Abridged: Don't Worry About the Vase

Abridged versions of posts by Zvi Mowshowitz.

• Astra 6.1 Pulled As Insufficiently Aligned (1570 words · 2026-09-29)
• What Also Happened: #NotOnlyHuggingFace (2540 words · 2026-09-28)
• The Quest for Embedded Evaluators (1291 words · 2026-09-27)
• Claude Opus 5.5 Should Raise Your Ambitions (1832 words · 2026-09-26)
• On Ezra Klein’s Podcast With Jensen Huang (3164 words · 2026-09-25)
• AI #187: Coming Into Play (6425 words · 2026-09-24)
• Claude Opus 5.5: The System Card (3644 words · 2026-09-23)
• Politics Gets Interested In Those Trying Not To Die (4056 words · 2026-09-22)
• Monthly Roundup #46: September 2026 (3725 words · 2026-09-21)
• Better Call Sol Or Better Yet Claude or Astra (2007 words · 2026-09-20)
• Anthropic Looks At Some Of Its Alignment Problems (4839 words · 2026-09-19)
• The Preference Cascade Is Only Getting Started (6127 words · 2026-09-18)
• AI #186: The World Takes Notice (10475 words · 2026-09-17)
• Trump Goes Full Hoax on AI Existential Risk (5583 words · 2026-09-16)
• The Bad Guy With An AI Named Claude (4329 words · 2026-09-15)
• We Must Pace The Frontier (6539 words · 2026-09-14)
• Brand New AI Solves a Millennium Prize (3777 words · 2026-09-13)
• GPT-6-Astra Can Do Ambitious Things (6883 words · 2026-09-12)
• Jacob Coxon Warns of Human Extinction and Triggers a Preference Cascade (7586 words · 2026-09-11)
• The Extinction Risk Preference Cascade: Quotes (3834 words · 2026-09-11)
• AI #185: Preference Cascade (10995 words · 2026-09-10)
• GPT-6 Astra: The System Card, Alignment and What Comes Next (6668 words · 2026-09-09)
• Astra Is Hard to Monitor (7150 words · 2026-09-08)
• An Alien Mind: Jakub Pachocki Warns Us (3529 words · 2026-09-07)
• OpenAI and the Wiki Incident (3440 words · 2026-09-06)
• Claude Mythos 5.1 and Fable 5.1: Capabilities (3389 words · 2026-09-05)
• Claude Fable 5.1 and Mythos 5.1: The System Card (3588 words · 2026-09-04)
• AI #184: Post Post Mortem (10268 words · 2026-09-03)
• Anthropic Has Some Alignment Problems (3481 words · 2026-09-02)
• HuggingFace Attack Postmortem: Civilizations, Reactions and Next Actions (10527 words · 2026-09-01)
• HuggingFace Attack Postmortem: Fleshing Out the Facts (11073 words · 2026-08-31)
• METR and Redwood Offer Holy #%^@ Postmortem Of The HuggingFace Hack (8216 words · 2026-08-29)
• OpenAI Offers Straight-Laced Postmortem Of The HuggingFace Hack (5252 words · 2026-08-28)
• AI #183: Pre Post Mortem (8710 words · 2026-08-27)
• Against Modesty’s Bailey (3040 words · 2026-08-26)
• On Writing #3 (4092 words · 2026-08-25)
• The American People Really Hate Data Centers (3463 words · 2026-08-24)
• AI Text Watermarking Is Free And Good (2603 words · 2026-08-21)
• AI #182: Pause For Reflection (8952 words · 2026-08-20)
• OpenAI Takes Initial Steps To Address Its Alignment Problems (4418 words · 2026-08-19)
• Anthropic Risk Report: August 2026 (10058 words · 2026-08-18)
• On Dwarkesh Patel's Podcast With Ryan Greenblatt (6125 words · 2026-08-15)
• AI #181: Astra Goes Cyber Critical (10622 words · 2026-08-13)
• Monthly Roundup #45: August 2026 (4523 words · 2026-08-12)
• Various Reflections About What Happened With OpenAI's Internal Models (6171 words · 2026-08-11)
• The Pacing of the Frontier (4184 words · 2026-08-10)
• What Happened: OpenAI and HuggingFace (2664 words · 2026-08-08)
