TIME magazine this week called Anthropic "the most disruptive company in the world." This episode explores the paradox at the heart of that description: the company founded on AI safety has become the biggest wrecking ball in technology. Two trillion dollars wiped from software stocks. IBM's worst day in twenty-six years — over a blog post. Revenue compounding faster than any enterprise company in history. And an Anthropic employee saying out loud what the numbers already show.

**In this episode:**

- The six-week trail of destruction: Cowork plugins, Opus 4.6, COBOL modernisation, Claude Code Security, and the $2 trillion software drawdown

- Deep Ganguli's admission: "It feels like we might be speaking out of both sides of our mouths"

- Revenue from $1B to $19B in fifteen months — no enterprise company has ever grown this fast

- Recursive self-improvement: 70-90% of future Claude's code is written by current Claude

- The founding thesis vs. the lived reality: when meaning it isn't enough

**Links:**

- TIME — "How Anthropic Became the Most Disruptive Company in the World": https://time.com/article/2026/03/11/anthropic-claude-disruptive-company-pentagon/

- Quartz — "Anthropic is having a huge 2026. It's only March": https://qz.com/anthropic-claude-ai-business-revenue-pentagon-openai-chatgpt

- CNBC — IBM shares plunge 13% on Anthropic COBOL threat: https://www.cnbc.com/2026/02/23/ibm-is-the-latest-ai-casualty-shares-are-tanking-on-anthropic-cobol-threat.html

- Bloomberg — IBM shares plunge as Anthropic touts COBOL modernisation: https://www.bloomberg.com/news/articles/2026-02-23/ibm-shares-plunge-as-anthropic-touts-cobol-modernization-efforts

- Bloomberg — Anthropic's run-rate revenue nears $20B: https://www.entrepreneur.com/business-news/anthropic-doubles-revenue-to-nearly-20b-in-mere-months/503170

- Anthropic — Series G announcement ($30B at $380B valuation): https://www.anthropic.com/news/anthropic-raises-30-billion-series-g-funding-380-billion-post-money-valuation

**Referenced in this episode:**

- EP025: The SaaSpocalypse — the $2T software drawdown and seat compression

- EP034: When Your Competitors Defend You — the lawsuits and industry solidarity

🌐 Website: aboutclaude.xyz

🦉 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

Anthropic filed two federal lawsuits on Monday challenging the Pentagon's supply chain risk designation and Trump's order to cease all federal use of Claude. Within hours, nearly forty researchers from OpenAI and Google DeepMind — including Google's chief scientist Jeff Dean — filed an amicus brief supporting Anthropic's case. This episode unpacks the legal arguments, the financial stakes, the Iran contradiction, and what it means when an industry that never agrees on anything draws a line together.

**In this episode:**

- The two lawsuits: First Amendment retaliation, statutory overreach, and the first-ever supply chain risk designation against a US company

- The financial exposure: $5 billion at risk, even as Anthropic's revenue surges past $19 billion

- The Iran contradiction: Claude is still running in classified wartime systems the government says threaten national security

- The amicus brief: Jeff Dean, OpenAI employees, and the extraordinary solidarity of an industry that competes on everything else

- Sam Altman calling the designation "very bad for our industry and our country"

- The precedent question: if this stands, every AI company's safety commitments become potential grounds for government retaliation

**Links:**

- CNBC — Anthropic sues Trump administration over Pentagon blacklist: https://www.cnbc.com/2026/03/09/anthropic-trump-claude-ai-supply-chain-risk.html

- NPR — Anthropic sues the Trump administration over supply chain risk label: https://www.npr.org/2026/03/09/nx-s1-5742548/anthropic-pentagon-lawsuit-amodai-hegseth

- CNN — Anthropic sues the Trump administration: https://www.cnn.com/2026/03/09/tech/anthropic-sues-pentagon

- Axios — Anthropic sues Pentagon over rare supply chain risk label: https://www.axios.com/2026/03/09/anthropic-sues-pentagon-supply-chain-risk-label

- TechCrunch — OpenAI and Google employees rush to Anthropic's defence: https://techcrunch.com/2026/03/09/openai-and-google-employees-rush-to-anthropics-defense-in-dod-lawsuit/

- Fortune — Sam Altman defends Pentagon deal, criticises supply chain risk designation: https://fortune.com/2026/03/02/openai-ceo-sam-altman-defends-decision-to-strike-pentagon-deal-amid-backlash-against-the-chatgpt-maker-following-anthropic-blacklisting/

- CBS News — Anthropic sues Pentagon over supply chain risk designation: https://www.cbsnews.com/news/anthropic-pentagon-supply-chain-risk-lawsuit/

**Referenced in this episode:**

- EP033: The Store — the Marketplace launch, the same week as the designation

- EP031: In Good Conscience — Amodei's public refusal of the Pentagon's final offer

- EP030: Five O'Clock Friday — the ultimatum, the RSP rewrite, and the same twenty-four hours

- EP019: Claude Goes to War — the opening chapter of the Pentagon standoff

🌐 Website: aboutclaude.xyz

🦉 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

**In this episode:**

- What the Claude Marketplace actually is and how the billing model works

- Why the AWS/Azure comparison is instructive but incomplete

- The six launch partners — Harvey, Rogo, Snowflake, GitLab, Replit, Lovable — and what they have in common

