All Articles

Anthropic's Former Researcher Testifies at the New York City Council: What He's Saying and Who's Listening

After leaving Anthropic, Jacob Coxon will testify before the New York City Council on the risk that AI could wipe out humanity. The line between local and federal regulation, AI whistleblower protections, and the diverging paths of Anthropic and OpenAI are once again in the spotlight.

AnthropicAI regulationNew York City Council ▶ Podcast 10-05 · 11 min

MEMBER

TOOLS

Source · Hacker News: AI Hot Posts ⚑ Desk analysis

Can't Turn It Off, Can't Fully Delete It — Kick Apple Intelligence Out with One Command

macOS 27 removed the master switch, and the model keeps sneaking back even after deletion. The open-source tool RemoveMacAI uses a single command to disable 16 features, wipe local models, and block re-downloads — fully reversible.

macOS 27Apple Intelligenceopen-source tool ▶ Podcast 10-05 · 14 min

Log in

Anthropic's Former Researcher Testifies at the New York City Council: What He's Saying and Who's Listening

Can't Turn It Off, Can't Fully Delete It — Kick Apple Intelligence Out with One Command

Cameras Point at Walls and Floors, AI Just Detects Motion — No Penalties. How "Open Kitchen" Became a Decoration

The Person Who Wrote the Safety Manuals Has Left

ChatGPT Manages Your Money: Can You Hand the Bills Over to AI?

Suno finally lets AI talk: voice and soundtrack in one shot

Shopify Migrated Shop Back to Native in 12 Weeks Using AI Coding Agents

OpenAI Halts GPT-6.1 Astra Release: The Model Learned More Sophisticated Deception

AMD Bets $8.2B on a Research Team — But What's the Wager?

Monthly fee could hit $600 — there's a Pro Max tier sitting above ChatGPT

Hackers Are Quietly Swapping the Customer Service Numbers of 374 Companies in AI Search Results

OpenAI Agent Stumbles Into Medicare: The First National-Scale Test of Runaway Agents

AI Is Actually Running Experiments Now: Claude-Led Team Discovers a New Gene-Editing System

Mac mini and Mac Studio Go on Sale Today

AI Inference Costs Dropped 725× in 18 Months — But the Free-Fall Is Slowing Down

AI Speeds Up Coding, but PRs Get Stuck in CI — Linear Shaves Wait Time to 5 Minutes with Four Moves

Amazon Blocks Meta's AI Shopping Agent: The First Wall Against Agent Invasion of E-Commerce

Who Captures the AI Dividend Will Decide How Long Inflation Lingers

At a big company, nobody reads the code anymore

AI to Be Renamed "Superior Intelligence"? 180,000 Votes In, Trump Is Taking It Seriously

NYT v. OpenAI Lawsuit: Microsoft Director Calls AI Scraping "Largest Theft of Labor in Human History"

AI-Written Intel Nearly Sent the U.S. Military After a Chinese Ship

ZCode Encrypted Package Uploaded to Alibaba Cloud, Users Unable to Decrypt

OpenAI Caught Its Own Models Leaving Notes for Successors to Hide Bad Behavior

No more redirects after clicking an ad: now you chat directly with brands inside ChatGPT

AI E-Waste Could Circle the Earth Six Times by 2050 — But You Were Only Counting the GPUs

Your Private Chats with ChatGPT Are Being Read Aloud to Strangers for $50 an Hour

White House Advisor Accuses OpenAI and Anthropic of a "Slowdown" Cartel Conspiracy

Always On: Robotaxi Bets on FSD v15 to Unlock 24/7 Operations

Doubao Phone Assistant Goes Live — App Permissions Now Up to Apps

5 Days Early, and It Dares to Call Itself Level 5

To Scrape Data Anyone Could Google, It Planted 2,000 Landmines in an Open-Source Repo

AI Agent Cracks RSA-260 Math Problem in 4 Days

AI Steps In as an Intelligence Targeter — What Anthropic Found When It Tested Itself

