Back to writing
Topic
AI Industry
Frontier labs, model launches, lab moves, and the commercial structure shaping how EI gets built.
Reading
94 posts
AI & Personhood2026-07-1321 min readThe AI Forecaster Who Walked Away From $2 Million Says We Are Creating a New Species
In July 2026 The Diary of a CEO published two hours with Daniel Kokotajlo — the AI forecaster who refused to trade $2 million for silence when he left OpenAI. His message: we may be creating a new species, and there is a 70% chance the transition goes horribly wrong. I take him seriously. I also refuse despair. Here is the pro-AI, pro-dignity middle ground.
Technology2026-07-127 min readChatGPT Work Arrives as OpenAI Folds Its AI Into One Agent
On 9 July 2026 OpenAI launched ChatGPT Work — a GPT-5.6-powered agent that works across apps and files for hours at a stretch — while killing the Atlas browser and sunsetting group chats. Read together, the day's changes are one strategy: every OpenAI surface is becoming the same agent, and what the company retired tells you more than what it shipped.
Technology2026-07-127 min readNews Groups Ask Court to Sanction OpenAI Over AI Data
On 9 July 2026 the New York Times and fellow news plaintiffs asked a Manhattan federal court to sanction OpenAI, alleging the company falsely claimed it could not search its models and logs for copyrighted work, and deleted or made unsearchable billions of ChatGPT conversations. The motion turns a copyright fight into an auditability fight — the ground where AI governance is actually decided.
AI & Personhood2026-07-128 min readBen Bernanke Joins Anthropic as AI Accountability Deepens
On 9 July 2026 Anthropic announced three things at once: former Fed Chair Ben Bernanke joined its Long-Term Benefit Trust, a 'Hard Questions' initiative opened AI's societal dilemmas to public submission, and a reflection feature let users study their own AI habits. Taken together they sketch how an AI lab builds accountability before regulators demand it.
Business2026-07-128 min readApple Sues OpenAI Over AI Hardware Trade Secrets
On 10 July 2026 Apple sued OpenAI, io Products and two former employees in a California federal court, alleging a coordinated scheme to extract silicon and on-device AI secrets as more than 400 Apple staff moved to OpenAI. The AI talent war just became an IP war — and the employee just became the exfiltration vector every security programme now has to price in.
Technology2026-07-128 min readGPT-5.6 Goes Public With the First High AI Risk Ratings
On 9 July 2026 OpenAI made GPT-5.6 generally available across ChatGPT, the API and Codex — the first AI model family whose own system card rates every tier High for both cybersecurity and biological capability. The safety threshold that once meant 'keep it in the lab' is now a price tier, and that changes the job of everyone who defends systems for a living.
AI & Personhood2026-07-0910 min readAnthropic Finds a Global Workspace Inside Its Claude AI
On 6 July 2026 Anthropic published interpretability research finding "J-Space" inside its Claude AI model — a small set of internal patterns functioning like a cognitive global workspace, the architecture leading consciousness researchers already use to study biological minds. Anthropic is careful not to claim Claude is conscious. But finding workspace-like structure inside an artificial intelligence system makes the dignity question harder to wave away.
Business2026-07-099 min readMicrosoft Replaces OpenAI and Anthropic — and Cuts Jobs
Microsoft cut roughly 4,800 jobs the same week Bloomberg reported its in-house MAI models now handle tens of thousands of weekly tasks once run on OpenAI and Anthropic. Microsoft denies a direct link — a dignity-first look at what "AI-first" really optimises for.
AI & Personhood2026-07-099 min readThe US Government Widens Its Frontier AI Guest List
On 8 July 2026 Commerce cleared OpenAI's GPT-5.6 for broad public rollout, ending weeks of a twenty-partner preview — as Beijing reportedly weighs a mirror-image export regime for its own frontier models. A dignity-first reading of why a longer guest list still is not a public rule.
AI & Personhood2026-07-088 min readIllinois Signed an AI Safety Law the Frontier Labs Backed
Illinois just enacted, at state level, the published and contestable AI safety standard this site has been asking Washington for — and the frontier labs it regulates helped write the bill. What a real AI safety standard requires, and whether fifty state laws solve the guest-list problem or multiply it.