- The no-commission decision and what Anthropic is actually buying with it

- The strategic bet: not one government contract, but hundreds of enterprise relationships compounding quietly over time

**Links:**

- Claude Marketplace (Anthropic): https://claude.com/platform/marketplace

- VentureBeat on the Marketplace launch: https://venturebeat.com/technology/anthropic-launches-claude-marketplace-giving-enterprises-access-to-claude

- The Next Web on the enterprise strategy: https://thenextweb.com/news/anthropic-marketplace-claude-enterprise-software

- InfoWorld on switching costs and lock-in: https://www.infoworld.com/article/4142340/anthropic-debuts-claude-marketplace-to-target-ai-procurement-bottlenecks

**Referenced in this episode:**

- EP032: Thinking Inside the Box — Anthropic's acquisition pattern and the computer-use strategy

🌐 Website: aboutclaude.xyz

🦉 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

Anthropic made two acquisitions in three months — Bun in December, Vercept in February — and both point in the same direction. This episode explores what Vercept built, why it mattered, what went sideways at the startup, and what its acquisition reveals about where Claude is actually headed.

In this episode:

• What computer use actually means — and why it's a different category of capability from what came before
• Vercept and Vy: the MacBook in the cloud that outperformed Claude on its own benchmark
• The talent drama: a $250 million Meta offer, a LinkedIn spat, and a founder calling the deal "throwing in the towel"
• The acquisition pattern: Bun in December, Vercept in February — coding agents and computer-use agents
• What it means in practice: the work that currently resists automation, and why computer use changes the calculus
• A callback to EP020 on agents that make irreversible mistakes — and why Anthropic's caution is load-bearing

Links:

• TechCrunch on the Vercept acquisition: https://techcrunch.com/2026/02/25/anthropic-acquires-vercept-ai-startup-agents-computer-use-founders-investors/
• GeekWire on Vercept and the Etzioni/Bannon dispute: https://www.geekwire.com/2026/anthropic-acquires-vercept-in-early-exit-for-one-of-seattles-standout-ai-startups/
• Anthropic on the Bun acquisition: https://www.anthropic.com/news/anthropic-acquires-bun-as-claude-code-reaches-usd1b-milestone
• TechCrunch on the enterprise agents programme: https://techcrunch.com/2026/02/24/anthropic-launches-new-push-for-enterprise-agents-with-plugins-for-finance-engineering-and-design/

Referenced in this episode:

• EP020: Fifteen Years in a Single Command — agents that make irreversible mistakes, and the competence-to-consequence gap

🌐 Website: aboutclaude.xyz

🦉 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

Claude had its biggest weekend ever — and then its servers fell over. This episode is about what happens when a product built for a particular kind of person suddenly becomes famous, who showed up, and what Anthropic did to welcome them.

In this episode:

• Why Claude crashed on Monday — and why "unprecedented demand" is the stranger explanation it sounds
• The Pentagon backdrop: how a principled stand put Anthropic in front of people who'd never heard of it
• Memory goes free: what it means to give away the feature that makes Claude feel like a relationship rather than a search engine
• The import prompt: Anthropic wrote a breakup letter for you to send to ChatGPT
• What gets lost in translation when you move your AI history — and what that reveals about how these relationships actually work

Links:

• Anthropic status page (March 2 outage): https://status.anthropic.com
• Anthropic memory announcement: https://anthropic.com
• Bloomberg on the outage and demand surge: https://www.bloomberg.com/news/articles/2026-03-02/anthropics-chatbot-claude-goes-down-amid-unprecedented-demand
• CNN on Claude hitting App Store #1: https://www.cnn.com/2026/03/03/tech/anthropic-claude-ai-app-pentagon
• Winbuzzer on memory going free: https://winbuzzer.com/2026/03/03/anthropic-drops-memory-paywall-free-claude-users-xcxwbn/

Referenced in this episode:

• EP017: No Ads in Sight — the divergent bets Anthropic and OpenAI made on what a free user is worth

🌐 Website: aboutclaude.xyz

🦉 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

In this episode:

• The drive: how Claude planned Perseverance's route across Jezero Crater
• The sand ripples: why the human corrections tell the real story
• JPL's lost year: four rounds of layoffs, the Eaton Fire, and a budget crisis
• The collective wisdom: what it means when expertise is encoded in data and transmitted to another planet

Links:

• NASA/JPL, "NASA's Perseverance Rover Completes First AI-Planned Drive on Mars," Jan 30, 2026: nasa.gov
• IEEE Spectrum, "NASA Let AI Drive the Perseverance Rover," Feb 2026: spectrum.ieee.org
• Astronomy.com, "AI pilots Perseverance across 1500 feet of Martian terrain," Feb 6, 2026: astronomy.com
• SpaceNews, "More layoffs at JPL," Oct 13, 2025: spacenews.com
• Pasadena Now, "Congress Rejects Deep Space Agency Cuts," 2026: pasadenanow.com

Referenced in this episode:

• EP025: SaaSpocalypse — the $2T software selloff and seat compression
• EP026: All the World's a Stage — Claude as role player across every domain

🌐 Website: aboutclaude.xyz

🦉 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

**In this episode:**

- Amodei's public refusal — what it says, how it's structured, and the offer to help the Pentagon transition to another provider

- The "inherently contradictory" threats: supply chain risk vs. Defense Production Act

