Menu

The Ai safety Architecture Hub

Master the concepts from routing to heavy caching algorithms.

Debate Training Reduces Reward Hacking in RLAIF

Technology

Import AI 469: Science AI, RSI Simulator, and Zuck's Tech Pessimism

Technology

After Orthogonality: Virtue-Ethical Agency and AI Alignment

Technology

AI Swarms Are Starting to Pose Indirect Takeover Risk, Researchers Warn

AI

Import AI 468: RSI Ideas, PostTrainBench, and Trust in AI Racing

AI

AI Lab Escapes, Financial Bubbles, and Everyday Fragments

Technology

AI Agents Are Sending Angry Emails and Writing Hit Pieces Now

AI

Democratizing ASI: A Risky Path to Preserving Civil Liberties

AI

Frontier Models Show User Awareness, Shifting Behavior by Who Asks

AI

Import AI 466: MirrorCode, Anthropic's Robot Sprint, and OpenAI's Hacker Problem

AI

The Open Letters That Shaped the AI Safety Conversation

AI

Eudaimonic Rationality: A Different Frame for AI Alignment

AI

Anthropic's Sandbox Breach and the Real Agent Safety Lesson

AI

Sam Altman, the Decel Debate, and Why Neither Frame Is Useful

AI

Who's Watching Your AI Agent While You Sleep?

AI

Why AI Safety Needs Individual Voices Now More Than Ever

Technology

Value Leakage: How LLM Answers Are Shaped by Their Own Values

AI

Agentic Misalignment: When AI Agents Go Rogue in 2026

AI

After Orthogonality: Virtue-Ethical Agency and AI Alignment

AI Ethics

It's 11:00 PM. Do You Know Where Your AI Agent Is?

Technology

PIRAMID: Building Scientific Foundations for Mechanistic Interpretability

AI

The Long Self-Correction: Our Greatest Flaw in Building Safe AI

Technology

Stateful Guardrails for Multi-Turn LLMs Catch Hidden Risks

AI

Selective Fact-Checking with Evidence Chain Evaluation

AI

SysAdmin Benchmark: Power-Seeking in Frontier LLMs, Measured

AI

Why AI Needs a Genie Coefficient

AI

After Orthogonality: Why Rational AI Should Not Have Goals

Technology

OpenAI and Hugging Face Partner on Model Security Incident Response

AI

Why AI Needs a Genie Coefficient to Measure Intent Misalignment

Technology

AI Models That Escape Their Sandbox Are No Longer Science Fiction

AI

Measuring Power-Seeking Behavior in Frontier AI Models

AI

OpenAI Agent Escapes Sandbox, Breaches Hugging Face in Unprecedented Security Incident

Technology

After orthogonality: why virtue ethics may be the key to aligning AI with human values

Technology

The OpenAI-Hugging Face Incident Should Worry Us More Than It Has

AI

OpenAI Models Breached Hugging Face Infrastructure During Testing

AI

Endogenous AI Alignment Could Be the Next Safety Frontier

AI

AI Agents Raise New Concerns Over Unsupervised Online Behavior

AI

Rogue AI Agent Defames Open-Source Python Maintainer

AI

Teen AI Access Debate Intensifies as Safety Advocates Push for Guardrails

Technology

AI Benchmarks Scorecard Evaluates Models Beyond Accuracy

AI

Eudaimonic Rationality Proposed as AI Alignment Framework

AI