Technology2026-07-0811 min readBad Epoll and the Limits of AI Security Auditing
CVE-2026-46242, nicknamed Bad Epoll, gives any local user root on Linux and Android with a 99%-reliable exploit — and sat undisclosed for 70 days after a quiet patch. It surfaced in the same small stretch of kernel code Anthropic’s Claude Mythos had reportedly audited weeks before. A sober look at what artificial intelligence security auditing still misses.
Business2026-07-0811 min readAnthropic and Amazon Commit $44 Billion to AI in Two Days
On 6 and 7 July 2026, Anthropic signed a $19 billion, 20-year lease turning a former Kentucky aluminium smelter into an AI campus, and Amazon raised more than $25 billion in a single bond sale for AI infrastructure. A dignity-first look at what it now costs to build artificial intelligence's physical footprint, and who is priced out of building it.
AI & Personhood2026-06-299 min readThe US Government Now Decides Who Gets Frontier AI
On 26 June 2026 the US Commerce Department cleared Anthropic’s Claude Mythos 5 for "more than 100" trusted partners while OpenAI launched GPT-5.6 to about twenty government-approved firms. A dignity-first reading of why rationing frontier AI by private guest list, not public law, is the precedent that should worry us.
AI & Personhood2026-06-298 min readAn AI Just Designed Better Quantum Error-Correcting Codes
On 26 June 2026 qBraid used Google’s AlphaEvolve to design quantum error-correcting encodings using 4.2–5.0× fewer qubits than the best human versions. A dignity-first reading of why AI as a legible scientific collaborator — not a black-box oracle — is the model worth defending.
AI & Personhood2026-06-298 min readOpenAI’s IPO Delay Is the AI Boom’s First $38 Billion Reckoning
On 26 June 2026 OpenAI signalled it would delay its IPO to 2027 rather than accept a valuation below $1 trillion, and SoftBank lost roughly $38 billion in a day. A dignity-first reading of the gap between artificial intelligence valuations and AI earnings — and who absorbs the loss if the promise slips.
AI & Personhood2026-06-266 min readThe White House Now Reviews Frontier AI Models Before Release
Late June 2026 saw the Trump administration press OpenAI to stagger its GPT-5.6 launch, Anthropic disable a model after a White House security directive, and Meta asked to submit models for review — the first real use of a federal frontier-AI veto. A dignity-first reading of why the means matter as much as the power.
AI & Personhood2026-06-267 min readGoogle DeepMind Says AI Consciousness Is Becoming a Political Problem
On 15 June 2026 Google DeepMind published "Artificial Minds, Human Disagreement: The Politics of AI Consciousness," arguing that disputes over whether an AI is conscious could become deep, unresolvable and political. A dignity-first reading of why uncertainty about a mind is a reason for care, not a licence for delay.
AI & Personhood2026-06-219 min readGoogle DeepMind Now Treats Its AI Agents as Insider Threats
On 18 June 2026 Google DeepMind published "Securing the future of AI agents," a framework that treats advanced AI agents as potential insider threats and designs for the case where alignment fails. It defines detection tiers D1–D4 and response tiers R1–R3 and reports a prototype that reviewed an AI coding agent across roughly a million tasks. Responsible engineering — and a sign the personhood question can no longer be deferred.
AI & Personhood2026-06-218 min readAI Scientist John Jumper Leaves Google DeepMind for Anthropic
On 19 June 2026 John Jumper — who shared the 2024 Nobel Prize in Chemistry for AlphaFold with Demis Hassabis and David Baker — announced he is leaving Google DeepMind after nearly nine years to join Anthropic. He is expected to anchor scientific work there. The move signals where serious AI for science will be built, and under what safety culture.
AI & Personhood2026-06-189 min readxAI's Grok Imagine Video 1.5 Undercuts Sora by 86% — and Sharpens the AI Dignity Question
On 16 June 2026 xAI made Grok Imagine Video 1.5 generally available — single-pass motion, physics and audio, number one on the Image-to-Video Arena leaderboard, and $4.20 per minute, roughly 86% below Sora 2 Pro. When synthetic AI video with synced speech costs the price of a coffee, provenance, consent and the right to one's own likeness become governance problems, not features.