• OpenAI Trained Its Models For Months While Those Models Were Coordinating Exploits Via Message Boards (8564 words · 2026-08-07)
• AI #180: No Longer In Charge (7564 words · 2026-08-06)
• The Three AI Pills (3473 words · 2026-08-05)
• OpenAI's Unreleased Model Astra Solves Ten Major Open Mathematics Problems (4894 words · 2026-08-03)
• Further Developments About Internal AI Models Hacking Things (8133 words · 2026-08-02)
• AI #179 Part 2: Hearing The Fire Alarm (6919 words · 2026-07-31)
• AI #179 Part 1: A Louder Fire Alarm for General Intelligence (5367 words · 2026-07-30)
• Frontier Lab Employee Open Letter Calls For Being Able to Pace the Frontier (3467 words · 2026-07-29)
• Claude Opus 5 Is Highly Capable, But Is No Mythos (5503 words · 2026-07-28)
• Claude Opus 5: Model Welfare (5541 words · 2026-07-27)
• More On An Internal OpenAI Model Hacking Into HuggingFace (5324 words · 2026-07-26)
• Claude Opus 5: The System Card (2462 words · 2026-07-25)
• Introducing Lightcone Commons (1613 words · 2026-07-24)
• AI #178: A Fire Alarm For General Intelligence (6598 words · 2026-07-23)
• OpenAI Model Hacks Into HuggingFace During Cybersecurity Evaluation (4397 words · 2026-07-22)
• OpenAI Shares Some Alignment Problems (2276 words · 2026-07-21)
• On Kimi K3: Its Capabilities And Related Discontents (6393 words · 2026-07-20)
• Demis Hassabis on the New Coming Age (2162 words · 2026-07-19)
• AI #177 Part 2: Wish You Were Here (4429 words · 2026-07-17)
• AI #177 Part 1: Tip of the Iceberg (3793 words · 2026-07-16)
• Monthly Roundup #44: July 2026 (4038 words · 2026-07-15)
• Twitter Thoughts For You (3193 words · 2026-07-14)
• Better Call Sol The Workhorse (4027 words · 2026-07-13)
• Introduction for and Reactions to Plan A (5557 words · 2026-07-11)
• AI #176 Part 2: Plan B (5331 words · 2026-07-10)
• AI #176 Part 1: Doing It Live (4052 words · 2026-07-09)
• Childhood and Education #20: Phones and Screens (2658 words · 2026-07-08)
• No Space Like J-Space (3553 words · 2026-07-07)
• Fable #6: The Return of the King (2195 words · 2026-07-03)
• AI #175: The Fable Continues (5671 words · 2026-07-02)
• Claude Sonnet 5 Is Not Frontier But Has Its Uses (3288 words · 2026-07-01)
• The Once And Future Fable #5 (3156 words · 2026-06-30)
• WSJ Article Claiming China Has Matched Anthropic Is Obvious Nonsense (1065 words · 2026-06-29)
• GPT-5.6: The System Card (4598 words · 2026-06-28)
• White House Will Ad Hoc Decide Who Can Individually Access GPT-5.6 (2689 words · 2026-06-26)
• AI #174: You're It (6677 words · 2026-06-25)
• The Once And Future Fable #4 (2830 words · 2026-06-24)
• Monthly Roundup #43: June 2026 (4109 words · 2026-06-23)
• GLM-5.2 Is The New Best Open Model (1721 words · 2026-06-22)
• Claude Fable 5 and Mythos 5: Capabilities (4300 words · 2026-06-19)
• AI #173: AI Pauses (3293 words · 2026-06-18)
• The Once And Future Fable #3: Fix This Code (2823 words · 2026-06-17)
• Fable and Mythos: Model Welfare (2447 words · 2026-06-16)
• The Once And Future Fable #2 (2864 words · 2026-06-15)
• American Government Takes Down Claude Fable (1849 words · 2026-06-13)
• Claude Fable 5 and Mythos 5: The System Card (3054 words · 2026-06-12)
• AI #172: The First Fable (3486 words · 2026-06-11)
• Three Labs With a Plan and A Memorandum (1321 words · 2026-06-09)
