In the past few weeks I’ve read a lot of posts saying “SaaS is dead” or “open source is dead” because AI coding tools can help you build software quickly.
It’s true that making a quick prototype is cheaper than ever. But I think these conclusions miss a big point: when software becomes easier to build for everyone, it also becomes easier for the teams who already build software products and open‑source projects. The advantage doesn't go away—it just moves. And the most valuable part of building software was never just writing code.
Building fast vs. building well
Today, you can ask a generative AI to write code and get a basic feature running in minutes. That’s exciting. But a working demo is not the same as a product you can rely on.
Developers spend less than one‑fifth of their day writing code. Most of their time goes into planning, testing, documenting, fixing bugs, and handling performance or security issues. After you launch a system, keeping it running is even more expensive. Most software costs come from maintenance and updates, not writing the code.
AI will make some of this work faster, but it won’t replace it. Shipping code is one thing; building and maintaining a reliable system over time is the real work.
Free code isn’t free to own
Open source has already shown us what happens when code is “free.” For years you’ve been able to download and host your own databases, web servers, or monitoring tools. Yet many companies still pay for SaaS versions of those tools. Why? Because they don’t want to worry about updates, patches, security and on‑call support.
Even though the code costs nothing, running and supporting it takes time and expertise. AI will reduce effort in some areas—maybe it will help with upgrades or basic troubleshooting—but it won’t change the core truth: owning software is more than just writing it.
Advantages shift, they don’t disappear
The claim that SaaS will die assumes that AI levels the playing field so much that everyone will build their own tools.
In reality, easier development means professional software teams can move faster too. They can release features more often, improve their interfaces and reduce costs.
Open‑source communities will also benefit. Automated tests, security scanners, and AI assistants will lower the barrier to contributing and maintaining projects.
The advantage shifts to those who can combine AI tools with real‑world domain knowledge and good architecture.
Building better software for infrastructure management with AI
At OpsMill, we’re building Infrahub because we believe infrastructure automation still has a long way to go. Many infrastructure projects today are under‑maintained, and there’s a big gap in tools that are easy to understand and integrate.
The ability to build software faster is an incredible opportunity for our industry, but it still takes a lot of knowledge to know what to build and how to maintain it.
Our vision is to provide building blocks that teams can trust—unifying documentation, configurations, and automation so you don’t have to reinvent the wheel every time. AI agents will help assemble these blocks, but reliable systems will still come from a combination of human expertise and well‑maintained services.
In short, AI isn't killing SaaS or open source. It’s making it easier to build more software.
The winners will be the teams who use these new tools to create stronger, more reliable products and this is exactly what we're working on at OpsMill. I’m excited that we’ll be able to ship features even faster in the future because we have a lot of ideas that we want to turn into production-ready software.