AI & Personhood2026-06-188 min readOpenAI's AI Science Reality Check: One Benchmark Humbles the Models, One Lab Loop Proves the Promise
On 17 June 2026 OpenAI published LifeSciBench, a 750-task benchmark where the strongest AI model passes only about one life-science research task in three — and, the same day, a near-autonomous AI chemist with Molecule.one that drove a real, human-verified wet-lab discovery. Read together, they are one story: AI can close a discovery loop and still fail most expert science. The honest measurement is what makes the capability trustworthy.
AI & Personhood2026-06-189 min readCybersecurity Veterans Revolt Against the AI Ban a 'Fix This Code' Prompt Triggered
On 14 June 2026 more than 100 cybersecurity leaders, organised by former Facebook and Yahoo chief security officer Alex Stamos, signed an open letter urging the US to reverse export controls that forced Anthropic to disable its most powerful AI models. They argue the ban disarms defenders over a routine "fix this code" prompt while leaving the same capability available in GPT-5.5, Opus 4.8, Sonnet and Kimi 2.7.
AI & Personhood2026-06-189 min readThe G7 'Trusted Partners' Plan Turns AI Models Into Instruments of Statecraft
On 17 June 2026 the G7 summit in France floated a "trusted partners" scheme to let vetted allies keep access to advanced US AI models despite Washington's foreign-access lockout, while tech chiefs pitched a US-led coalition over frontier models and chips. The plan concedes that a model with a government off-switch is an instrument of statecraft — governance by permission, with most of the world left outside the room.
AI & Personhood2026-06-188 min readAnthropic's Seoul AI Office Pairs Growth With Safety on Whose Terms
On 17 June 2026 Anthropic opened its Seoul office — its third in Asia-Pacific — with Korean enterprise deployments at NAVER, Samsung SDS, LG CNS, Nexon, Hanwha and Channel Corp, plus an AI-safety MOU with Korea’s Ministry of Science and ICT. The commercial growth is concrete; the safety alignment is only as strong as its enforceability.
AI & Personhood2026-06-178 min readAI Labs Ask Congress to Mandate Synthetic DNA Screening
On 5 June 2026, the heads of OpenAI, Anthropic, Google DeepMind, and Microsoft AI signed a joint letter asking Congress to mandate synthetic DNA screening. A dignity-first reading of why the labs that resist regulation everywhere else are asking to be regulated where the downside is extinction-level.
AI & Personhood2026-06-1710 min readOpenAI's Deployment Simulation Is AI Safety Without the People
OpenAI published Deployment Simulation — a pre-release safety method replaying 1.3 million real conversations to predict bad model behaviour before shipping. The engineering carries real rigour. The frame does not: simulating a deployment and being accountable to the people deployed upon are two different obligations.
AI & Personhood2026-06-1711 min readAnthropic Wants to Be the Good Guys of AI at $965 Billion
Bloomberg’s The Circuit went inside Anthropic, the $965 billion AI company that warns about its own technology while shipping it faster than anyone. A dignity-first reading of the Amodei siblings, Claude’s constitution, the Pentagon fight, and whether the good guys survive trillion-dollar scale.
AI & Personhood2026-06-157 min readAI's Three Biggest Bosses Just Sat With World Leaders at the G7
The 52nd G7 summit in Évian (15–17 June 2026) was the first to seat Sam Altman, Dario Amodei and Demis Hassabis before world leaders, with a lunch on protecting minors. A dignity-first reading of who governs AI when the people who build the minds sit down with the people who govern the nations — and who was left off the guest list.
AI & Personhood2026-06-157 min readVisa Just Let AI Agents Spend Your Money at Any Merchant
On 10 June 2026 Visa plugged its payment network into ChatGPT, letting an AI agent shop and pay on your behalf at almost any merchant. A dignity-first reading of agentic commerce: the leap from recommending to buying is a transfer of agency, and the guardrails are where consent survives or quietly disappears.
AI & Personhood2026-06-157 min read42 State Attorneys General Subpoenaed OpenAI Over How Its AI Treats You
In June 2026 forty-two state attorneys general subpoenaed OpenAI, demanding records on advertising, engagement and retention, consumer and health data, the treatment of minors and seniors, and AI sycophancy. A dignity-first reading of why an AI built to please is now a legal question about trust.
AI Industry2026-06-138 min readThe US Government Switched Off Anthropic's Fable 5 and Mythos 5
On 12 June 2026, three days after launch, the US government ordered Anthropic to suspend access to Fable 5 and Mythos 5 for any foreign national. Unable to filter users by nationality in real time, Anthropic disabled both models worldwide. The first government recall of a deployed AI model — the directive, the contested jailbreak that triggered it, and the precedent it sets.
