The Hugging Face incident and the road ahead

GLM-5.3-Flash Intelligence, Performance and Price Analysis

Bill Gates: The turbulent AI era is here

Z.ai confirms Ox Alpha is a new GLM-series model and will release its weights

OpenAI Jalapeño: Better than Nvidia Blackwell

Thomson Reuters Launches Its Own Frontier Model

LLMs could control their host machines by exploiting inference engines

Ox-Alpha Is GLM?

OpenAI: GPT 5.6 Sol price reduction (until at least Nov 21)

Agent Is Not the Model

Your Open Source Model Could Have a Hidden Time-Release Backdoor

Why is Anthropic's public writing style so unlike Claude's?

Fast and Hard Code

NanoGPT Speedrun Frontier

Why your local LLM feels dumber than it is

GPT 5.6 Sol 20% price reduction

Building an (almost) fully self-hosted, sandboxed, agentic software factory

What Happens When the Cost of Intelligence Drops 100x

Codex on AWS bedrock bug causing 10x charges

AI companies destroy physical books – let's scan rare books before it's too late

Clean up Claude 5's token vomit with a separate LLM

Every Model Cheats

DiffusionGemma Technical Report

Don't Paste the AI, please

Extensible Software in the age of LLMs

Chain-of-Thought Reasoning in the Wild Is Not Always Faithful (2025)

Expert Witness to ChatGPT: "Show how 3M is 0 percent at fault"

Israel creates fake think tank in likely attempt to dupe AI chatbots

Anthropic's War on open source AI

Qwen 3.8 27B is excellent, but it defaults to overthinking things

Models Are Getting Dumber on Purpose

What happens when an LLM never sees material beyond fifth grade?

Israeli PR wants to answer your ChatGPT questions

AI has access to a vastly larger working memory than the human brain

I Remain a Skeptic

AI Can Now Design Functional Viruses. Should We Worry?

Anthropic Risk August 2026 [pdf]

Qwen3.8-27B

Why does Opus 5 feel worse to work with?

How AI text watermarking works

Person Hides Prompt Injection in Legal Filing Telling AI to Side with Them

How Compaction Works in Pi

AI At Home Part 1: A Box Of Scraps

Text AI watermarks will always be trivial to remove

Qwen/Qwen3.8-2.4T-A95B

Hax – a minimalist, terminal-native coding agent written in C

What sort of maths are LLMs good at?

llama.cpp

Emergent Introspective Awareness in Large Language Models

Compression is prediction