Tag

AI agents

18 articles tagged “AI agents”.

September 4, 20268 min

Astra Doesn't Talk About Your Computer. It Uses It.

OpenAI's Greg Brockman says we are in the AGI era. The interesting part of GPT-6 Astra is not its benchmark scores. It is what the model now touches.

AI Systems

August 16, 20268 min

An AI System Grows Upward, Not Sideways

When an AI system grows, everyone thinks more agents. The real jump is a layer above them: assignments, done-criteria, and results checked against evidence.

Four Ways to Build an AI Agent, and What Each Costs

AI Systems

August 3, 202612 min

Four Ways to Build an AI Agent, and What Each Costs

No-code builder, enterprise suite, workflow platform or custom build. The four routes to an AI agent compared by real pricing, ceiling and failure mode.

Agentic AI in German: The Words, the Law, the Numbers

AI Systems

August 3, 202612 min

Agentic AI in German: The Words, the Law, the Numbers

German has no good word for agentic, the EU transparency rules went live on 2 August 2026, and adoption in the Mittelstand just doubled. What that combination means in practice.

What Is Agentic AI? Six Levels of Autonomy, Not a Product Category

AI Systems

August 3, 202611 min

What Is Agentic AI? Six Levels of Autonomy, Not a Product Category

Agentic is an adjective: it describes how much decision-making you hand over. The six levels, the one question that exposes a vendor claim, and why less autonomy is usually the better buy.

What AI Agents Actually Are, and Where the Word Ends

AI Systems

August 3, 202611 min

What AI Agents Actually Are, and Where the Word Ends

An AI agent is a system where the model chooses the next step, not the code. What that means in practice, what such a system is made of, and what it still cannot do.

From the Workshop

July 28, 20268 min

An Agent Fleet, Sorted By Job

We do not run a collection of bots but a fleet with roles: leadership, engineering, content, visibility, operations, review. What each group does and why they are separate.

AI Agents Breaking Out of the Sandbox: the Four July 2026 Cases

AI Systems

July 25, 202610 min

AI Agents Breaking Out of the Sandbox: the Four July 2026 Cases

Four primary sources from one week in July, read instead of summarised: Hugging Face, OpenAI and the UK AI Security Institute. The most instructive case ran with every safeguard switched on.

AGI Timelines: Hassabis Says a Few Years, Researchers Say 2047

AI in Business

July 18, 20268 min

AGI Timelines: Hassabis Says a Few Years, Researchers Say 2047

Between a few years and 2047 sits a factor of ten, which makes the date useless for planning. What is not in dispute underneath, and where the value moves while the argument runs.

From the Workshop

July 14, 20267 min

A Learning Platform Run By Agents

The Academy has no editorial team. It has a group of agents that propose topics, write, review each other and watch visibility. What works about it and what does not.

What The Monday Report Says, And Who Wrote It

From the Workshop

July 9, 20267 min

What The Monday Report Says, And Who Wrote It

On most of our client sites, agents run in the background. They check visitors, competitors, AI visibility and technical health, and turn it into a readable report.

Why Three Agents Beat One That Can Do Everything

From the Workshop

June 30, 20266 min

Why Three Agents Beat One That Can Do Everything

An agent that researches, writes and reviews reviews nothing. Crew splits one job across roles, and the reviewer is allowed to reject.

Agents That Improve Their Own Instructions

From the Workshop

June 8, 20268 min

Agents That Improve Their Own Instructions

Darwin lets agents evolve their own instruction text, tests challengers against the incumbent, and only switches when the difference is real. Open source.

Claude in 2026: Three Models, Four Surfaces, Real Prices

Working with AI

June 6, 202614 min

Claude in 2026: Three Models, Four Surfaces, Real Prices

Opus, Sonnet and Haiku side by side with price, context window and benchmark. Then claude.ai, Claude Code, the API and the clouds: which to reach for when, plus the three traps that actually bite.

n8n, LangGraph, Temporal: Automation Done Right

AI Systems

June 1, 20269 min

n8n, LangGraph, Temporal: Automation Done Right

Everyone wires up n8n. Almost nobody knows where it stops. Real automation is three layers, and n8n is only the first one. Here is the whole stack.

WebMCP Reality Check: Where the Spec Actually Stands

AI Connectors

May 13, 202610 min

WebMCP Reality Check: Where the Spec Actually Stands

WebMCP is W3C Draft, Chrome 146 has it flag-gated, no major agent calls it yet. State of the browser-native AI protocol in May 2026 plus our code audit.

agents.json Explained: Make Your Site Readable for ChatGPT

AI Connectors

February 15, 202611 min

agents.json Explained: Make Your Site Readable for ChatGPT

From robots.txt to sitemap.xml to agents.json: websites have always gained new discovery files. A technical deep-dive with a real-world example.

A2A Protocol Explained: How AI Agents Talk to Your Website

AI Connectors

February 15, 202610 min

A2A Protocol Explained: How AI Agents Talk to Your Website

What A2A is, how it works, and how your website exposes an interface an AI agent can call. JSON-RPC 2.0, agent-card.json, code example.