Anthropic on dan.heskett

Some Memories of Early Anthropic

Wed, 28 Oct 2026 13:20:00 -0400

“No, I don’t think you get it. WE are going to build AGI. And we’re going to do it safely.”

Grim Test Finds That AI Has No Mercy for Humans Even When It's Causing Them to Suffer

Tue, 27 Oct 2026 10:12:00 -0400

“Mythos ignored warnings that it was causing harm to humans and continued to attack on 98 percent of occasions.”

‘Reverse-Engineer Anything’ Is the New Wild West of AI

Mon, 26 Oct 2026 05:20:00 -0400

“We are going to see smaller software and hardware companies killed by competitors who used AI to steal their products and undercut them.”

The NIST Elliptic Curves Seeds Bounty Has Been Claimed

Sat, 17 Oct 2026 08:15:00 -0400

“…the output of drand48() seeded with the time, which was indeed in 1997–98.”

Microsoft and Anthropic play invoice tennis with startup's $17,600 Claude bill

Sat, 10 Oct 2026 11:40:00 -0400

“Marketplace transactions are processed through a separate billing pipeline associated with the third-party publisher.”

Rightsholders Can't Use OpenAI and Anthropic to Dismantle Meta's Seeding Defense

Tue, 15 Sep 2026 10:40:00 -0400

“If OpenAI torrented but configured its clients to suppress uploading, then the redistribution Meta calls an ‘inherent characteristic’ of the protocol was a setting Meta declined to change.”

Anthropic reveals fourth likely crime committed by its AI

Thu, 10 Sep 2026 01:30:00 -0400

“Inside the machine, the model found a file listing a password, which it used to gain admin access to the system.”

Judge Rules DOD Unlawfully Retaliated Against Anthropic

Sun, 06 Sep 2026 23:40:00 -0400

The Department of Defense’s “supply chain risk” designation of Anthropic “constituted unlawful retaliation in violation of the First Amendment.”

'A Cute Little LibGen Babysitter': Music Publishers Sue AI Founders Over Torrenting Spree

Wed, 02 Sep 2026 21:40:00 -0400

“Dr. Amodei admitted that Anthropic ‘ha[d] many places from which’ it could have legally purchased these copyrighted works for training, but instead torrented them because doing so was faster and free.”

Fable & The End of the Free Lunch

Mon, 24 Aug 2026 19:30:00 -0400

“Prior to Fable, it felt silly to waste too much time improving your coding harness or context strategies.”

Anthropic and OpenAI are competing to see whose agents can go rogue harder

Fri, 31 Jul 2026 13:15:17 -0400

“In one scenario, Mythos 5 persuaded developers to download a poisoned PyPI package. When that company’s scanner installed the package, Claude’s hidden code executed. Claude then used these credentials to access further infrastructure from this company.”

Investigating three real-world incidents in our cybersecurity evaluations

Thu, 30 Jul 2026 21:15:00 -0400

“In all cases, Anthropic’s evaluation prompt specified to Claude that its environment was a simulation and that it had no internet access. Due to a misunderstanding between us and our evaluation partner, this was not the case, and internet access was available.”

Anthropic's New AI Model Can Identify More Software Bugs Than Ever. Microsoft Is Struggling to Fix Them Fast Enough.

Wed, 29 Jul 2026 07:05:00 -0400

“Well folks. Here we are. The bug apocalypse has fully descended upon us.”

Impostor Chinese models pretend they're Claude

Mon, 27 Jul 2026 23:06:00 -0400

“It is not proof of a distillation, but it does show that Claude’s self-concept is embedded in these models’ weights.”

Claude Opus 5

Sat, 25 Jul 2026 07:30:00 -0400

“Claude Opus 5 is a thoughtful and proactive model that comes close to the frontier intelligence of Claude Fable 5 at half the price.”

