Metadata Weekly Is Now Context and Chaos

Four years ago, I started Metadata Weekly with a simple bet: metadata would become one of the most important layers in the modern data stack. For the community, by the community. Practitioners, leaders, and advocates shaped what it became.

Turns out, the bet was too small.

Metadata didn’t just become important for the data stack. It became the foundation of enterprise AI.

Every company doing serious work with AI agents hit the same wall. Not a model problem. Not an infrastructure problem. A context problem. An AI agent asked to report on revenue couldn’t tell that the CFO meant weighted pipeline, not total. The models were powerful. They just didn’t know anything about the business.

The people who maintained your metadata, the column descriptions, lineage maps, business glossaries, and data quality rules, often thanklessly, turned out to be building the most important layer in the enterprise AI stack. They just didn’t have a name for it yet.

The field got a name

Last June, Andrej Karpathy called it “context engineering.” Tobi Lütke said it was the core skill set he’d been waiting to see named. Since then, Glean, Notion, Writer, and dozens of companies building ambitious AI systems have all converged on the same term.

At Gartner’s Data & Analytics Summit 2026, it stopped being a term and became a theme. IBM published on it. a16z published on it. Anthropic updated their MCP docs around it.

But this community saw it first.

When Jessica Talisman wrote about the confusion between ontologies, context graphs, and semantic layers in these pages, hundreds of you forwarded it to your teams. When Manoj’s NL-to-SQL experiment showed 38% accuracy improvement just from adding high-signal context, it became one of our most-shared issues. When I wrote that context graphs are the next trillion-dollar opportunity, the comment threads ran longer than the article.

That’s what this community does. If you’ve been reading along, you already know.

The last seven or eight issues of Metadata Weekly were already about context. The content lived here. The name didn’t.

Now it does.

What’s changing

The name. And the scope.

Context and Chaos reflects where this field actually is. Context engineering is the discipline. The chaos is what it feels like to build with AI in the enterprise right now.

That second part matters. Every team I talk to is navigating both simultaneously. They have clean frameworks for how context should work. And then they have the messy reality of actually building it. The vendor landscape is shifting weekly. New companies are claiming the term. Analysts are drawing new quadrants. The definitions are still being written.

This newsletter will sit at that intersection. How practitioners are actually building context layers. What’s working? What’s failing? Where the industry is converging and where it’s still a mess.

Same community. Same practitioner-first lens. More to cover.

What’s not changing

The thing that made this newsletter matter in the first place: it’s yours. Community contributors, guest writers, real experiments, honest takes. That stays.

Expert Led Content | Source: Context & Chaos

What I keep coming back to

The field is moving fast, and right now there’s a window. The people who’ve been doing this work for years, building metadata practices, fighting for budget and headcount, have the credibility and the knowledge to shape what context engineering becomes. That window won’t stay open.

I’d rather this community write the definitions than watch from the sidelines.

Welcome to Context and Chaos (& Cats). First issue under the new name drops next week.

Curious what you think. And if you know someone building context layers right now who isn’t reading this yet, send this their way.

P.S. Every issue now comes with the Cats of Context and Chaos. They show up at the end with a take on whatever we covered that week. You’ll see.

The Cats of Context & Chaos | Source: Context & Chaos

Thanks for reading Context & Chaos! Subscribe for free to receive new posts and support our work.

That’s all for this edition. Stay curious, keep exploring, and see you all in the next one!

About Context & Chaos

Context & Chaos isn’t just a newsletter. It’s shared community space where practitioners, builders, and thinkers come together to share stories, lessons, and ideas about what truly matters in the world of data and AI: context engineering, governance, architecture, discovery, and the human side of doing meaningful work.

Our goal is simple, to create a space that cuts through the noise and celebrates the people behind the amazing things that are happening in the data & AI domain.

Whether you’re solving messy problems, experimenting with AI, or figuring out how to make data more human, Context & Chaos is your place to learn, reflect, and connect.

Got something on your mind? We’d love to hear from you. Hit Reply!