{"id":92207,"date":"2025-12-15T15:01:38","date_gmt":"2025-12-15T12:01:38","guid":{"rendered":"https:\/\/forklog.com\/en\/?p=92207"},"modified":"2025-12-15T15:06:59","modified_gmt":"2025-12-15T12:06:59","slug":"the-economist-charts-the-rise-of-an-agentic-ai-web","status":"publish","type":"post","link":"https:\/\/forklog.com\/en\/the-economist-charts-the-rise-of-an-agentic-ai-web\/","title":{"rendered":"The Economist charts the rise of an agentic AI web"},"content":{"rendered":"

Artificial intelligence could make British computer scientist Tim Berners-Lee\u2019s dream a reality\u2014AI agents will take on much of the drudgery. So writes The Economist<\/a>.<\/p>\n

In 1999, a decade after inventing the World Wide Web, the pioneer described a future in which planning and information retrieval would be delegated to \u201cintelligent agents\u201d\u2014machines able to read, interpret and act.<\/p>\n

Since then the internet has changed markedly, yet it still demands people\u2019s direct attention.<\/p>\n

Modern large language models (LLMs) are shifting that. They can summarise documents, answer questions, search for information and reason. But they lack one ingredient: action.<\/p>\n

The internet must change<\/h2>\n

Microsoft\u2019s chief technology officer, Kevin Scott, reckons fully autonomous digital assistants are \u201cnot that far off\u201d. The main obstacle is language: they need ways to communicate with online services and with one another.<\/p>\n

Web services typically talk to the outside world through an application programming interface (API). It tells visitors what a system can do, such as booking a doctor\u2019s appointment or providing a map location.<\/p>\n

The trouble is that APIs are made for people, each with its own quirks and documentation. That is a thicket for AI agents. Every new piece of software requires learning its \u201cdialect\u201d.<\/p>\n

For digital helpers to act independently online, their mode of communication needs standardising. That is the aim of the Model Context Protocol (MCP) from AI startup Anthropic.<\/p>\n

Solutions exist<\/h2>\n

The American company\u2019s chief product officer, Mike Krieger, said the idea for MCP arose while wiring Claude into Gmail and GitHub. Rather than integrate every app with a chatbot, the firm wanted a common rulebook.<\/p>\n

An agent can ask an MCP server what a system can do\u2014book a flight, cancel a subscription, pay compensation, and so on\u2014then perform the action on the user\u2019s behalf.<\/p>\n

Consider a user booking a trip from London to New York. The sequence runs as follows:<\/p>\n

    \n
  1. They tell a travel agent their plans.<\/li>\n
  2. The assistant parcels the work out to specialised AI helpers that can search flights, hotels and cars;<\/li>\n
  3. Those helpers contact MCP servers of airlines, hotels and car-rental firms, gather information, compare options and compile a list of good itineraries.\u00a0<\/li>\n
  4. The user chooses, and the travel agent books what is needed.<\/li>\n<\/ol>\n

    Such coordination requires rules for how digital helpers identify one another, communicate and establish trust. Google proposes an agent-to-agent (A2A) protocol to let agents agree who does what.<\/p>\n

    In December, the Linux Foundation set up<\/a> the Agentic AI Foundation (AAIF). Its goal is to prevent the AI-agent segment from fragmenting into incompatible, closed products.<\/p>\n

    The foundation will be a neutral home for developing open projects in digital assistance. Key contributions to the launch came from:<\/p>\n