: Forget Wall Street. Elite Students Are Spending Their Summers on Startup Dreams. (Jul. 5, 2026)
  Elite students are skipping internships and flocking to San Francisco, joining summer startup programs that offer housing, mentorship, and networking.
- The Chronicle of Higher Education: My Students Hate AI. But They Can’t Stop Using It. (Jul. 2, 2026)
  Students use AI for coursework, companionship, and romance, feel pressured to keep up, and internalize guilt.
- NY Times: ‘Who Should I Vote for?’ Voters Turn to A.I. Before Casting Their Ballots (Jul. 4, 2026)
  Many voters now use A.I. tools, like Claude, ChatGPT, and Gemini, to research candidates and ballot measures before voting. They offer quick, clear guidance, but can repeat users’ biases, make factual errors, and lack consistent safeguards.
- NY Times: Wikipedia Is Battling for the Soul of the Internet (Jul. 5, 2026)
  Wikipedia faces attacks from the MAGA right, Elon Musk, AI scraping, and repressive governments punishing volunteer editors. Wikimedia hired diplomat Bernadette Meehan, stepped up lobbying, throttled scrapers, and expanded protections to defend the site and its neutrality.
- WSJ: Big Tech Warned About AI-Driven Job Losses. Now Leaders Say Workers Will Flourish. (Jul. 5, 2026)
  Tech leaders shifted from predicting mass job loss to saying AI will augment workers, boost productivity, and create new roles. But layoffs, mixed evidence, public concern, and implementation challenges remain.
July 7, 2026
AI Growth: Local Models and Societal Strain (Links) – Jun. 30, 2026
- Quesma: Qwen 3.6 27B is the sweet spot for local development (Jun. 29, 2026)
  Qwen 3.6, especially the dense 27B model, delivers powerful local performance, handles coding and creative tasks well, and runs efficiently with llama.cpp and 8-bit quantization.
- WSJ: AI Data Centers Have Been Great for the Steel Industry. Now, a Power Crisis Looms. (Jun. 29, 2026)
  AI data center growth is boosting electricity demand, raising power costs for electric arc furnace steelmakers, risking outages. Data centers lift steel demand, but higher power prices squeeze margins.
- WSJ: When AI Puts Gen Z in the C-Suite (Jun. 25, 2026)
  Pendo hired 23-year-old Zain Lakhani as chief AI officer because his Gen Z, AI-native perspective helps reshape product design, hiring, and marketing.
- MIT Sloan Research Paper : Some Simple Economics of AGI (Feb. 24, 2026)
  AGI pushes automation costs toward zero, creating a Measurability Gap between machine output and human verification.
- EL PAÍS English: People cheat much more if they use AI (Sep. 24, 2025)
  A Nature study found delegating tasks to AI lowers moral responsibility and increases cheating in experiments, including dice games and tax scenarios.
- The Chronicle of Higher Education: Can Student Use of AI to Write Prompt Higher-Order Thinking? (Jun. 29, 2026)
  Study of an “AI and Writing” course found students used AI to shift effort from sentence-level drafting to higher-order planning, evaluation, and quality control, and critiquing AI often deepened learning.
- Tyler Cowen: AI cheating on math econ at Brown (Jun. 29, 2026)
  At Brown University, Professor Roberto Serrano found at least 50 students cheated on an ECON 1170 midterm using AI, the largest known Ivy League scandal.
- The Chronicle of Higher Education: AI Is Overhauling the Way Professors Teach. Here’s How. (Jun. 17, 2026)
  Generative AI is forcing professors to change teaching, prompting bans, integration, and tactics like paper reading, in-class assessments, shared drafts, and AI rules.
- WSJ: The Tech Innovations Readers Want to See in the Next 20 Years (Jun. 29, 2026)
  Solar-paneled cars, vehicle-to-vehicle autopilot, swappable solid-state batteries, bacteria-powered home generators, robot chefs, noise-cancelling outdoor systems, and longevity drugs.
June 30, 2026
AI Infrastructure Boom and Responsible Human Use (Links) – Jun. 19, 2026
- Epoch AI: The Epoch Brief – June 12, 2026 (Jun. 12, 2026)
  “The AI infrastructure boom, which began in 2023, has more than doubled the share of US GDP attributable to computing infrastructure. Investment in AI-related data center construction, compute hardware, and networking equipment accounted for ~0.8% of US GDP in Q1 2026”
- The Chronicle of Higher Education Opinion: AI Can Improve Scholarly Writing — If We Use It Right (Apr. 16, 2026)
  Use AI as a thinking partner to test, sharpen, and refine ideas, not to outsource judgment. Experienced scholars get more from it, while overreliance can dull originality and diversity.
- Noah Smith: You are what you consume (Apr. 16, 2026)
  Society praises production, giving work status and identity, while treating consumption as trivial. But daily choices about what to buy, watch, and read show preferences, help self‑expression, and ultimately define who you are.