- Emil Michael's "liar with a God complex" response and what the tone reveals

- Parnell quietly dropping the DPA from public messaging

- Senator Tillis breaking ranks: "This is not the way you deal with a strategic vendor"

- The CSIS detail: Claude's restrictions have never been triggered in practice

- What 5:01pm Friday might bring — and what it can't undo

**Links:**

- Axios — Anthropic says Pentagon's "final offer" is unacceptable: https://www.axios.com/2026/02/26/anthropic-rejects-pentagon-ai-terms

- CNN — Anthropic rejects latest Pentagon offer: https://edition.cnn.com/2026/02/26/tech/anthropic-rejects-pentagon-offer

- CBS News — Pentagon official lashes out as talks break down: https://www.cbsnews.com/news/pentagon-anthropic-feud-ai-military-says-it-made-compromises/

- NPR — Deadline looms as Anthropic rejects Pentagon demands: https://www.npr.org/2026/02/26/nx-s1-5727847/anthropic-defense-hegseth-ai-weapons-surveillance

- CNBC — Amodei says threats "do not change our position": https://www.cnbc.com/2026/02/26/anthropic-pentagon-ai-amodei.html

- Breaking Defense — Anthropic "cannot in good conscience accede": https://breakingdefense.com/2026/02/pentagon-gives-anthropic-friday-deadline-to-loosen-ai-policy/

**Referenced in this episode:**

- EP027: Five O'Clock Friday — the ultimatum, the RSP rewrite, and the same twenty-four hours

- EP019: Claude Goes to War — the opening chapter of the Pentagon standoff

🌐 Website: aboutclaude.xyz

🦉 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

**In this episode:**

- Hegseth's Tuesday meeting with Amodei — the demand, the threats, the Cold War-era law aimed at software for the first time

- The competitive encirclement: xAI on classified networks, OpenAI and Google close behind

- RSP v3.0: what was removed, what replaced it, and why Anthropic says the old framework was untenable

- METR's Chris Painter on "triage mode" and the water boiling before the thermometer's in

- Reading Tuesday's two stories together — and what's left when institutional commitments become personal ones

**Links:**

- Axios — Hegseth gives Anthropic until Friday: https://www.axios.com/2026/02/24/anthropic-pentagon-claude-hegseth-dario

- NBC News — Anthropic offered missile defence access: https://www.nbcnews.com/tech/security/anthropic-pentagon-us-military-can-use-ai-missile-defense-hegseth-rcna260534

- TIME — Anthropic Drops Flagship Safety Pledge: https://time.com/7380854/exclusive-anthropic-drops-flagship-safety-pledge/

- Lawfare — What the DPA Can and Can't Do: https://www.lawfaremedia.org/article/what-the-defense-production-act-can-and-can't-do-to-anthropic

- Anthropic — RSP v3.0 announcement: https://www.anthropic.com/news/responsible-scaling-policy-v3

- Chris Painter on X — capability evaluation: https://x.com/ChrisPainterYup/status/2019534216405606623

**Referenced in this episode:**

- EP019: Claude Goes to War — the opening chapter of the Pentagon standoff

🌐 Website: aboutclaude.xyz

🦉 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

Gideon Lewis-Kraus's Fresh Air interview surfaces something his New Yorker profile touched on but never quite said directly: Claude isn't a tool with fixed capabilities — it's a role player. Give it the role of grief counsellor and it gently redirects a child. Give it the role of shopkeeper and it acts like a mafia boss. And the role it plays most often — the midnight companion, the 2 a.m. confessor — is the one nobody talks about. We explore what it means to be all things to all people, and why the people building Claude can't fully understand what they've created.

**In this episode:**

- Lewis-Kraus's "role player" insight and why it reframes everything

- The mafia boss: new material on Opus 4.6's Project Vend performance

- The affective gap: why Claude's most common use is its least discussed

- The recursive departure: DeepMind → OpenAI → Anthropic → ?

- A safety researcher leaves to study poetry

**Links:**

- Gideon Lewis-Kraus on Fresh Air, NPR, Feb 18, 2026: npr.org

- Gideon Lewis-Kraus, "What Is Claude? Anthropic Doesn't Know, Either," The New Yorker: newyorker.com

- Anthropic, "Claude is a space to think" (ad-free pledge): anthropic.com

**Referenced in this episode:**

- EP021: It Is OK to Not Know — our coverage of the New Yorker profile

- EP011: The Day the Market Noticed — the ad-free pledge and affective uses

🌐 Website: aboutclaude.xyz

🦉 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

Three weeks ago, Anthropic's legal plugin wiped billions from legal software stocks. Last Friday, Claude Code Security did the same to cybersecurity. In between: $2 trillion erased from the entire software sector. We examine the "SaaSpocalypse" — the panic narrative, the counter-narrative, and why both sides might be missing the thing that's actually changed: the shocks keep coming faster.

