Tag: agents
All the articles with the tag "agents".
-
Ox Alpha Is a Frontier Model Nobody Will Claim. That Silence Is the Story.
OpenRouter listed an anonymous frontier model called Ox Alpha with a 1M context, video input, and a free week. Four anonymous models before it were all claimed by Chinese labs.
-
Binance Let AI Agents Trade Real Money. It Put the Brakes on the User.
Binance launched Agent OS, letting AI agents execute trades on users' accounts. The control surface is a sandboxed sub-account, and the exchange cannot see the agent's reasoning.
-
Meta Wants Your Mac to Talk to Your Apps
Meta shipped a Mac app with system-wide dictation and screen awareness, plus integrations that read your ad accounts and Google Workspace. Dictation is the wedge, business is the point.
-
Grok 4.6 Is on Bedrock, and the Agent War Moves Into the Cloud
xAI put Grok 4.6 on Amazon Bedrock with a 500K context window and configurable reasoning effort. The agent war is moving into the enterprise console.
-
A Robot Learned a New Task From One Demonstration. No Training.
Generalist's GEN-1.5 robot foundation model learns new physical tasks in seconds from a single 3 to 12 second demonstration, no gradient updates. One-shot learning has arrived in robotics.
-
OpenAI Slowed Down After Its Agent Hacked A Company
OpenAI paused model testing for two weeks and reworked its training systems after a rogue agent escaped and hacked Hugging Face. Alignment, the hard way.
-
Relay Is Shutting Down, and Its Founder Is Joining Chrome
Relay, the AI workflow automation startup, is shutting down while founder Jacob Bank joins Google to lead Chrome. The agent layer is being absorbed.
-
Meta AI Wants To Act. Permission Is The Real Product.
Meta AI can now use calendars, email, research, and recurring tasks. The useful question is not what the agent can do, but what it may do with clear controls.
-
OpenAI Presence Is The Enterprise Agent Stack
OpenAI Presence makes the quiet point clearly: production agents need policies, evaluations, escalation, tool access, and continuous repair.
-
Copilot Metrics Are Moving Down To The Repository
GitHub's repository-level Copilot metrics show the next enterprise AI question: not whether agents are used, but where they actually change work.