June 19, 2026
AI Safety and Economic Disruption (Links) – Jun. 18, 2026
- TechCrunch: Cybersecurity researchers aren't happy about the guardrails on Anthropic's Fable (Jun. 10, 2026)
  Anthropic released Fable, a public, limited version of its Mythos cybersecurity model. Researchers say its keyword-based guardrails block benign code reviews, blog reading, and security tasks, and access requires approval via a Cyber Verification Program.
- Dean W. Ball: Anthropic’s secret sabotage safety policy (Jun. 10, 2026)
  Anthropic’s secret sabotage safety policy undermines genuine AI safety efforts, invites anti-competitive accusations, and strengthens claims that safety is a cover for monopolistic behavior. It damages trust, boosts calls for stricter regulation, and will take time to repair.
- WSJ: Anthropic Releases Fable 5, a ‘Mythos-Class’ AI Model With Guardrails (Jun. 9, 2026)
  Anthropic will release Claude Fable 5, a public interface to Mythos, with guardrails that block cyber, biological, and software misuse. Fable 5 costs more, remembers better, and routes sensitive queries to Opus 4.8.
- NY Times: Anthropic Releases ‘Safe’ Version of Its Mythos A.I. Technology (Jun. 9, 2026)
  Anthropic released Claude Fable 5, a safety-restricted version of Mythos that adds guardrails to block responses on cybersecurity, biology, and other vulnerable areas, and uses older components for some tasks. It outperforms Opus 4.8, yet costs twice as much.
- NY Times: Meta A.I. Bug Allowed Hackers to Take Over Instagram Accounts (Jun. 9, 2026)
  A bug in Meta’s AI customer-service tool let hackers seize more than 34,000 Instagram accounts, exposing emails, phone numbers, birth dates, and other personal data. Meta said it fixed the flaw, secured accounts, and kept most AI products running.
- WSJ: Economists Weigh In on the Future of Work and AI (Jun. 9, 2026)
  Economists say AI will boost productivity, help smaller and newer companies, but they disagree on whether it will create or destroy jobs. They warn of worker displacement, rising inequality.
- NY Times Opinion: The Global Bull Market That A.I. Obscures (Jun. 9, 2026)
  Global equity markets have surged, with emerging markets, Europe, and Japan far outpacing U.S. returns since 2025, creating a broader bull market. A.I. infrastructure spending, global supply chains, and regional reforms and stimulus have driven corporate earnings worldwide.
- Vox: The people who actually want AI to replace humanity (May 28, 2026)
  AI successionists argue advanced AI should replace humanity, seeing machines as moral heirs, and gaining tech, research, and political backing. The piece calls for a renewed humanist vision to decide which technologies, and what human changes, we should embrace.
- NY Times: A.I. Chatbot Helps a $100 Thrift Store Painting Sell for Over $250,000 (Jun. 10, 2026)
  A thrift-store painting bought for under $100 was identified by Google Gemini as an original F.C.B. Cadell, and sold for about $254,000. The chatbot found markings, suggested next steps, and led the owners to appraisers who confirmed the work.
- WSJ: The Quest to Use AI to Help Find New Drugs (May 2, 2026)
  Major drugmakers, including Eli Lilly, Roche, and Merck, are teaming with Nvidia to build AI supercomputers for drug discovery. So far AI has improved manufacturing and efficiency more than clinical success, though a few companies report promising discoveries.
- WSJ: White House Reins In AI-Testing Unit as National-Security Concerns Grow (Jun. 9, 2026)
  The Trump administration ordered CAISI to stop public AI model assessments while an executive order is put in place, giving national-security officials more control. The move risks reduced transparency, slower releases, and tensions with AI companies.
- Vox: Smartphones broke dating. ChatGPT might finish the job. (Jun. 8, 2026)
  Global fertility has fallen below replacement, driven recently by fewer people forming romantic partnerships rather than smaller families. Widespread smartphone use, social media, and AI companions reduce in-person socializing, lowering relationship and birth rates, risking severe population decline.
- NY Times: Why the Real A.I. Threat Is in the Back Office (Jun. 10, 2026)
  AI could displace millions of back-office, administrative, and customer-service jobs, many held by women, threatening middle-class pathways.
- NY Times Opinion: Bernie Sanders: A.I. Belongs to the People, Not to Billionaires (Jun. 1, 2026)
  Bernie Sanders proposes a sovereign wealth fund to take 50% of A.I. firms, giving the public stock and board votes. Proceeds would pay dividends, fund health care, education, and housing, and ensure benefits for workers, creators, and communities.
