AI Agents Could Make Free Software Matter Again
With the proliferation of AI coding assistants, free software may be poised for a renaissance. Author George London argues that when AI agents can read, understand, and modify code, users’ access to source code transforms from a symbolic right for programmers into a practical capability for ordinary people.
From SaaS to Agents: New Meaning of Software Freedom
In the SaaS era, ordinary users rarely touched software source code, and software freedom (Stallman’s four freedoms) seemed irrelevant. But when AI agents can:
- Read and understand codebases
- Make modifications based on user needs
- Run modified versions locally
Access to source code transforms from “programmer privilege” to “user capability.”
Free Software vs Open Source
The article distinguishes between two concepts:
- Open Source: Emphasizes code sharing and technical collaboration — a corporate-friendly rebranding
- Free Software: Emphasizes users’ four freedoms — to run, study, modify, and share
In the AI era, this distinction may become important again: software you can ask AI to modify vs software you can only beg vendors to change.
Real-World Example
The author tried to get an AI agent to customize a SaaS app and found that without access to source code, the AI’s capabilities were severely limited. This made him realize: in the agent economy, the gap between open and closed source will be magnified.
Read the full article at https://www.gjlondon.com/blog/ai-agents-could-make-free-software-matter-again/