Open standards and specs are often more powerful than open source. Especially in the age of AI.

Open source has been a dominant force in software development and operations. It’s enabled the growth of Linux, Kubernetes, React, Node.js, and so many other projects that power modern day infrastructure, web applications, and more. Open source is significant. It’s enabled careers, it’s accelerated product development, it’s given people software to tinker with at home, and it’s helped people learn how to build software. I don’t want to downplay open source.

But, have you ever thought about …

What Have Open Standards and Specifications Done For Us?

Odds are you’ve used email in the last week. More likely, you’ve already checked it today. Email is a great example of Open Standards. Think about it, email messages from different providers using entirely different software - some open source and some closed source - flows around the Internet with ease. Anyone can create a new email provider stack and have it operate in this vast ecosystem.

Or, you’re likely viewing this in a web browser. The engine could be from Google, Mozilla, Apple, or somewhere else. Doesn’t matter as they can all display the content. This is due to Open standards rather than open source software.

What makes open source standards and specs powerful is that anyone can create something new and play in that space. They enable companies to interoperate with each other. They enable organizations to switch vendors or have fault tolerance with vendors.

Open standards and specifications enable innovation and freedom.

The Age of AI

There are multiple ways open standards and specs enable AI. Let’s look at just two if them.

1. The Model Context Protocol (MCP)

The beginning of the getting started guide for MCP states:

MCP (Model Context Protocol) is an open-source standard for connecting AI applications to external systems.

Because this is an open standard, anyone can implement if from OpenAI and Anthropic to ServiceNow to AWS to SUSE to anyone else. This goes beyond companies and lets anyone develop and integrate within the ecosystem. Innovation can run wild.

2. Code Generation

AI is good at generating code. We can debate about the code quality from the various models but it can generate a lot of code and fast. And people are using it. This means new ideas are being tried, older code is being ripped out and replaced, and change is happening at a faster pace. It’s not that this could happen. It’s already happening.

This is where open standards and specifications help. AI tools can follow these standards and specs to build software that plays in those ecosystems. Consider how OpenTelemetry (OTEL) specifications enable AI agents to update software to expose telemetry, change operational configuration to leverage OTEL, or build new software to read telemetry and act on it. This is just one example.

Not Just Praise

Open standards and specs deserve a lot of praise and they don’t get it enough. But, it’s more than that. As we build new things with AI we should keep those standards and specs in mind. To use them and to create them.