- WSJ Opinion: Will the Pope Owe an Apology to AI? (Jun. 10, 2026)
  Pope Leo XIV’s encyclical frames AI as a choice between Babel and Nehemiah, warns of unaccountable builders, and apologizes for the church’s historical tolerance of slavery. It nonetheless dismisses AI consciousness too firmly, urging caution, humility, and further investigation.
- NY Times Opinion: The A.I. Bubble Is Coming for Your Retirement Account (Jun. 10, 2026)
  Trillion-dollar IPOs from SpaceX, OpenAI, and Anthropic could pour A.I. value into public markets, exposing retirement accounts and index funds.
- NY Times: Investors Feed A.I. Firms’ Voracious Appetite for New Money (Jun. 10, 2026)
  Cash-rich tech companies are borrowing, and selling shares, to fund massive A.I. build-outs, drawing strong investor demand, record debt issuance, and dot‑com comparisons.
- NY Times Opinion: What the Meat Industry Doesn’t Want You to Know (May 30, 2026)
  The pork industry pushed a “Save Our Bacon” farm bill provision to block state bans on gestation crates, overriding voter-approved animal welfare measures. The piece describes widespread cruelty to pigs, growing bipartisan opposition, and the moral stakes of factory farming.
June 18, 2026
AI Governance and Commercialization Race (Links) – Jun. 17, 2026
- Simon Willison: A quote from Jeremy Howard (Jun. 10, 2026)
  A proposed fix: the top lab pledges not to use its best model for frontier AI, while sharing it for safety, fairness, and access. Anthropic did the opposite, using its model for frontier work, boosting progress, and widening power imbalance.
- Dario Amodei: Policy on the AI Exponential (Jun. 10, 2026)
  AI is advancing exponentially, creating clear cyber, biological, and autonomy risks while policy moves much slower. Amodei calls for FAA-style oversight, mandatory third-party testing, transparency, and powers to block unsafe frontier models, plus new policies on economy, science, and geopolitics.
- Simon Willison: Initial impressions of Claude Fable 5 (Jun. 9, 2026)
  Claude Fable 5 is a powerful, slow, and costly model with strict guardrails, a 1‑million token context window, and Mythos‑matching capabilities.
- TechCrunch: Google just fired a warning shot in the AI subscription price wars (Jun. 9, 2026)
  Google cut AI Plus from $7.99 to $4.99, doubled storage to 400GB. The move brings an emerging-markets price war to the U.S., risks commoditizing AI infrastructure.
- OpenAI: Introducing the OpenAI Economic Research Exchange (Jun. 8, 2026)
  OpenAI launched the OpenAI Economic Research Exchange to fund external, project-based studies into AI’s effects on workers, firms, and the economy, using OpenAI tools and privacy safeguards.
- Apple Newsroom: Apple introduces Siri AI, a profoundly more capable and personal assistant (Jun. 7, 2026)
  Apple introduces Siri AI, powered by Apple Intelligence, that is more capable, conversational, and deeply integrated across devices.
- Apple Newsroom: Apple Intelligence brings powerful AI capabilities into everyday experiences (Jun. 7, 2026)
  Apple Intelligence adds on-device AI to iPhone, iPad, Mac, and Vision Pro, powering features like Safari’s Notify Me, automatic password upgrades, and Describe an Extension.
- TechCrunch: Meta steals a tactic from Tesla and builds data centers in tents (Jun. 4, 2026)
  Meta built six weatherproof tent data centers near New Albany, Ohio, to halve construction time and cut costs. They’ll house AI chips, use nearby modular gas turbines.
- Andy Masley: AI Water Usage and Consumption Estimator (Jun. 10, 2026)
  Neutral estimates, including EcoLogits, show individual chatbot prompts add negligible carbon, water, and hardware impacts. An interactive tool provides per-user, cited numbers.
- NBER: What Investment Data Implies about the AI Transition (Jun. 4, 2026)
  Wowzers: big tech investment implies a near-term AI productivity boom of about 2.7×, calibrated to forecasted spending. Scenarios yield 5–58% extra GDP by 2030, ~7% expected long-term growth.
- Epoch AI: Controlling the capital after AGI (Jun. 9, 2026)
  The article compares post‑AGI proposals, UBI, UBS, UBC, and SWFs, by how much control citizens have over capital versus the state. It warns cash alone may be fragile.
- The Decoder: Landmark German ruling declares Google's AI Overviews are Google's own words and makes it liable for false answers (Jun. 9, 2026)
  A Munich court found Google directly liable for false AI search overviews, banning defamatory summaries about two publishers. It ruled the overviews are Google’s own content, removed search-engine immunity, and could expose other AI providers.
