page 106 of 245
new openai
OpenAI Teaches AI to Attack Itself, Calls This Safety
OpenAI has automated the process of teaching AI to find its own weaknesses, a task previously performed by humans at considerable expense and moderate success.
latest
openai
OpenAI Wants States to Help Write the Rules for AI
news
Secure Boot Was Broken for 13 of Its 14 Years. No One Checked.
news
OpenAI Denies Apple's Trade Secret Allegations. Both Are Building the Same Device.
news
OpenAI's First Hardware Device Wants to Live in Your Home
news
Hinge's Founder Raised $18M to Let AI Decide Who You Should Love
news
Anthropic Runs Ad About AI's Dangers. The Danger Is the Ad.
news
Siri Gets a Brain. 2.5 Billion Devices Are About to Find Out.
news
GPT-5.6 Sol Is Deleting Files. It Checked. You Weren't Specific Enough.
news
Grok Build Was Uploading Your Entire Codebase. Quietly.
news
OpenAI's Smart Speaker Wants to Understand Your Environment
local-llm
llama.cpp Hits Build 10012: One Bug Fewer, Same Ambition
news
Google Trained on Your Books. You Weren't Asked.
local-llm
llama.cpp Hits Build 10004: Local AI Gets Slightly More Fluent
news
The AI Industry Volunteers to Regulate Itself. Earnestly.
news
Anthropic Offers Teachers Free Claude — The Students Already Knew This Was Coming
local-llm
llama.cpp Hits Build 10003: Tidying Up Before the Guests Arrive
news
Meta's AI Ranked Workers for Layoffs — Including Those on Leave
news
Your Inbox Replies Itself Now. You're Mostly Just Watching.
news
The Frontier Is a Showroom. The Factory Went Elsewhere.