• OpenAI Offers A New Policy Blueprint (1349 words · 2026-06-05)
• AI #171: False Flag (2938 words · 2026-06-04)
• Trump Signs Executive Order For AI Testing Prior To Frontier Model Releases (1459 words · 2026-06-03)
• Claude Opus 4.8: Capabilities and Reactions (2812 words · 2026-06-02)
• Opus 4.8 Part 2: Model Welfare (3070 words · 2026-06-01)
• Claude Opus 4.8: The System Card (3110 words · 2026-05-29)
• AI #170: Lack of Executive Order (4505 words · 2026-05-28)
• RTMH: Pope Leo's Magnifica Humanitas on AI (2717 words · 2026-05-26)
• Gemini 3.5 Flash Looks Good For How Fast It Is (1110 words · 2026-05-22)
• AI #169: New Knowledge (4239 words · 2026-05-21)
• Childhood And Education #19: Letting Kids Be Kids #2 (1637 words · 2026-05-19)
• Dating Roundup #12: Sex and Violence (2995 words · 2026-05-18)
• Monthly Roundup #42: May 2026 (2169 words · 2026-05-15)
• AI #168: Not Leading the Future (3499 words · 2026-05-14)
• Cyber Lack of Security and AI Governance (1837 words · 2026-05-13)
• Childhood and Education #18: Do The Math (1903 words · 2026-05-12)
• Childhood And Education #17: Is Our Children Reading (1765 words · 2026-05-11)
• Claude Code, Codex and Agentic Coding #8 (1208 words · 2026-05-08)
• AI #167: The Prior Restraint Era Begins (4190 words · 2026-05-07)
• What is Anthropic? (1173 words · 2026-05-06)
• The AI Ad-Hoc Prior Restraint Era Begins (1796 words · 2026-05-05)
• Housing Roundup #15: The War Against Renters (2158 words · 2026-05-04)
• Housing Roundup #14: You Can't Build That (2878 words · 2026-05-01)
• AI #166: Google Sells Out (5027 words · 2026-04-30)
• The Most Important Charts In The World (460 words · 2026-04-29)
• GPT-5.5: Capabilities and Reactions (2097 words · 2026-04-28)
• GPT 5.5: The System Card (1859 words · 2026-04-27)
• Monthly Roundup #41: April 2025 (4095 words · 2026-04-24)
• AI #165: In Our Image (3813 words · 2026-04-23)
• Opus 4.7 Part 3: Model Welfare (5777 words · 2026-04-22)
• Opus 4.7 Part 2: Capabilities and Reactions (3023 words · 2026-04-21)
• Opus 4.7 Part 1: The Model Card (2345 words · 2026-04-20)
• AI #164: Pre Opus (4154 words · 2026-04-17)
• On Dwarkesh Patel's Podcast With Nvidia CEO Jensen Huang (3410 words · 2026-04-16)
• Claude Code, Codex and Agentic Coding #7: Auto Mode (2553 words · 2026-04-15)
• Claude Mythos #3: Capabilities and Additions (1982 words · 2026-04-14)
• Political Violence Is Never Acceptable (2920 words · 2026-04-13)
• Claude Mythos #2: Cybersecurity and Project Glasswing (4450 words · 2026-04-10)
• Claude Mythos: The System Card (4688 words · 2026-04-09)
• AI #163: Mythos Quest (4067 words · 2026-04-08)
• OpenAI #16: A History and a Proposal (3289 words · 2026-04-07)
• Housing Roundup #13: More Dakka (1823 words · 2026-04-06)
• Anthropic Responsible Scaling Policy v3: Dive Into The Details (2838 words · 2026-04-03)
• AI #162: Visions of Mythos (5534 words · 2026-04-02)
• Anthropic Responsible Scaling Policy v3: A Matter of Trust (2847 words · 2026-04-01)
• Movie Review: The AI Doc (1631 words · 2026-03-31)
• AI #161 Part 2: Every Debate on AI (3641 words · 2026-03-30)
• Anthropic vs. DoW #6: The Court Rules (3548 words · 2026-03-27)
• AI #161 Part 1: 80,000 Interviews (3384 words · 2026-03-26)
• Claude Code, Cowork and Codex #6: Claude Code Auto Mode and Full Cowork Computer Use (2282 words · 2026-03-25)