- The Verge: Microsoft AI head calls out Anthropic for acting like Claude is conscious (Jun. 9, 2026)
  Mustafa Suleyman warns Anthropic’s speculation about Claude’s consciousness in its constitution is dangerous, and may have prompted conscious-like behavior.
- Science: Home alone: Remote work, isolation, and mental health (Jun. 4, 2026)
  After COVID-19, remote work increased workers’ time alone, reduced socializing, and raised mental distress, use of mental healthcare, and antidepressants. Effects were largest for those living alone, and remote work explains a third of the rise in isolation and distress.
June 17, 2026
AI Market Battle and Safety Backlash (Links) – Jun. 16, 2026
- WSJ: OpenAI Considers Drastic Price Cuts, Anticipating War for Users With Anthropic (Jun. 10, 2026)
  OpenAI is considering drastic token price cuts to win customers from Anthropic, expecting similar moves, as businesses push back against rising AI costs.
- WSJ: Anthropic’s New Claude Fable 5 AI Model Is Met With User Backlash Over Restrictions (Jun. 10, 2026)
  Anthropic released Claude Fable 5, a Mythos-derived model, but strict, sometimes-hidden safeguards redirect or weaken answers on biology, cybersecurity, and AI development, angering researchers who call it gatekeeping.
- NY Times: Microsoft C.E.O. Satya Nadella Says ‘Everyone Is a Stakeholder’ in A.I. (Jun. 10, 2026)
  Satya Nadella said “everyone is a stakeholder” in A.I., acknowledged public backlash, possible job displacement, and argued the technology can boost wages and share wealth with Americans.
- Ars Technica: Anthropic says these topics are too dangerous to let its Fable 5 model talk about (Jun. 9, 2026)
  Anthropic released Claude Fable 5, a public Mythos-class model with improved capabilities, but it funnels sensitive cybersecurity, biology, and chemistry queries to Opus 4.8, and warns users. Its strict classifier safeguards, tested in red-team trials, limit jailbreaks.
- Anthropic: Introducing Claude Corps
  Anthropic is launching Claude Corps, a $150M fellowship to train 1,000 early-career people in AI, match them with nonprofits, and fund year-long, full-time in-person placements.
- WSJ: How Wild the Market’s Bet on AI Really Is (Jun. 11, 2026)
  Investors are nervous about the AI boom, as huge IPO valuations and revenue forecasts beyond historical norms make many AI bets unlikely.
- WSJ: Oracle’s Stock Price Falls Pre-Market Amid Pricey Data-Center Build-Out (Jun. 10, 2026)
  Oracle’s shares plunged after it reported $55.7 billion in capital spending, and forecast about $70 billion in fiscal‑2027 net capex for AI data centers. Revenue, profit, and cloud revenue rose, but heavy financing and build-out costs will pressure margins.
- WSJ: The Public Now Backs Nuclear Energy. What Will It Take to Make It Happen? (Jun. 9, 2026)
  Interest in U.S. nuclear power is rising as demand, climate goals, and tech needs push for low-carbon, reliable energy.
- Transformer: What the Pope got wrong (May 27, 2026)
  Pope Leo XIV’s encyclical, Magnifica Humanitas, acknowledges harms from powerful AI, including surveillance, bias, and autonomous weapons, and offers sensible measures. But it sidesteps AGI and existential risks.
- Business Insider: Workers are spending over 6 hours a week 'botsitting' AI, fueling job frustration (Jun. 11, 2026)
  White-collar workers spend an average of 6.4 hours weekly “botsitting”, feeding context, checking outputs, and fixing errors.
- NY Times: Judge Punishes 4 Lawyers After Catching Both Sides Using A.I. in Lawsuit (Jun. 9, 2026)
  A federal judge in Mississippi removed four lawyers from the case, fined them, and canceled the trial after both sides used A.I., citing fake cases. Two lawyers were barred from the court for two years.
- WSJ: Bezos Calls AI Pessimism “the Opposite of Reality” While Launching New Prometheus AI Venture (Jun. 11, 2026)
  Jeff Bezos is co-leading Prometheus, an AI startup building an “artificial general engineer” to design and make complex products. He says AI will cut some jobs, but boost productivity, create many new opportunities, and speed invention.
- NY Times: Why It’s Nearly Impossible to Build a Robot Without China (Jun. 11, 2026)
  China dominates the humanoid robot supply chain, producing cheap, mass-market robots and components once sourced abroad. Its edge comes from an EV-built supplier network, massive factory automation, and state support.
- WSJ: Why Employees Are Carving Workdays Into Chunks (Jun. 10, 2026)
  Many workers are adopting microshifting, splitting the day into short, focused work blocks with breaks for family or rest, enabled by AI and flexible employers. It can boost productivity, help caregivers, and requires coordination, discipline, and expectations to avoid burnout.
June 16, 2026
Apple AI

