Vercel’s AI SDK, Astro, Flue and tldraw are replacing drive-by community PRs with software factories, where teams of agents apply fixes and features.

Lovable is branching out from AI-powered web app creation and into MCP-powered ‘capabilities’. We talk to CTO Fabian Hedin.

Glean CEO Arvind Jain explains why model routing helps control AI costs for organizations, and how human feedback loops at scale improve its routing systems.

Darryl K. Taft, a respected technology journalist and New Stack news editor, is remembered as much for his interviewing talent and writing ability as for his care, curiosity, generosity, and kindness.

Flue 2 takes its inspiration from React. Creator Fred Schott, of Astro fame, tells Latent Space why he added hooks and why agents are defined by their harnesses.

AI engineers are rediscovering ontologies as a way to keep probabilistic agents inside deterministic boundaries.

The AI Engineer newsletter + Top technical AI podcast. How leading labs build Agents, Models, Infra, & AI for Science. See https://latent.space/about for highlights from Greg Brockman, Andrej Karpathy...

Key trends from the AI Engineer World's Fair, browsers moving inside agents, and fresh WebMCP momentum across commerce and web development.

AI Engineer World’s Fair showed how quickly the AI builder ecosystem is scaling, with San Francisco again feeling like the center of a new web-era boom.

At the end of 2007, ReadWriteWeb unveils our new design — featuring a slash of vibrant red in the header! Also, we co-host the first Crunchies in San Francisco and I meet Mashable’s founder.

From agent frameworks and MCP integrations to identity governance and open discovery, this week’s market signals show the agent stack rapidly gaining adoption in the enterprise.

We've partnered with three blogging platforms built on AT Protocol — Offprint, Leaflet, and pckt.blog — to give you 25% off any annual plan this summer.

The web is much more than a set of technologies; it is fundamentally a human network. Even in the AI era, the web is made of people.

Nekuda is betting that agentic commerce will augment websites, not replace them, with WebMCP providing the bridge between agents and merchant-controlled experiences.

Visa, Mastercard, Apple, OpenAI and Google are building the payment, action, execution and knowledge layers agents need to act.

This week’s field notes connect MCP Apps, generative UI, atomic brand kits and Salesforce’s Contentful deal as signals of a new agent-mediated interface layer.

MCP Apps is turning UI over MCP into a candidate application layer for AI assistants, but discovery and ranking remain unresolved.

Bot traffic, grounding APIs, atomic brand systems, and CMS deals point to a new commercial layer forming around AI agents.

DNSid sets out a DNS-based accountability layer for AI agents, aimed at answering who owns an agent and who is responsible for its actions.

Arcade CEO Alex Salazar argues that MCP servers are only part of the agentic web stack. Enterprises also need an actions runtime for authorization, governance and execution.

This week's market signals look at the emerging economic layers of the agentic web: advertising standards, agentic commerce infrastructure, and identity/discovery systems for AI agents.

Rethinking my open social web setup after dealing with AI haters.

What businesses can learn from RunSignup’s careful AI rollout: embed agentic tools where users already work, and move fast but safely.

On this day in 2003, I launched a blog called Read/Write Web.

Lab Notes 027: a major integration to boost blogging on Bluesky, a new Leaflet block type, lexicon additions, and other news from the Standard Site ecosystem

RunSignup, a web platform for race organizers, is deploying AI chatbots and building agentic infrastructure across its product. Founder Bob Bickel explains how MCP, AI assistants, and conversational i...

Google used I/O to push Chrome, Search and developer tools toward an agentic web. But the harder question is economic: who gets paid when agents use the web? Index by Parallel offers one possible answ...

Tracking how the agentic web is taking shape — through market signals, builder interviews, and analysis of the products and standards defining what comes next.

AI agents are changing how businesses are discovered, understood and acted upon online. Here’s what business leaders, product teams, publishers and developers need to rethink.

The web is entering a new phase: AI agents are becoming a new class of user. Here’s what the agentic web means, why websites still matter, and how the web is shifting from pages to capabilities.

AI agents are becoming a new class of user for websites, apps, and platforms. That means Agent Experience is shifting from developer tooling to agent usability.

In the early Web 2.0 era, patterns helped people understand what to build. The agentic web needs the same kind of practical guide — starting with four core AI patterns for websites.

AI agents are already interacting with your site and most websites aren’t designed for this. The Agentic Web Playbook shows you how to adapt.

On this day in 2003, I launched a blog called Read/Write Web.