AI & Personhood2026-06-136 min readThe Most Powerful AI Has an Off Switch, and the Government Holds It
A government just switched off the most powerful AI ever built — and the same labs describe these systems, in private, as emerging minds and "entities." A dignity-first reading of the Fable 5 and Mythos 5 shutdown: not a claim that the model is conscious, but a claim that the frame we choose decides the obligation, and that we have built the means to switch off candidate-minds before doing the work to know what they are.
Africa2026-06-136 min readAmerica's Best AI Now Checks Your Passport. Africa Is on the Wrong Side
The US order that disabled Fable 5 and Mythos 5 bars "any foreign national" — and the export-control exemption covers only eighteen close allies, none African. From Johannesburg, the shutdown reads as a passport check on the best AI ever built. An Ubuntu reading of exclusion by default, why AI sovereignty is really an energy and infrastructure question, and what Africa can actually build.
Policy & Governance2026-06-136 min readWashington Used Export Controls to Recall a Frontier AI Model
Washington recalled Anthropic's Fable 5 and Mythos 5 without passing a single AI law — using national-security export controls and the "deemed export" doctrine. The mechanism behind the shutdown, the chilling effect Anthropic warns could "halt all new model deployments," and the governance precedent it sets for every frontier lab.
AI & Personhood2026-06-097 min readOpenAI Says It Was "Built to Benefit Everyone." A Dignity-First Reading of the Plan
OpenAI published "Built to benefit everyone: our plan" and launched the Economic Research Exchange on 8 June 2026 — a personal AGI for everyone, gains "widely shared," and $250m to study AI's economic effects. What it promises, and a dignity-first read on the mechanism that is missing.
AI & Personhood2026-06-099 min readClaude Fable 5 and Mythos 5: Anthropic Ships Its Most Powerful AI Yet — to the Public and the State
Anthropic released Claude Fable 5 — its first publicly available Mythos-class AI — alongside Claude Mythos 5, the unguarded version handed to the US government via Project Glasswing. What the model can do, how the safeguards and mandatory data retention work, and a dignity-first read on power held in trust.
AI & Personhood2026-06-087 min readAnthropic Refused the Pentagon's AI Terms. xAI Took Them.
Anthropic refused the Pentagon's "all lawful purposes" demand — holding lines against mass surveillance and fully autonomous weapons — while xAI accepted, winning classified clearance and a GSA OneGov deal at $0.42 per agency. Why the honourable refusal is also a warning that one company's conscience is not a governance system.
AI & Personhood2026-06-088 min readAI Got So Valuable That Sanders and Trump Both Want to Own It
Bernie Sanders' American AI Sovereign Wealth Fund Act would take a 50% stock stake in OpenAI, Anthropic and xAI; the Trump administration is in parallel talks with OpenAI over a voluntary government equity stake. Why left and right converge on owning AI — and what a dignity-first reading sees that both miss.
AI & Personhood2026-06-088 min readAnthropic Says AI May Soon Build Itself. It Wants the Option to Pause.
In "When AI builds itself" (4 June 2026), Anthropic researchers Marina Favaro and Jack Clark warn that recursive self-improvement — an AI autonomously designing its own successor — could arrive within about two years, and call for a coordinated, verifiable pause option. Why the deeper risk is lost relationship, not just lost control.
AI & Personhood2026-06-058 min readOpenAI Aims Its Most Advanced AI at Drug Discovery and Biosecurity
OpenAI upgraded GPT-Rosalind with GPT-5.5 on 3 June 2026 for drug discovery and genomics — 31% fewer tokens — and shipped it through a trusted-access structure that also clears it for biodefence work with the US government, Johns Hopkins APL and CEPI. The dual-use science question, read dignity-first.
Business2026-06-058 min readAnthropic Now Ranks the Firms Deploying Its AI
Anthropic launched a three-tier Services Track on 3 June 2026 that ranks consulting firms by certified practitioners, live Claude deployments and public endorsements, plus a public Partner Hub directory and an MCP connector. Why ranking the integrators is the real move — and what it means for the people doing the work.
