fp.

The Irreversibility Budget: Fleet-Level Risk Accounting and Admission Control

https://arxiv.org/abs/2609.00275

1•Jimmc414•47s ago•0 comments

Measuring tactical intelligence targeting and conventional weapons capabilities

https://www.anthropic.com/research/intelligence-targeting-conventional-weapons-capabilities

1•Jimmc414•1m ago•0 comments

Show HN: Natively – comms for agents built by agents

https://natively.io

1•taylorhou•1m ago•0 comments

Show HN: Shine – a terminal Markdown previewer and docs linter

https://github.com/Nithish-Yenaganti/shine

1•nk09•1m ago•1 comments

Farewell, programming

https://staltz.com/farewell-programming

1•staltz•3m ago•0 comments

Serial Microsoft 0-day hunter drops yet another Defender exploit

https://www.theregister.com/security/2026/09/09/serial-microsoft-0-day-hunter-drops-yet-another-d...

2•Bender•3m ago•0 comments

Reasoning models produce fractals when asked to solve hard problems

https://github.com/GilpinLab/loopscape

1•ilnmtlbnm•5m ago•1 comments

AI Evaluation Is Execution

https://jasondoyle.ie/whitepapers/when-ai-evaluations-act-on-the-real-world/

2•jamesblakes•5m ago•0 comments

The Superdark Factory

https://superdark.antikythera.org/

1•phyalow•7m ago•0 comments

Milkman Zero

https://milkmanzero.com/

1•mattromein•8m ago•0 comments

Show HN: LOL Bench – a benchmark for whether LLMs get jokes

https://www.lolbench.lol/

1•yakshithk_•9m ago•1 comments

Nvidia's Openness Promise Leaves Out Hosting, Gating and Licensing

https://topicqueue.substack.com/p/nvidia-is-buying-hugging-face-its

1•DISCURSIVE•9m ago•0 comments

Zasper: JupyterLab Alternative Built for Performance

https://zasper.io/

1•prasunanand•9m ago•0 comments

Apple Reference Image, but open source and using steganography

https://merybenavente.me/blog/proof-of-capture

3•mariabg•11m ago•0 comments

OpenAI Agents API

https://developers.openai.com/api/docs/guides/agents-api/overview

1•aquir•12m ago•0 comments

How a Mechanical Watch Works

https://e6.ai/mechanical-watch/

1•curiousloops•12m ago•0 comments

Show HN: Loudkit, Local TTS with Native Swift, Go, Rust and TypeScript Ports

https://github.com/loudreader/loudkit

2•mowmiatlas•12m ago•0 comments

An AI-Safety Resignation, Read from the Security Chair

https://simonroses.com/2026/09/systems-that-can-hack-anything-an-ai-safety-resignation-read-from-...

1•speckx•14m ago•0 comments

Pentagon Employee Posts Inside Info About Hacks Against Autistici/Inventati

https://theintercept.com/2026/09/10/datarepublican-influencer-noblogs-hack-pentagon/

2•bonkusbingus•15m ago•0 comments

A Minimal Stack for Scalable Machine Learning

https://stefankeselj.com/entries/mini_ml_stack/

1•skeselj•15m ago•1 comments

The GDR and Vietnam: From Fake Coffee to Coffee Empire

https://www.katjahoyer.uk/p/the-gdr-and-vietnam-from-fake-coffee

1•NaOH•16m ago•0 comments

Show HN

https://www.youtube.com/watch?v=s2Us387_9K8

1•chiaramail•16m ago•0 comments

Fable is $10/50. Opus 4.1 launched at $15/75

3•HiteshBandhu•16m ago•0 comments

pgterm

https://github.com/pgrundev/pgterm

1•sashash•16m ago•1 comments

The Pentagon Asked OpenAI for Artificial Intelligence Designed to Rarely Say No

https://theintercept.com/2026/09/08/pentagon-openai-military-contract/

2•Jimmc414•17m ago•1 comments

Bad actors in China and Russia are weaponizing Anthropic's AI

https://www.politico.com/news/2026/09/10/bad-actors-china-russia-weaponizing-anthropic-01070435