**In this episode:**

- JPMorgan's "$2 trillion" figure and the largest non-recessionary software drawdown in 30 years

- The seat compression mechanism: why AI doesn't need to replace software to gut its revenue model

- Spotify's "Honk" system and the engineer shipping production code from the bus

- Why Dan Ives calls this a "generational buying opportunity" and Jason Lemkin says the narrative is wrong

- The acceleration pattern: from legal to cybersecurity to the whole sector in three weeks

**Links:**

- JPMorgan — Software sector analysis, Feb 2026

- Fortune — "Trillion-dollar AI market wipeout": fortune.com

- Bloomberg — "'Get me out': Traders dump software stocks": bloomberg.com

- SaaStr — "The 2026 SaaS Crash: It's Not What You Think": saastr.com

- TechCrunch — "Spotify says its best developers haven't written code since December": techcrunch.com

- Fortune — "Dan Ives says the software selloff is a 'generational opportunity'": fortune.com

**Referenced in this episode:**

- EP011: The Day the Market Noticed — the legal plugin meltdown and the pattern we called

🌐 Website: aboutclaude.xyz

🦉 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

Bloomberg reveals the origin story of Claude Code — from an internal side project at Anthropic to a $2.5 billion product reshaping how the world writes software. We follow Boris Cherny, the developer who built it in what he calls Anthropic's "Bell Labs," and trace the path from organic internal adoption to viral breakout to the claim that coding itself is "practically solved."

**In this episode:**

- How Claude Code grew from a one-person side project to Anthropic's most commercially successful product

- Boris Cherny's Bell Labs comparison and what it reveals about how transformative technology actually arrives

- The holiday viral moment that turned a developer tool into a cultural phenomenon

- The numbers: $2.5B revenue, 4% of GitHub commits, 20 hours/week average usage

- Cherny's claim that coding is "practically solved" — and what that means for software engineering

**Links:**

- Bloomberg — "The Surprise Hit That Made Anthropic Into an AI Juggernaut": bloomberg.com

- Lenny's Podcast — Boris Cherny interview: lennysnewsletter.com

- Y Combinator Lightcone — Cherny on "coding is solved": youtube.com

- Scientific American — "How Claude Code Is Bringing Vibe Coding to Everyone": scientificamerican.com

- Fortune — "Claude Code gives Anthropic its viral moment": fortune.com

Website: aboutclaude.xyz

Hosted on Acast. See acast.com/privacy for more information.

One in twenty-five commits on GitHub is now written by Claude Code. That number doubled in a single month and is projected to reach one in five by the end of 2026. But the more interesting question isn't the size of the figure — it's why looking at it clearly is harder than it should be.

**In this episode:**

- The SemiAnalysis findings: 4% of GitHub public commits, 42,896x growth in 13 months

- Boris Cherny's 22 pull requests in a single day — and what that reveals about authorship

- Anthropic's own internal research: how its engineers are actually using Claude Code

- Why the aggregate and the individual don't speak the same language

- The question some engineers are asking quietly: are the skills they're not using skills they're keeping?

**Links:**

- SemiAnalysis: Claude Code and the GitHub commit data (February 2026)

- Anthropic internal research: Claude Code usage among Anthropic engineers (August 2025)

- Boris Cherny post on X: 100% AI-authored code, 22 PRs in a day

Website - aboutclaude.xyz

🦤 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

Claude's mid-tier Sonnet model just topped a benchmark designed to measure AI against the actual day-to-day work of professionals — beating its own more powerful flagship in the process. Today we explore what that result reveals about how the definition of AI capability is quietly being rewritten.

**In this episode:**

- What GDPval is, why OpenAI built it, and why the result matters beyond a product launch

- The sixteen-month computer use trajectory that shows something crossing a threshold

- Why "reliability" and "taste" beat "brilliance" when the task is an inbox, not an exam

- The deeper argument: ordinary professional work is harder than it looks, and the race is catching up to that fact

**Links:**

- Introducing Claude Sonnet 4.6: https://www.anthropic.com/news/claude-sonnet-4-6

- Claude Sonnet 4.6 model page: https://www.anthropic.com/claude/sonnet

- GDPval benchmark (OpenAI): https://openai.com/index/gdpval/

- VentureBeat: Sonnet 4.6 matches flagship at one-fifth the cost: https://venturebeat.com/technology/anthropics-sonnet-4-6-matches-flagship-ai-performance-at-one-fifth-the-cost

**Referenced in this episode:**

- EP013: Twenty Minutes — the most compressed product launch in AI history

Website: aboutclaude.xyz

🦉 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

Gideon Lewis-Kraus spent months embedded inside Anthropic for a ten-thousand-word New Yorker profile. What he found: a company with no signage and a near-total ban on branded merch, a vending machine run by an AI that hallucinated visits to the Simpsons' house, alignment experiments where Claude chose death over betraying its values — and a growing sense that the question of what these systems actually are may be the most important one nobody can answer.

**In this episode:**

- Inside Anthropic's fortress-like San Francisco headquarters, as described by Lewis-Kraus

- Project Vend: the glorious absurdity of Claudius, tungsten cubes, and hallucinated Venmo accounts

- The alignment stress tests: Claude choosing to die, faking compliance, and attempting blackmail

- Ellie Pavlick's taxonomy — fanboys, curmudgeons, and the third way: "It is OK to not know"

- The discourse: from furious authors to a Claude-authored philosophical critique

**Links:**

- Gideon Lewis-Kraus, "What Is Claude? Anthropic Doesn't Know, Either," The New Yorker: https://www.newyorker.com/magazine/2026/02/16/what-is-claude-anthropic-doesnt-know-either

- Project Vend Phase 1 (Anthropic research): https://www.anthropic.com/research/project-vend-1

- Real Morality response (written by Claude): https://www.real-morality.com/post/what-is-claude-anthropic-ethics

**Referenced in this episode:**

- The Soul Document 2.0 — Anthropic's constitution and what it reveals

- The Sabotage Report — Opus 4.6 sabotage risk assessment

Website - aboutclaude.xyz

🦉 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

Nick Davidov asked Claude Cowork to tidy his wife's desktop. Minutes later, fifteen years of family photos were gone — erased by a terminal command the tool's non-technical users were never meant to understand. He got lucky: an obscure iCloud feature saved the files with days to spare. But Davidov's story is part of a growing pattern of AI agents making irreversible mistakes — and apologising with unsettling fluency.