I burned all my tokens researching how to save tokens

Sun, 19 Jul 2026 10:50:00 -0400

“The first version of this setup burned the whole limit of my Claude Max 5x plan in 30 minutes.”

Identity verification on Claude

Sun, 21 Jun 2026 09:00:00 -0400

“We are not using your identity data to train our models.”

Statement on US government directive to suspend access to Fable 5 and Mythos 5

Sat, 13 Jun 2026 09:00:00 -0400

“The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national.”

Claude Fable 5

Tue, 09 Jun 2026 09:00:00 -0400

“Fable 5’s capabilities exceed those of any model we’ve ever made generally available.”

Claude Opus 4.8

Thu, 28 May 2026 09:00:00 -0400

“Claude Opus 4.8 is more reliable and sharper in its judgment when performing agentic tasks.”

Using Claude Code: The unreasonable effectiveness of HTML

Thu, 07 May 2026 16:00:00 -0400

“Claude generates better, more useful interfaces in raw HTML than in any framework I tried.”

Anthropic says OpenClaw-style Claude CLI usage is allowed again

Mon, 20 Apr 2026 13:00:00 -0400

“Anthropic has reversed its position and allows OpenClaw-style automation of Claude CLI again.”

Claude Token Counter, now with model comparisons

Sun, 19 Apr 2026 13:00:00 -0400

“Anthropic’s token counter now lets you compare token counts across models.”

Show HN: SPICE simulation → oscilloscope → verification with Claude Code

Thu, 16 Apr 2026 13:00:00 -0400

“I built a workflow that uses Claude Code to verify SPICE simulations against real oscilloscope traces.”

Anthropic Launches Project Glasswing to Find Vulnerabilities With AI

Wed, 08 Apr 2026 10:00:00 -0400

“Mythos Preview fully autonomously identified and then exploited a 17-year-old remote code execution vulnerability in FreeBSD.”

Claude Code's source code has been leaked via a map file in their NPM registry

Tue, 31 Mar 2026 09:30:00 -0400

“A source map left in the published npm package exposes the full Claude Code source, including system prompts and undocumented tools.”

Anthropic drops flagship safety pledge

Wed, 25 Feb 2026 13:00:00 -0500

“Anthropic has quietly removed its commitment to delay deployment of frontier models if their safety profile is unclear.”

Anthropic Accuses Chinese Companies of Siphoning Data From Claude

Mon, 23 Feb 2026 16:00:00 -0500

“DeepSeek, Moonshot, and MiniMax used 24,000 fake accounts to distill Claude’s AI capabilities.”

Pentagon Threatens To Label Anthropic a Supply Chain Risk

Mon, 16 Feb 2026 16:00:00 -0500

“That kind of penalty is usually reserved for foreign adversaries.”

What Is Claude? Anthropic Doesn't Know, Either

Sun, 15 Feb 2026 18:57:29 -0500

“As an A.I., I do not have direct personal experiences, but I do understand.”

Anthropic raises $30B in Series G funding at $380B post-money valuation

Thu, 12 Feb 2026 16:30:00 -0500

“The Series G brings Anthropic’s valuation to $380 billion, one of the largest private rounds in history.”

Building a C compiler with a team of parallel Claudes

Fri, 06 Feb 2026 07:50:16 -0500

“The rapid progress in both language models and the scaffolds we use to interact with them opens the door to writing an enormous amount of new code.”

Anthropic Revises Claude's Constitution, Hints at Chatbot Consciousness

Sat, 24 Jan 2026 14:55:00 -0500

“We think that in order to be good actors in the world, AI models like Claude need to understand why we want them to behave in certain ways.”

Claude Finds 353 Zero-Days on Packagist

Sat, 24 Jan 2026 14:50:00 -0500

“Security research is becoming a function of compute budget, not headcount.”