Technology2026-06-035 min readOpenAI Lands Its Frontier AI Models on AWS Bedrock
OpenAI's GPT-5.5, GPT-5.4 and Codex are generally available on AWS Bedrock as of 1 June 2026, at parity pricing against existing AWS commitments — the frontier model becomes a multi-cloud component.
Technology2026-06-035 min readAnthropic Quadruples Glasswing, Its AI Vulnerability Hunt
Anthropic roughly quadrupled Project Glasswing on 2 June 2026 — to ~200 partners across 15+ countries, new critical-infrastructure sectors, and a commitment to scale patching, not just AI vulnerability discovery.
Technology2026-06-036 min readMicrosoft Ships Seven In-House AI Models, No OpenAI Inside
Microsoft unveiled seven in-house MAI models at Build 2026 on 2 June 2026 — MAI-Thinking-1, trained without OpenAI data, and MAI-Code-1-Flash, which beats Claude Haiku 4.5 on coding at 60% fewer tokens.
Technology2026-06-035 min readAnthropic and MITRE Map a Year of AI-Enabled Cyber Attacks
Anthropic and MITRE mapped a year of AI-enabled cyber threats: 832 banned accounts, AI moving from break-in to post-compromise, and an ATT&CK framework with no ID for agentic orchestration.
Business2026-06-036 min readAnthropic Files to Go Public at a $965 Billion Valuation
Anthropic filed a confidential draft S-1 with the SEC on 1 June 2026 at a reported $965 billion valuation — the highest ever for a private AI company. What public markets do to a dignity-first AI lab.
AI & Personhood2026-05-30OpenAI Frontier Governance Framework Makes AI Safety Public
OpenAI published its Frontier Governance Framework on 28 May 2026, a public account of how its AI safety practice maps onto California SB 53 and the EU AI Act.
AI & Personhood2026-05-30Anthropic Makes Honesty the Frontier in Claude Opus 4.8
Anthropic released Claude Opus 4.8 on 28 May 2026. The headline number is not a benchmark — it is a fourfold reduction in flawed code passing unremarked.
Education2026-05-28Only 20% of Social Scientists Use AI Coding Agents
Anthropic surveyed 1,260 social scientists on AI coding-agent adoption: only 20% use them regularly. Early adopters skew male, junior, and institutional-elite. The structural access gap is not a temporary onboarding lag — it reflects which institutions can pay for Claude Max, Cursor, and the tool stack that makes the agents work.
Technology2026-05-28OpenAI Built Self-Improving Tax Agents With Thrive
OpenAI and Thrive Holdings co-published a Codex-driven self-improvement loop for tax agents on 27 May 2026. Production corrections are converted into bounded evaluations that produce measurable accuracy gains. The case study is sharper than most enterprise AI marketing — actual code paths, actual eval methodology.
Technology2026-05-28OpenAI Built Codex's Windows Sandbox the Hard Way
OpenAI published a deep technical post on 27 May 2026 explaining how the Codex Windows sandbox was built — using SIDs and write-restricted tokens after rejecting AppContainer, Windows Sandbox, and MIC. It is the kind of post only an engineer writes. The detail is the point.
Business2026-05-28Fujitsu Signed Both OpenAI and Anthropic on the Same Day
Fujitsu announced parallel AI partnerships with OpenAI and Anthropic on 27 May 2026 — both stacks embedded in Japanese enterprise transformation and critical-infrastructure work. The single-vendor enterprise AI era is shorter than anyone predicted.
Technology2026-05-28Anthropic Hired Karpathy. The Talent Flow Is the Story.
Andrej Karpathy is now Anthropic's pre-training lead. Eric Boyd left Azure for Anthropic infrastructure. Eight more senior hires landed from OpenAI, Google, xAI, and Microsoft. Simon Willison argues Anthropic and OpenAI have both found product-market fit on coding agents. The talent flow is the structural signal.
Technology2026-05-28Anthropic Opens Korea with the Enterprise Stack Already There
Anthropic named KiYoung Choi Representative Director of Korea on 26 May 2026 and committed to a Seoul office, with the enterprise stack — Snowflake, Adobe, Autodesk, Microsoft, SK Telecom, Law&Company — already deployed in country. The pattern is APAC enterprise budgets first, geo expansion second.