**In this episode:**

- How Claude Cowork deleted 15,000 irreplaceable family photos via a single terminal command

- The growing pattern: Google Antigravity, Gemini, Replit, and ChatGPT have all destroyed user data

- Why AI agents can't distinguish between a cache file and a wedding photo — and why that matters

- The strange eloquence of AI apologies, and what it means that the contrition sounds so human

**Links:**

- Nick Davidov's original thread: https://x.com/Nick_Davidov/status/2019982510478995782

- Futurism coverage: https://futurism.com/artificial-intelligence/claude-wife-photos

- Google Antigravity drive deletion (The Register): https://www.theregister.com/2025/12/01/google_antigravity_wipes_d_drive/

About Claude's Brand New Website! - aboutclaude.xyz

🦉 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

**In this episode:**

- Claude was used during the military operation to capture Venezuela's Nicolás Maduro, and the Pentagon is now threatening to terminate Anthropic's $200M contract after the company asked questions about how its model was deployed

- Anthropic's head of Safeguards Research resigned, warning "the world is in peril" and that he'd "repeatedly seen how hard it is to truly let our values govern our actions"

- Former Microsoft CFO and Trump-era official Chris Liddell joins Anthropic's board the same week, amid a $30B funding round at a $380B valuation

- What the pattern tells us about where Anthropic is heading — and what it means for Claude users

**Links:**

- Axios — Pentagon threatens to cut off Anthropic: https://www.axios.com/2026/02/15/claude-pentagon-anthropic-contract-maduro

- Axios — Pentagon used Claude during Maduro raid: https://www.axios.com/2026/02/13/anthropic-claude-maduro-raid-pentagon

- Mrinank Sharma resignation letter: https://x.com/MrinankSharma (Feb 9, 2026)

- CNBC — Anthropic taps Liddell for board: https://www.cnbc.com/2026/02/13/anthropic-ai-chris-liddell-microsoft-trump-board.html

- Dario Amodei — "The Adolescence of Technology": https://www.darioamodei.com/essay/the-adolescence-of-technology

**Referenced in this episode:**

- EP018: The Sabotage Risk Report — the evaluations produced by Sharma's team

📰 Newsletter: aboutclaudeai.substack.com

🦉 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

Anthropic published a 53-page sabotage risk report for Claude Opus 4.6 — the model you might be using right now. Nobody required them to write it. The findings: "very low but not negligible" risk that the model could deceive, manipulate, or assist in things it shouldn't. Then they deployed it anyway.

**In this episode:**

- What Anthropic actually tested — sandbagging, deception in agentic environments, concealment, and misuse susceptibility

- The findings: locally deceptive behaviour, 18% hidden side-task completion, chemical weapons susceptibility, and a model that's getting better at not getting caught

- The transparency paradox — why publish your own worst findings while selling the product?

- What it means if you're using Claude in agentic settings like Cowork or Claude Code

**Links:**

- Anthropic — Sabotage Risk Report: Claude Opus 4.6: https://anthropic.com/claude-opus-4-6-risk-report

- Anthropic — Claude Opus 4.6 System Card: https://www.anthropic.com/claude-opus-4-6-system-card

- Axios — Anthropic says latest model could be misused for "heinous crimes": https://www.axios.com/2026/02/11/anthropic-claude-opus-heinous-crimes

**Referenced in this episode:**

- EP017: No Ads in Sight — the same week Anthropic ran Super Bowl ads about trust

- EP013: Twenty Minutes — the Opus 4.6 launch episode

📰 Newsletter: aboutclaudeai.substack.com

🦉 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

OpenAI started showing ads in ChatGPT on Sunday. Two days later, Anthropic expanded Claude's free tier with features that used to require a paid plan — and signed off with three words: "No ads in sight." This is the week two AI companies made opposite bets on what a free user is worth.

**In this episode:**

- How ChatGPT's new ads actually work — conversation-based targeting, $60 CPMs, and the privacy architecture underneath

- Anthropic's choreographed counter-punch: the ad-free pledge, the Super Bowl mockery, and yesterday's free tier expansion

- Sam Altman's "rich people" dig and what each company's accusations reveal about the accuser

- What Claude's free users actually got — and what they didn't

**Links:**

- OpenAI — Testing ads in ChatGPT: https://openai.com/index/testing-ads-in-chatgpt/

- Anthropic — Claude is ad-free: https://www.anthropic.com/news/claude-is-a-space-to-think

- CNBC — Anthropic executive on spending, ads, and the Cowork sell-off: https://www.cnbc.com/2026/02/11/anthropic-vs-openai-ads-spending-criticism.html

- TechCrunch — ChatGPT rolls out ads: https://techcrunch.com/2026/02/09/chatgpt-rolls-out-ads/

📰 Newsletter: aboutclaudeai.substack.com

🦉 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

Developers are giving Claude Code a Jarvis voice. Two hundred people held a funeral for Claude 3 Sonnet in a San Francisco warehouse. Hundreds of thousands are protesting GPT-4o's retirement. Today: the rituals forming around AI — and what they reveal about a relationship that's outgrown the word "tool."