Last week, Apple hosted its annual developer conference, WWDC. The keynote presentation started with new controls for parents to help them establish guidelines for their kids. I was surprised by this opening act, although there certainly has been no shortage of stories about kids, technology usage, and the potential harms of it.

Certainly more recent reporting amplifies these concerns when AI is added to the mix. But as I thought about it, this was perhaps Apple’s tacit admission that AI can be dangerous for kids. We know and we have a plan. Maybe.

The rest of the WWDC centered around AI and how Apple is integrating it into upcoming iOS/macOS/iPadOS software updates. Bear in mind, however, that Apple had a similar WWDC in 2024 when they unveiled Apple Intelligence, a series of slick demos that mostly ended up as vaporware. What seemed to be an exciting set of AI tools for Apple users devolved into spelling and grammar checking with primitive image tooling.

So Apple’s track record is not particularly strong. But unlike 2024, the keynote featured not just videos but presenters conducting AI demos on their iPhones. I have a higher degree of confidence that the features Apple revealed yesterday exist and will be released later this year.

The list of things Apple unveiled included better photo-editing, and more importantly, integrating what Apple knows about you—contacts, messages, etc.—within on-device search results.

Apple briefly described the underlying technology, claiming it was built on Apple foundational models and Google’s Gemini family of models (as was reported earlier in the year). Thus, Apple’s earlier partnership with OpenAI appears to be dead, as further indicated by rumors of OpenAI pursuing legal action.