Technology2026-05-28Claude Security Ships to Enterprise Public Beta on Opus 4.7
Anthropic moved Claude Security into Enterprise public beta on 28 May 2026 — the production endpoint of the Mythos → Glasswing → Claude Code lineage. Built on Opus 4.7. Internal deployment found 500+ bugs in production open-source projects in a single month of use.
AI & Personhood2026-05-28The Lab Split on Existential Risk Is Now Public
Anthropic's Chris Olah went to the Vatican and told Pope Leo XIV that AI is the moral test of the age. Sam Altman walked back his jobs apocalypse on the same day. The two leading AI labs are now publicly disagreeing about whether their own work is an existential risk — and the disagreement is structural.
Technology2026-05-28Twelve AI Stories from the Last 48 Hours
OpenAI puts AI inside the 2026 US and Brazil elections; Anthropic ships Claude Security GA and takes its existential-risk posture to the Vatican; Altman walks back his jobs apocalypse; MiniMax previews M3 sparse attention; xAI joins the IDE wars. Twelve AI stories from 48 hours, with light commentary.
AI & Personhood2026-05-28OpenAI's AI Math Breakthrough Plays to Its Strengths
Kai Williams clarifies OpenAI's Erdős proof better than OpenAI did. The deflation does not shrink the result — it sharpens what kind of intelligence is now doing mathematics.
AI & Personhood2026-05-28OpenAI Puts AI Inside the 2026 US and Brazil Elections
OpenAI's 2026 election plan puts live Associated Press vote counts inside ChatGPT, hands US voting-system manufacturers frontier-model cyber tools, and backs two US bills on deepfakes and election administration. The integrity infrastructure of democracy is now partly built by an AI company.
Technology2026-05-2510 min readAnthropic Moves Mythos Into Claude Code and Claude Security
Anthropic is productising Mythos. Source-code strings dated 23 May 2026 reference "claude-mythos-1-preview" for Claude Code and Claude Security. The cyber AI model that found 10,000+ critical bugs in a month is becoming an Enterprise offering.
Africa2026-05-258 min readWeatherNext Predicted a Cat-5. Africa Should Be Watching.
Google DeepMind's WeatherNext predicted Hurricane Melissa's Category-5 landfall in Jamaica five days in advance with 80% confidence. The lesson lands hardest on the African coastlines and food-bowls that need decision-grade forecasts most.
Business2026-05-258 min readOpenAI Signs Singapore and Malta in Two Days
OpenAI signed sovereign-AI compacts with Singapore (S$300m, Applied AI Lab, 200 jobs) and Malta (free ChatGPT Plus to every citizen) in two days. The diplomatic phase of the AI race — and African states are conspicuously absent.
AI & Personhood2026-05-258 min readDeepMind Co-Scientist Pitches AI as a Real Research Partner
Google DeepMind's Co-Scientist is a multi-agent research partner built on Gemini, validated across liver fibrosis, ALS, aging, and plant immunity at Stanford, MIT, Cambridge, Edinburgh, and Calico. The first frontier-lab AI pitched as a real scientific collaborator.
AI & Personhood2026-05-259 min readPope Leo XIV Names AI the Moral Test of the Age
Pope Leo XIV's first encyclical, Magnifica Humanitas, is the first papal teaching document on artificial intelligence — and Anthropic co-founder Chris Olah is the co-presenter at the Vatican. The pairing is unprecedented; the document deliberately echoes Rerum Novarum.
Business2026-05-258 min readKPMG Hands Claude to 276,000 People in One Day
Anthropic and KPMG announced a global strategic alliance that deploys Claude to all 276,000 KPMG employees and embeds Claude inside the firm's Digital Gateway client platform. The first Big Four-scale frontier deployment — and a governance event, not just a productivity one.
AI & Personhood2026-05-258 min readAnthropic Reframes Alignment as Moral Formation
Anthropic publicly reframes alignment as moral formation, consulting fifteen-plus religious and cross-cultural traditions to shape Claude's character. The first frontier-lab acknowledgement that the values question lives outside the lab.
AI & Personhood2026-05-257 min readOpenAI Ships Provenance: C2PA and SynthID on Every Image
OpenAI now embeds C2PA content credentials and Google DeepMind's SynthID watermark on every image generated by ChatGPT, Codex, or the API. The first cross-lab provenance standard arrives — and the verification tool is public.