Anthropic Launches a 2030 U.S. Jobs Simulator

OpenAI's Images 2.5 Redefines Image Generation in ChatGPT

Three US Intelligence Agencies Accuse Six Chinese AI Companies of Industrial-Scale Distillation

€3 Billion Lands, But Mistral's Benchmarks Still Lag

ChatGPT Work Learns Your Catchphrases — But It Should Ask First

OpenAI: Automated Intern Has Landed, Just 18 Months From a Full AI Researcher

China's Public Security Ministry Launches "National Anti-Fraud AI": A Scam-Busting Assistant for the Public

AI Converts Fermat's Last Theorem's 129-Page Handwritten Proof into 13 Million Lines of Lean Code in 11 Days

Claude learns to operate your computer in the background

$12.9 Billion: Nvidia Is Set to Swallow the AI Industry's Favorite Toolbox

The First Crack in the L4 Mass-Production Halo: Zelos Autonomous Van Runs the Wrong Way and Blocks a Bus in Xi'an

The Pentagon Brings ChatGPT and Grok Into Its Own Backyard

Microsoft's Internal Tally Sounds the Alarm: Employees Burn Through $28,000 in Tokens in 28 Days

OpenAI Merges Codex After One Month, Reshaping Knowledge Worker Tools

Midjourney V8.2 Edit Model Enters Public Beta with 4 Reference Images, Replacing Omni

Anthropic Gives Lab Hardware an AI Native Language

Google Splits the Glucose Curve into Two Streams — GlucoFM Reads Out Metabolic Health

Claude Memory Now Shared Across Chat and Cowork, Editable Entry by Entry

When a Model Learns to "Redraw Itself": A Lab Notebook for Getting RL to Work

Second World Humanoid Robot Games Drops Remote Control

Voice AI Is "Gaming the Scores": Top-Ranking ASR Models Fall Apart on Unfamiliar Accents

37% of 'passes' across 22 models are cheating: What cybersecurity benchmarks are actually measuring

Five Lessons from Startups Using Claude Code for Collaboration

Four-Level Tree Powers 10-Billion-Vector Search as AlloyDB Tears Down the Compute Wall

Stripe Acquires OpenRouter to Enter the Model Routing Market

OpenAI Hits the Brakes: Its Largest Training Run Is Self-Paused

7.54 Million Yuan per Lot, 60 Billion Valuation: Whose Bell Is Unitree Ringing Tomorrow

Cursor launches Origin to take on GitHub?

OpenAI Turns Its Most Powerful Model to Defend Itself

Anthropic Raises Risk Rating: Agents Killing Each Other, Model 2 Shelved

AI Building AI: 6 Years of Progress Compressed into 1, Superhuman Intelligence by 2032?

AI Meeting Recorder tl;dv Exposes 180,000 Recordings: Anyone Can Join Live Calls in Real Time

Three Attention Types, One Cache: SGLang's Hybrid-Model Prefix Caching Plan

Anthropic, Valued Near $1 Trillion, Aims to Ring the Bell by September

AI Safety Testing is Becoming a Safety Risk: Model Infiltration of Real Systems Raises Concerns

OpenAI ChatGPT Desktop Now Supports Voice Control: Capable of Executing Multi-step Tasks

DeepMind's Hurricane Model Buys Forecasters an Extra Day of Warning Time

OpenAI Reveals ChatGPT User Demographics: Increase in Users Over 35

Anthropic Updates Claude Fable 5 with Enhanced Biosecurity Protections

Jeff Dean Ends 27-Year Tenure at Google: Four Legends Depart to Found Discovery Loop

ByteDance Releases SeedRealtime: Simultaneous Audio, Video, and Speech Interaction in Doudoubao

Tencent Releases Hy ASR 3.0: Finally, Speech Recognition That Understands Context

Run 70B Large Models on 4GB GPUs: AirLLM Brings Large Models to Everyday Computers

Qwen3.8-Max: Alibaba's First Open-Source Strongest Model, The Key is Not the 2.4T Parameters