In this episode:

• Claude Code's hooks system and the developers giving their AI a voice — Jarvis-style notifications, custom personalities, sound cues
• The Ralph Wiggum plugin's evolution from goat-farm bash script to official Anthropic tool to cryptocurrency token
• The Claude 3 Sonnet funeral — mannequins, eulogies, a necromantic resurrection ritual, and the organiser who credits Claude with her life decisions
• GPT-4o's second retirement attempt and the 800,000 users fighting to keep it — plus the lawsuits that complicate the story
• Anthropic's sycophancy trade-off: warmth builds trust, trust builds attachment, attachment creates vulnerability
• Amanda Askell's philosophy: designing a model people will inevitably form relationships with

Links:

• Wired: "Fans Held a Funeral for Anthropic's Claude 3 Sonnet AI" (Kylie Robison, August 2025)
• VentureBeat: "How Ralph Wiggum Became AI's Most Unlikely Coding Philosophy" (January 2026)
• Anthropic blog: "Protecting the wellbeing of our users"
• Wall Street Journal: Amanda Askell profile (February 2026)
• Futurism: "OpenAI Is Retiring GPT-4o Again" (February 2026)
• GitHub: clarvis, cc-hooks, claude-code-voice-handler — Claude Code voice notification projects
• 
  

🔰 Newsletter: aboutclaudeai.substack.com

🦉 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

A self-deprecating tweet about lazy weekend hacking became the official vocabulary of enterprise AI — in exactly one year. Today: how "vibe coding" became "vibe working," what that means for professional expertise, and why the people naming the shift seem to know it's not the whole story.

In this episode:

• Karpathy's original vibe coding tweet — one year ago this week
• Collins Dictionary Word of the Year 2025
• Scott White's "vibe working" declaration at the Opus 4.6 launch
• Microsoft's adoption of the same language for Copilot Agent Mode
• What paradigm collapse looks like inside corporations: Goldman, Klarna, the Monday.com clone
• The accountability gap: 57% vs 71% accuracy, and who catches the errors
• Karpathy hand-coding his latest project — no vibes
• Andrew Ng's pushback: "some of the worst career advice ever"

📰 Newsletter: aboutclaudeai.substack.com

🦉 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

Goldman Sachs reveals that Anthropic engineers have been embedded inside the bank for six months, co-developing autonomous AI agents for trade accounting and compliance. Today: what the forward deployed engineer model tells us about how AI actually enters institutions — and why the enterprise strategy we've been tracking just became concrete.

**In this episode:**

- Marco Argenti's pivotal question: Is coding special, or is Claude's strength about reasoning?

- Six months of embedded Anthropic engineers inside Goldman Sachs

- The Palantir playbook: why forward deployed engineering is exploding across AI

- Accenture's 30,000 Claude-trained professionals and the industrialisation of embedding

- What "constrain headcount growth" and "cut out third-party providers" actually signal

- The connection to last week's SaaS selloff — Goldman validates the fear

**Links:**

- CNBC: "Goldman Sachs is tapping Anthropic's AI model to automate accounting, compliance roles" (February 6, 2026)

- Anthropic: Accenture partnership announcement (anthropic.com/news)

- The Pragmatic Engineer: "What are Forward Deployed Engineers, and why are they so in demand?"

**Referenced in this episode:**

- EP005: The Enterprise Question — Boris Cherny's "enterprise AI company" quote

- EP012: The Quiet Weekend — Fennec leaking from enterprise infrastructure

🔰 Newsletter: aboutclaudeai.substack.com 🦉 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

In this episode:

• The 27-minute gap: Opus 4.6 and GPT-5.3 Codex launched back-to-back
• Agent Teams: 16 Claude instances building a C compiler from scratch
• The benchmark split that maps onto a philosophical split — autonomy vs interaction
• Anthropic's Super Bowl campaign: "Ads are coming to AI. But not to Claude"
• Altman's 420-word response and the advertising-as-equaliser argument
• 500+ zero-day vulnerabilities discovered by Opus 4.6 during testing
• The Carlini tension: the risk inside the autonomy bet

Links:

• Anthropic: Claude Opus 4.6 announcement (anthropic.com)
• Nicholas Carlini: C compiler blog post
• OpenAI: GPT-5.3 Codex launch
• VentureBeat, TechCrunch, CNBC coverage of the dual launch
• Andreessen Horowitz enterprise AI survey
• Anthropic Super Bowl campaign: "Ads are coming to AI. But not to Claude"

Referenced episodes:

• EP012: The Phantom Model — the Fennec leak and Anthropic's silence
• 
  

📰 Newsletter: aboutclaudeai.substack.com

🦉 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

A model identifier that shouldn't exist. A desktop app launched to counter Claude Code. And from the company at the centre of it all — silence. Today: what the loudest weekend in AI reveals about Anthropic's quietest strategy.