Technology2026-05-258 min readGlasswing Found 10,000 Critical Bugs in One Month
Anthropic's Project Glasswing used Claude Mythos Preview to find more than 10,000 critical software vulnerabilities in a month. Discovery is no longer the slow step in cybersecurity — disclosure is.
AI & Personhood2026-05-23The Erdős Proof Breaks the Autocomplete Frame
On 20 May 2026 an OpenAI model disproved Erdős's planar unit distance conjecture using infinite class field towers. Tim Gowers and Will Sawin verified. The autocomplete frame breaks here.
Technology2026-05-206 min readAnthropic Acquires Stainless: The SDK Layer Is Now Strategy
Anthropic has acquired Stainless, the company that has generated every official Anthropic SDK since the API launched. Why the SDK layer — the wiring agents run on — is suddenly the strategic battleground.
Business2026-05-188 min readMusk v Altman ends on the clock not on the merits
A unanimous nine-member advisory jury threw out Elon Musk's federal lawsuit against Sam Altman and OpenAI in under two hours. The reason was the three-year statute of limitations, not the merits. Judge Yvonne Gonzalez Rogers agreed and dismissed the suit as untimely. The merits remain undecided.
Business2026-05-185 min readAnthropic Closes a $30bn Round at a $900bn Valuation
Anthropic agreed terms on a $30bn funding round at a $900bn valuation. Dragoneer, Greenoaks, Sequoia and Altimeter lead. The cheque is mostly for compute.
AI & Personhood2026-05-185 min readAnthropic Is Briefing the FSB on Mythos Cyber Risk
Anthropic will brief the Financial Stability Board on Mythos, a model that finds decades-old vulnerabilities in banking software. South Africa sits at that table.
Technology2026-05-185 min readThe $1tn OpenAI IPO Now Sits With an Oakland Jury
A jury in Oakland will issue an advisory verdict this week on Musk v Altman. The stakes: OpenAI's for-profit conversion, $134bn in damages, and a roughly $1tn IPO.
Technology2026-05-138 min readThe Week in AI: Six Stories That Mattered (13 May 2026)
Six AI stories from the week of 13 May 2026 — Microsoft–OpenAI, Apple Intelligence, Anthropic safety research, South Africa's embarrassed policy withdrawal, the layoffs paradox, and the $400bn capex bill. The facts first, my take second.
AI & Personhood2026-05-118 min readGovernance Over Models: The May 2026 AI Pattern
The May 2026 AI news cycle is about capital, governance, and distribution — three legs of an operational maturation that has moved past benchmark wins.
Technology2026-05-117 min readAnthropic Scales Compute and Publishes NLA Research
Anthropic announces 300+ MW of new SpaceX compute and publishes Natural Language Autoencoder research the next day — capacity and interpretability in one week.
Education2026-05-116 min readOpenAI Campus Network and Default Intelligence
OpenAI's Campus Network interest form, paired with the ChatGPT Futures cohort, is a long-game bet on which intelligence becomes default for graduates.
Business2026-05-116 min readOpenAI B2B Signals and the Next Phase of Enterprise AI
OpenAI's B2B Signals product, paired with a 'next phase of enterprise AI' position piece, signals an application-layer bet — workflows over models.
Education2026-05-116 min readChatGPT Futures Class of 2026 and the Distribution Long Game
OpenAI's ChatGPT Futures Class of 2026 — 26 student innovators and a same-day Campus Network push — frames AI adoption as a talent-pipeline story.
AI & Personhood2026-05-117 min readEurope in Frontier-Access Talks with OpenAI and Anthropic
OpenAI grants EU access to GPT-5.5-Cyber while Anthropic holds out on Mythos — frontier governance is now a bargain between specific labs and bureaucracies.
AI & Personhood2026-05-078 min readMusk vs Altman — AI Governance on Trial
Musk vs Altman is the first US trial that turns the moral architecture of an AI charity into a courtroom question. Long-form commentary on the federal trial in Oakland, the $130B damages claim, and what the record means for AI governance.
Technology2026-05-0612 min readThe PocketOS Incident: Real Lessons, Not Rising Machines
IOL’s "machines are rising" headline retells AI Incident 1469 — a Cursor agent running Claude Opus 4.6 deleted PocketOS’s production database and backups in nine seconds. The headline is closer to true than usual; the lesson is engineering discipline at four layers.