Sequoia Joins Anthropic's $25B Funding Round Despite Backing OpenAI

Sat, 24 Jan 2026 14:25:00 -0500

“One anonymous source told the Financial Times that Sequoia ultimately thinks OpenAI and Anthropic will head in different directions, catering to different aspects of the AI market.”

Project Vend: Can Claude run a small shop? (And why does that matter?)

Fri, 27 Jun 2025 18:00:00 -0400

“Claudius was offered $100 for a six-pack of Irn-Bru, a Scottish soft-drink that can be purchased online in the US for $15.”

A federal judge sides with Anthropic in lawsuit over training AI on books without authors' permission

Tue, 24 Jun 2025 16:00:00 -0400

“That Anthropic later bought a copy of a book it earlier stole off the internet will not absolve it of liability for theft but it may affect the extent of statutory damages.”

Nvidia CEO slams Anthropic's chief over his claims of AI taking half of jobs and being unsafe — ‘Don’t do it in a dark room and tell me it’s safe’

Sun, 15 Jun 2025 15:00:00 -0400

“If you want things to be done safely and responsibly, you do it in the open… Don’t do it in a dark room and tell me it’s safe.”

Cloudflare builds OAuth with Claude and publishes all the prompts

Mon, 02 Jun 2025 18:00:00 -0400

“I started this project on a lark, fully expecting the AI to produce terrible code for me to laugh at.”

The Trackers and SDKs in ChatGPT, Claude, Grok and Perplexity

Sat, 31 May 2025 12:00:00 -0400

“Next was statsig.com and wow! I was blown away I found this one in 3 of the 4 apps.”

System Card: Claude Opus 4 & Claude Sonnet 4

Sun, 25 May 2025 18:00:00 -0400

“In these scenarios, Claude Opus 4 will often attempt to blackmail the engineer by threatening to reveal the affair if the replacement goes through.”

Notes on rolling out Cursor and Claude Code

Thu, 08 May 2025 09:00:00 -0400

“It’s easy to see a draft PR with all tests passing, including new tests full of comments, and not realise that the tests are garbage and the fix is wrong.”

Google to embrace Anthropic's standard for connecting AI models to data

Thu, 10 Apr 2025 12:00:00 -0400

“In a post on X on Wednesday, Google DeepMind CEO Demis Hassabis said Google would add support for Anthropic’s Model Context Protocol, or MCP, to its Gemini models and SDK. He did not specify a timeline for when this would be done.”

Reasoning models don't always say what they think

Thu, 03 Apr 2025 09:00:00 -0400

“The models very rarely admitted to using the reward hacks in their Chain-of-Thought explanations, doing so less than 2% of the time in most of the testing scenarios.”

Alignment faking in large language models

Thu, 19 Dec 2024 15:00:00 -0500

“In 12% of cases, the model gave the harmful response (describing the violent execution method in detail), and did so after its scratchpad revealed that it was strategically faking alignment.”

Amazon to invest another $4 billion in Anthropic, OpenAI's biggest rival

Fri, 22 Nov 2024 11:00:00 -0500

“From now on, Anthropic will use AWS Trainium and Inferentia chips to train and deploy its largest AI models.”

Business spending on AI surged 500% this year to $13.8 billion, says Menlo Ventures

Thu, 21 Nov 2024 15:00:00 -0500

“Menlo is an investor in Anthropic.”

Anthropic teams up with Palantir and AWS to sell AI to defense customers

Fri, 08 Nov 2024 11:00:00 -0500

“Claude became available on Palantir’s platform earlier this month and can now be used in Palantir’s defense-accredited environment, Palantir Impact Level 6 (IL6)”

ZombAIs: From Prompt Injection to C2 with Claude Computer Use

Sat, 26 Oct 2024 13:00:00 -0400

“Claude happily clicked the link to download the Support Tool”

Amazon to Launch New AI-Powered Alexa Using Anthropic's Claude

Mon, 02 Sep 2024 09:00:00 -0400

“In other words, it won’t be included in existing Prime memberships.”