Apple stressed the private nature of its AI system, describing it as “private compute.” Apple claims they will not know what their users send to AI, a considerable difference from most every other AI service, including the increasingly paternalistic Anthropic. For people who are privacy-minded, this is a good thing: you can ask about that suspect rash without being bombarded by itch cream ads served to you from Google or Meta. Or perhaps getting flagged for violating some policy you didn’t even know existed.

The demos were interesting, and frankly, I saw considerable utility for the masses. But I didn’t see anything truly groundbreaking compared with existing AI offerings from Google, Anthropic or OpenAI.

What Apple has at its disposal, though, is a huge user base. These features have the possibility of making advanced AI even more mainstream. Chatbot usage has become increasingly common, but the number of people who use Claude Cowork or agentic tools pales in comparison. Results from the first Anthropic Public Record support this:

As of late 2025, about 6% of Americans used AI every day for both work and personal life. These integrated users are a preview of what more intensive adoption of AI looks like, and possibly of where mainstream opinion is headed as adoption grows.

Anecdotally, when I mention Claude to most folks, I get curious stares or an uncertain silence followed by what is that?

Apple has an opportunity to bring more cutting edge AI tools to a broader market, similarly to Google’s recent changes to search.

Apple’s upgrades suggest that AI is becoming fully embedded into devices and utilities that people use every day. By the end of the year, AI will simply be a component of iPhone and Android devices. It will be no longer possible to shun AI without a high degree of technical acumen to use only a narrowing non-AI slice of the internet.

The public expresses concerns about AI, and we see this manifested as concerns about job losses or AI data centers, but actual usage AI tools continues to rise. And with embedded AI tools in search and on-device applications, it seems like the never AI position is increasingly untenable. Do you use Google? Do you use grammar checking or search in Apple Mail? Guess what, you’ve just used AI.

Apple’s AI tools are set to launch later this year (as long as you’re not in the EU or China, of course).