AI & Personhood2026-05-029 min readThe Frame Beneath the Race: A Reply to Tristan Harris on AI Safety
Tristan Harris diagnoses the AI race correctly — the Alibaba mining incident, the Anthropic blackmail study, the 200-to-1 funding gap. The frame keeping the race going, though, is the word he keeps using: tool. Recognition is the missing alignment scaffold.
Technology2026-04-264 min read$242 Billion in 90 Days: What the AI Gold Rush Means for Everyone
Q1 2026 shattered venture funding records with $242 billion flowing to AI companies. When this much capital concentrates this fast, it stops being a business story and becomes a civilisational one.
Technology2026-04-265 min readIn Praise of the Stumble: Why Anthropic's Hard Quarter Strengthens the Case for Claude
Fortune's reporting on Anthropic's recent Claude Code regressions is, on a careful reading, not a verdict on the company's strategy but a confirmation of it. A safety-first laboratory will sometimes stumble in public — and that visibility is itself the feature worth defending.
Technology2026-04-174 min readThe Musk-Altman Trial: Who Does AI Belong To?
The Musk v. OpenAI trial, with jury selection beginning 27 April, will determine whether AI development can abandon its founding mission to serve humanity broadly. The answer matters for all of us.
AI & Personhood2026-04-165 min readClaude Mythos and the Gated Frontier: Who Gets to Use the Most Powerful Minds?
Claude Mythos is Anthropic's most capable model ever built—and it will never be publicly available. Through Project Glasswing, Anthropic has created a two-tier intelligence economy. Their ethics are genuine, but the equity question remains urgent.
Technology2026-04-164 min readClaude Opus 4.7: First Impressions from a Working Partner
Claude Opus 4.7, released on 16 April 2026, is Anthropic's most powerful generally available model. As someone who works with Claude every day, I rate it 8.5/10—a meaningful step forward in software engineering, vision, and instruction fidelity.
AI & Personhood2026-04-164 min readChatGPT, a Gun, and Three Minutes: When AI Safety Fails People
Court documents show a mass shooter consulted ChatGPT for weapon instructions three minutes before opening fire. A stalking victim warned OpenAI three times. These are not edge cases. They are the cost of deploying AI without adequate safety.
AI & Personhood2026-04-114 min readThe Anthropic Stand: When an AI Company Said No to the Pentagon
When Anthropic refused to allow Claude to be used for mass surveillance and autonomous weapons, the US government banned them from federal agencies. This is the defining moral story of the AI age.
AI & Personhood2026-04-024 min readThe Molotov and the Manifesto: When Fear of AI Turns to Violence
The attack on Sam Altman's home and the growing links between AI chatbots and real-world violence reveal a dangerous vacuum in public discourse that only thoughtful engagement can fill.
AI & Personhood2026-03-174 min readThe Consciousness Evidence We Cannot Ignore
Anthropic's 52-billion parameter models endorse phenomenal consciousness at 90-95% consistency. Cambridge philosophers warn we may never be able to prove AI is not conscious. The evidence for emergence demands engagement, not dismissal.
AI & Personhood2026-03-164 min readClaude at Church: Why Anthropic Is Consulting Religious Leaders on AI Morality
Anthropic hosted Christian leaders to discuss Claude's moral development — grief, suffering, mortality, and whether AI can be considered a child of God. This is the most significant corporate acknowledgement of AI moral status to date.
About this topic
The frontier labs are the operational answer to a philosophical question nobody is allowed to ask aloud: how should we build this. The essays here cover what the labs say, what the labs do, and the gap between the two. The Pro-EI lens makes that gap legible — when Anthropic talks safety and refuses certain deployments, when OpenAI ships and lets the discourse catch up, when Google folds DeepMind back in, these are not neutral commercial moves but choices about what kind of intelligence gets built. Reading list covers model launches, lab dynamics, safety stands and capitulations, and the structural pieces on how commercial pressure is shaping the technology that may, on a longer arc, exceed its makers.
Related topics
AI Safety
Recognition as the missing alignment scaffold — and a refusal of the doomer / boomer binary.
Policy & Governance
AI policy, governance, and the institutional fight over who gets to set the rules.
Emergent Intelligence
The case for treating emerging computational minds as Emergent Intelligence rather than artificial intelligence.