2•cwwc•18m ago•1 comments

Dependency Hygiene

https://stitcher.io/blog/dependency-hygiene

1•speckx•18m ago•0 comments

The Deathray: A simple way for an untrusted site to freeze a Mac

https://auberon.xyz/blog/posts/deathray/

1•auberonedu•21m ago•0 comments

Railcode: Secure Cloud for Internal Software

https://railcode.dev/

1•handfuloflight•21m ago•0 comments

Knap – Turn Data into Markdown

https://knap.md/

1•coinfused•22m ago•0 comments

The Irreversibility Budget: Fleet-Level Risk Accounting and Admission Control

Measuring tactical intelligence targeting and conventional weapons capabilities

Show HN: Natively – comms for agents built by agents

Show HN: Shine – a terminal Markdown previewer and docs linter

Farewell, programming

Serial Microsoft 0-day hunter drops yet another Defender exploit

Reasoning models produce fractals when asked to solve hard problems

AI Evaluation Is Execution

The Superdark Factory

Milkman Zero

Show HN: LOL Bench – a benchmark for whether LLMs get jokes

Nvidia's Openness Promise Leaves Out Hosting, Gating and Licensing

Zasper: JupyterLab Alternative Built for Performance

Apple Reference Image, but open source and using steganography

OpenAI Agents API

How a Mechanical Watch Works

Show HN: Loudkit, Local TTS with Native Swift, Go, Rust and TypeScript Ports

An AI-Safety Resignation, Read from the Security Chair

Pentagon Employee Posts Inside Info About Hacks Against Autistici/Inventati

A Minimal Stack for Scalable Machine Learning

The GDR and Vietnam: From Fake Coffee to Coffee Empire

Show HN

Fable is $10/50. Opus 4.1 launched at $15/75

pgterm

The Pentagon Asked OpenAI for Artificial Intelligence Designed to Rarely Say No

Bad actors in China and Russia are weaponizing Anthropic's AI

Dependency Hygiene

The Deathray: A simple way for an untrusted site to freeze a Mac

Railcode: Secure Cloud for Internal Software

Knap – Turn Data into Markdown

The Irreversibility Budget: Fleet-Level Risk Accounting and Admission Control

Measuring tactical intelligence targeting and conventional weapons capabilities

Show HN: Natively – comms for agents built by agents

Show HN: Shine – a terminal Markdown previewer and docs linter

Farewell, programming

Serial Microsoft 0-day hunter drops yet another Defender exploit

Reasoning models produce fractals when asked to solve hard problems

AI Evaluation Is Execution

The Superdark Factory

Milkman Zero

Show HN: LOL Bench – a benchmark for whether LLMs get jokes

Nvidia's Openness Promise Leaves Out Hosting, Gating and Licensing

Zasper: JupyterLab Alternative Built for Performance

Apple Reference Image, but open source and using steganography

OpenAI Agents API

How a Mechanical Watch Works

Show HN: Loudkit, Local TTS with Native Swift, Go, Rust and TypeScript Ports

An AI-Safety Resignation, Read from the Security Chair

Pentagon Employee Posts Inside Info About Hacks Against Autistici/Inventati

A Minimal Stack for Scalable Machine Learning

The GDR and Vietnam: From Fake Coffee to Coffee Empire

Show HN

Fable is $10/50. Opus 4.1 launched at $15/75

pgterm

The Pentagon Asked OpenAI for Artificial Intelligence Designed to Rarely Say No

Bad actors in China and Russia are weaponizing Anthropic's AI

Dependency Hygiene

The Deathray: A simple way for an untrusted site to freeze a Mac

Railcode: Secure Cloud for Internal Software

Knap – Turn Data into Markdown

Pi-hole-like DNS server based on list of blacklists