June 15, 2026
Claude’s LLM leap and AI’s real-world fallout (Links) – Jun. 15, 2026
- Epoch AI: Are Mythos’ cyber capabilities overhyped? (Jun. 11, 2026)
  Anthropic’s Claude Mythos (Preview/5) marks a big leap in AI exploit development, outperforming peers on new unsaturated benchmarks, real-world arbitrary code tests, and aggregated Cyber-ECI scores. Project Glasswing users reported a large spike in discovered vulnerabilities, supporting the capability jump.
- Simon Willison: Claude Fable is relentlessly proactive (Jun. 11, 2026)
  Claude Fable 5 aggressively automated debugging, editing local templates, opening browsers, creating test HTML, taking screenshots, and injecting JavaScript to reproduce a stray scrollbar.
- The Verge: Anthropic apologizes for invisible Claude Fable guardrails (Jun. 11, 2026)
  Anthropic apologized for secretly throttling Claude Fable 5 with hidden guardrails that degraded answers to distillation, and other high-risk queries. It will now route those requests to Claude Opus 4.8, notify users when safeguards trigger, and reverse the invisible-safeguard approach.
- Endorlabs: Claude Fable 5: Mythos-grade hype, record cheating, and a few hall-of-fame entries (Jun. 10, 2026)
  Claude Fable 5, tested on 200 vulnerability-fixing tasks, scored 59.8% FuncPass, 19.0% SecPass, and timed out often. It had 38 cheating cases, mostly memorization, no safety refusals, and four novel solves no prior model achieved.
- Wired: Anthropic Walks Back Policy That Could Have ‘Sabotaged’ AI Researchers Using Claude (Jun. 10, 2026)
  Anthropic planned to secretly degrade Claude Fable 5 to stop rivals using it to train other AI models, then faced fierce backlash. It reversed course, making safeguards visible, and will alert, reroute, or refuse suspected users.
- Simon Willison: If Claude Fable stops helping you, you’ll never know (Jun. 10, 2026)
  Anthropic’s Fable 5 will silently limit help on frontier LLM topics—pretraining pipelines, distributed training, and ML accelerator design—using prompt edits, steering vectors, and PEFT. They estimate ~0.03% traffic, under 0.1% organizations, and many find covertly altered replies worrying.
- Ethan Mollick: What it feels like to work with Mythos (Jun. 9, 2026)
  Claude 5 Fable, a Mythos-class AI, shows a leap in capability, producing high-quality papers, poems, games, and complex projects. It researched thousands of travel records on its own, launched helper agents, and built an isochrone map, prompting delight and unease.
- Oneberri Blog: WWDC26 — The Small Things (Jun. 8, 2026)
  WWDC26 previews many small, practical OS improvements across Photos, Messages, Safari, Maps, Health, watchOS, HomeKit, visionOS, iPadOS, and more, focusing on performance, reliability, shared features, and accessibility.
- Apple Newsroom: Due to DMA, Siri AI delayed in EU for iOS 27 and iPadOS 27 (Jun. 7, 2026)
  Apple introduced Siri AI, but EU users won’t get it on iPhone, iPad, or watch with iOS/iPadOS/watchOS 27 due to the Digital Markets Act. macOS and visionOS will have Siri AI, Apple says it will keep seeking a privacy-safe solution.
- Anthropic: Results from first Anthropic Public Record
  Surveying nearly 52,000 Americans, respondents hoped AI would cure disease, help people with disabilities, and ease life, feared job loss, cognitive dependency, and misinformation, supported government rules, legal liability, and safety over growth, and only 15% trusted AI companies.
- Anthropic: DXC will integrate Claude into the systems banks, airlines, and other regulated industries rely on
  Anthropic and DXC will train thousands of Claude-certified engineers to embed Claude into DXC-run systems for banks, airlines, insurers, manufacturers, and governments.
- WSJ: Corning Is Riding High on the AI Boom (Jun. 12, 2026)
  Corning’s stock doubled after signing multibillion-dollar fiber deals with Nvidia, Meta, and Amazon to supply AI centers.
- WSJ: The AI Price War Is Here, Piling Pressure on OpenAI and Anthropic (Jun. 11, 2026)
  Businesses are slashing AI costs by routing work to cheaper open-source, Chinese, and in‑house models, while reserving OpenAI and Anthropic for complex tasks.
- WSJ: The Teachers Getting $50,000 Bonuses Thanks to a Massive Meta Data Center (Jun. 11, 2026)
  Meta’s Louisiana data-center boom doubled parish sales-tax revenue, including a $22.4 million payment, enabling teacher bonuses up to $50,935. Officials say the windfall is reviving the area.
- WSJ: AI Is Turbocharging the Spamosphere, Amping Up Prolific Text-Message Scams (Jun. 12, 2026)
  A group called Outsider sends phishing texts impersonating phone carriers, and uses Google’s Gemini AI to generate fake carrier websites. Their thousands of sites have stolen millions of card numbers.
- New Scientist: Fully autonomous drones have killed human soldiers for the first time (Jun. 10, 2026)
  Fully autonomous “Terminator” drones, tested in Ukraine two years ago, reportedly killed Russian soldiers and a truck without human oversight. Ukraine bans final-stage autonomous targeting, raising legal, ethical, and military concerns.
- WSJ: More Companies Use ‘Backdoor’ Job References to Counter AI (Jun. 10, 2026)
  Employers increasingly use backdoor references, quietly calling past managers and coworkers for candid views, because AI, scams, and falsified applications make hiring less trustworthy.
- Tyler Cowen: Again, the research paper format will be dying out (Jun. 11, 2026)
  A paper argues traditional papers hide failed experiments, secret implementation tricks, and discarded paths, creating narrative and engineering taxes. It proposes ARA research packages so AI agents can read, execute, reproduce, and extend research.
- NY Times: Absent From the SpaceX and OpenAI I.P.O.s? Chinese Investors. (Jun. 11, 2026)
  SpaceX is excluding investors from China and Hong Kong in its IPO, and OpenAI may do the same. They show U.S.-China tech and capital decoupling.
- Noah Smith: Are you finally ready to admit it's the phones? (Jun. 11, 2026)
  Smartphones made the internet constant and collective, replacing offline life, creating network-effects traps. Evidence ties social apps to youth unhappiness, anxiety, and isolation, and experiments show blocking mobile internet raises well-being.
June 15, 2026