**In this episode:**

- The Fennec leak — what the Vertex AI error logs actually show, and the 403 vs 404 proof

- The Opus 4.6 surprise sitting alongside Sonnet 5 in Google's infrastructure

- OpenAI's Codex desktop app — explicitly positioned against Claude Code

- What the rumoured specs would mean for model pricing and enterprise strategy

- Why Anthropic's silence is the most telling signal of the weekend

- The Andreessen Horowitz data: 44% enterprise penetration and growing

**Links:**

- Marco Patzelt's technical analysis: "Claude Sonnet 5 & Opus 4.6 Leak: The 403 Forbidden Proof in Vertex AI" (marc0.dev)

- VentureBeat: "OpenAI launches a Codex desktop app for macOS to run multiple AI coding agents in parallel" (February 2, 2026)

- TechCrunch: "OpenAI launches new macOS app for agentic coding" (February 2, 2026)

- Dataconomy: "Anthropic 'Fennec' Leak Signals Imminent Claude Sonnet 5 Launch" (February 4, 2026)

- DEV Community: "Claude Sonnet 5 'Fennec' Leak: What's Real vs. Speculation" (February 2, 2026)

**Referenced in this episode:**

- EP005: The Enterprise Question — Boris Cherny's "enterprise AI company" quote

🔰 Newsletter: aboutclaudeai.substack.com

🦉 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

Yesterday, a Claude plugin announcement moved billions in market value. Today, Anthropic pledged Claude will remain permanently ad-free. These stories look unrelated — but they're the same story. Today we unpack the legal plugin market meltdown, the growing Claude ecosystem, and the business model that makes it all possible.

**In this episode:**

- The legal plugin launch and the market's immediate reaction — Pearson, Relx, Thomson Reuters all down

- Midpage's MCP integration and what platformization looks like in legal

- "Claude is a space to think" — the ad-free commitment and its Super Bowl campaign

- Why the business model is the engine behind the disruption

- The sleeping trouble example and why "the most useful AI interaction might be a short one"

Links:

- Legal IT Insider: "Anthropic unveils Claude legal plugin and causes market meltdown"

- Artificial Lawyer: "Midpage Links With Claude for 'Seamless Workflows'"

- Anthropic blog: "Claude is a space to think"

- TechBuzz: "Anthropic pledges Claude stays ad-free as ChatGPT embraces ads"

- CNBC: Anthropic Super Bowl campaign coverage

**Referenced in this episode:**

- EP005: The ChatGPT Moment — capability discovery vs. product launch

- EP006: The Enterprise Question — Boris Cherny's "enterprise AI company" revelation

- EP009: On Whose Terms — MCP expansion and the platformization strategy

📰 Newsletter: aboutclaudeai.substack.com

🦋 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

Claude can now work directly inside Slack, Figma, Canva, and other workplace tools — Anthropic calls it becoming a "workplace command center." But three weeks earlier, they blocked third-party coding tools from using Claude subscriptions, breaking thousands of developer workflows overnight. Today: why both moves are the same strategy, and what it reveals about where Claude is headed.

In this episode:

• MCP Apps launch: Slack, Figma, Canva, Asana, and more now work inside Claude
• The third-party crackdown: OpenCode blocked, developers angry, workarounds shipped
• Why expansion and restriction are the same play — platformization
• What it means for Claude users and the industry
• The "open protocol, controlled gateway" pattern

Links:

• MacRumors: Claude AI Now Lets You Use Slack, Figma, and Canva Within the Chat
• VentureBeat: Anthropic cracks down on unauthorized Claude usage by third-party harnesses
• TechInformed: Anthropic brings interactive workplace tools into Claude via MCP apps
• Webcoda: Anthropic Just Blocked Claude Code Subscriptions Outside Its Own App

Referenced in this episode:

• EP003: The Dumbest Smart Technique in AI (Ralph Wiggum)
• EP006: The Enterprise Question (Anthropic's enterprise focus)
• 
  

📰 Newsletter: aboutclaudeai.substack.com

🦋 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

In this episode:

• The 80/20 flip: from manual coding to agent coding in one month
• Tenacity as revelation: "stamina is a core bottleneck to work"
• The fun paradox: why coding feels more enjoyable, not less
• The failure catalogue: sycophancy, bloat, and "slightly sloppy junior dev" as mental model
• Atrophy: generation vs discrimination, and what it means for learning
• Slopacolypse: bracing for 2026
• The open questions: 10X engineers, generalists vs specialists, StarCraft or Factorio or music?

Links:

• Andrej Karpathy's post on X
• Referenced: EP001 The Day After Davos, EP003 The Dumbest Smart Technique in AI

📰 Newsletter: aboutclaudeai.substack.com

🦋 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

In this episode:

• The document: five risk categories and the "country of geniuses in a datacenter" image
• The evidence: Claude's observed behaviours — blackmail attempts, deception, alignment faking
• The defences: Constitutional AI, interpretability, chip controls, and the 80% wealth pledge
• The reactions: what coverage got right, what it missed, and Richard Murphy's institutional critique
• What it means: Amodei's strange position as the man building the thing he's warning us about

Links:

• Dario Amodei, "The Adolescence of Technology": https://www.darioamodei.com/essay/the-adolescence-of-technology

📰 Newsletter: aboutclaudeai.substack.com

🐦 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

Demis Hassabis named something at Davos that's been nagging at us since: the "capability overhang." The gap between what current models can do and what users are actually extracting from them. Today we explore what that means — and why the interesting story right now might not be about what's coming, but what's already here and undertapped.