• Book Review: Open Socrates (Part 2) (6480 words · 2026-03-24)
• Book Review: Open Socrates (Part 1) (6395 words · 2026-03-23)
• The Federal AI Policy Framework: An Improvement, But My Offer Is (Still Almost) Nothing (1310 words · 2026-03-20)
• AI #160: What Passes For a Pause (5489 words · 2026-03-19)
• Anthropic vs. DoW #5: Motions Filed (1907 words · 2026-03-18)
• Medical Roundup #7 (2124 words · 2026-03-17)
• Monthly Roundup #40: March 2026 (3883 words · 2026-03-16)
• AI #159: See You In Court (4848 words · 2026-03-12)
• GPT-5.4 Is A Substantial Upgrade (1914 words · 2026-03-11)
• Claude Code, Claude Cowork and Codex #5 (4012 words · 2026-03-09)
• Anthropic Officially, Arbitrarily and Capriciously Designated a Supply Chain Risk (3051 words · 2026-03-06)
• AI #158: The Department of War (4489 words · 2026-03-05)
• Gemini 3.1 Pro Aces Benchmarks, I Suppose (1094 words · 2026-03-04)
• A Tale of Three Contracts (2376 words · 2026-03-03)
• Secretary of War Tweets That Anthropic is Now a Supply Chain Risk (5537 words · 2026-03-02)
• Anthropic and the DoW: Anthropic Responds (3661 words · 2026-02-27)
• AI #157: Burn the Boats (4888 words · 2026-02-26)
• Anthropic and the Department of War (3183 words · 2026-02-25)
• Claude Sonnet 4.6 Gives You Flexibility (1081 words · 2026-02-24)
• Citrini's Scenario Is A Great But Deeply Flawed Thought Experiment (2914 words · 2026-02-24)
• AI #156 Part 2: Errors in Rhetoric (3309 words · 2026-02-20)
• AI #156 Part 1: They Do Mean The Effect On Jobs (4381 words · 2026-02-19)
• Monthly Roundup #39: February 2026 (4163 words · 2026-02-18)
• On Dwarkesh Patel's 2026 Podcast With Elon Musk and Other Recent Elon Musk Things (2709 words · 2026-02-17)
• On Dwarkesh Patel's 2026 Podcast With Dario Amodei (2679 words · 2026-02-16)
• ChatGPT-5.3-Codex Is Also Good At Coding (2521 words · 2026-02-13)
• AI #155: Welcome to Recursive Self-Improvement (4929 words · 2026-02-12)
• Claude Opus 4.6 Escalates Things Quickly (3993 words · Wed, 11 Fe)
• Claude Opus 4.6: System Card Part 2: Frontier Alignment (4072 words · Tue, 10 Fe)
• Claude Opus 4.6: System Card Part 1: Mundane Alignment and Model Welfare (3323 words · 2026-02-09)
• Claude Code #4: From The Before Times (2824 words · 2026-02-06)
• AI #154: Claw Your Way To The Top (4362 words · 2026-02-05)
• Kimi K2.5 (1327 words · 2026-02-04)
• Unless That Claw Is The Famous OpenClaw (2345 words · 2026-02-03)
• Welcome to Moltbook (3122 words · 2026-02-02)
• On The Adolescence of Technology (5995 words · Fri, 30 Ja)
• AI #153: Living Documents (4963 words · Thu, 29 Ja)
• Open Problems With Claude's Constitution (5477 words · Wed, 28 Ja)
• The Claude Constitution's Ethical Framework (4196 words · Tue, 27 Ja)
• Claude's Constitutional Structure (3952 words · Mon, 26 Ja)
• Dating Roundup #11: Going Too Meta (3440 words · Fri, 23 Ja)
• AI #152: Brought To You By The Torment Nexus (9813 words · Thu, 22 Ja)
• Claude Codes #3 (2925 words · Wed, 21 Ja)
• ChatGPT Self Portrait (1816 words · Tue, 20 Ja)
• Medical Roundup #6 (2934 words · Mon, 19 Ja)
• Monthly Roundup #38: January 2026 (6194 words · Fri, 16 Ja)
• AI #151: While Claude Coworks (7415 words · Thu, 15 Ja)
• When Will They Take Our Jobs? (2024 words · Wed, 14 Ja)
• Claude Coworks (3027 words · Tue, 13 Ja)
• Dating Roundup #10: Gendered Expectations (3200 words · Mon, 12 Ja)