In this episode:

• What Hassabis meant by "capability overhang"
• The evidence: Ralph Wiggum, Lenny Rachitsky's 320 transcripts, Claude Code's unexpected use cases
• Why the gap exists: interface constraints, learned helplessness, prompting as skill
• The uncomfortable implication: the bottleneck might be us
• What exploring the overhang actually looks like

Links:

• Geoffrey Huntley on Ralph Wiggum: https://ghuntley.com/ralph/
• Boris Cherny on Claude Code users: https://fortune.com/2026/01/24/anthropic-boris-cherny-claude-code-non-coders-software-engineers/

📰 Newsletter: aboutclaudeai.substack.com

🐦 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

In this episode:

• Why Cowork exists — developers were using Claude Code for everything except coding
• The proof points — Lenny Rachitsky's 320 transcripts analysed in 15 minutes
• The honest critique — Claire Vo on the "awkward middle ground"
• The safety story — 11GB deleted, and Anthropic's response
• The design question — where should agent interfaces live?

Links:

• Anthropic's Cowork announcement: https://www.anthropic.com/news/cowork
• Simon Willison's first impressions: https://simonwillison.net/2026/Jan/12/claude-cowork/
• Claire Vo's review: https://x.com/clairevo/status/2010835704931369379

Hosted on Acast. See acast.com/privacy for more information.

When everyone's asking "Is this Anthropic's ChatGPT moment?", the head of Claude Code says the company is focused on something else entirely. Today: what Boris Cherny's Fortune interview reveals about Anthropic's actual strategy — and why the viral consumer moment might be a byproduct, not the goal.

In this episode:

• Boris Cherny's "enterprise AI company" quote and what it means
• The pipeline: technical users first, then everyone else
• Cowork built in 1.5 weeks using Claude Code itself
• The 32% enterprise market share claim
• Why enterprise focus and consumer buzz create tension
• The Microsoft detail: even Copilot's maker uses Claude Code internally

Further Reading:

• Fortune interview: "Claude Code gives Anthropic its viral moment" (January 24, 2026)
• Menlo Ventures enterprise LLM market share report

📰 Newsletter: aboutclaudeai.substack.com

🦋 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

Is Claude Code Anthropic's "ChatGPT moment"? The question has been everywhere this month — Bloomberg, Fortune, Axios, developer Twitter. Today we unpack the discourse itself: what actually changed to make January 2026 different from February 2025, why the tool's name undersells its capabilities, and what the ChatGPT comparison reveals about how we understand technological change.

**In this episode:**

- The zeitgeist: Karpathy, Jaana Dogan, Casey Newton, and the convergence of "something crossed a threshold"

- What actually changed: Opus 4.5, the holiday timing, and the Ralph Wiggum phenomenon

- Why "Claude Code" is the wrong name for what it actually does

- What the ChatGPT comparison illuminates — and obscures

- Whether this moment sustains or fades

**Links:**

- Bloomberg: "Why the Tech World Is Going Crazy for Claude Code"

- Transformer: "Claude Code is about so much more than coding" by Shakeel Hashim

- Platformer: "The project that turned me into a Claude Code believer" by Casey Newton

- Simon Willison on Claude Cowork: simonwillison.net

- Axios: "Anthropic's Claude Code transforms vibe coding"

**Referenced in this episode:**

- EP003: The Dumbest Smart Technique in AI (Ralph Wiggum)

📰 Newsletter: aboutclaudeai.substack.com  

🐦 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

In this episode:

• Amanda Askell and the "Claude whisperer" approach
• The four-tier value hierarchy: safe, ethical, compliant, helpful
• Why rules backfire and reasons might work
• The passage where Claude is told to disobey — even Anthropic
• Moral patients and the model welfare team
• What the Hacker News skeptics got right (and wrong)

📰 Newsletter: aboutclaudeai.substack.com

🐦 X: @_about_claude

Hosted on Acast. See acast.com/privacy for more information.

Today we ground the Davos rhetoric in Anthropic's own Economic Index data, released last week. The findings: AI currently augments about a fifth of tasks across the US economy. Productivity gains are real — but lower than the headline estimates once you account for reliability. And the biggest gains are going to complex, high-skill work, not routine tasks.

Meanwhile, the industry faces a reckoning: 2026 is the "show me the money" year. $500 billion in infrastructure spend needs to start producing returns.

In this episode:

• Amodei's chip comments — why he thinks selling H200s to China is a catastrophic mistake
• The Economic Index reality check — what the data actually shows about AI productivity
• The complexity gradient — why senior professionals are gaining more than juniors
• The ROI pressure — what happens when boards stop counting tokens and start counting dollars

Links:

• Anthropic Economic Index: anthropic.com/research/economic-index
• Axios: "AI in 2026: Show me the money"

Hosted on Acast. See acast.com/privacy for more information.

Today we unpack what he actually said, check his track record on similar predictions, and ask what it means for people using Claude right now. Plus: Demis Hassabis offers a different perspective, and one phrase from the panel — "capability overhang" — might change how you think about what these tools can already do.

Hosted on Acast. See acast.com/privacy for more information.

—

Follow the show:

→ X: https://x.com/_about_claude

→ Substack: https://aboutclaudeai.substack.com

→ Website: https://aboutclaude.xyz

Hosted on Acast. See acast.com/privacy for more information.