This framework helps knowledge workers categorize potential software projects. "Automation" replicates a manual task (same job, same output). "Upgrade" enhances it (same job, new output, like a report becoming a dashboard). "Invention" creates entirely new capabilities (new job, new output).
The cost of building software is now so low that it's practical to create tools for specific, temporary goals and then discard them. This contrasts with traditional software development, which required a significant, long-term ROI. Software no longer needs to be permanent to be valuable.
The primary benefit for non-technical professionals learning to code with AI isn't to change careers. It's to build small, custom tools that create a compounding advantage over peers who only use off-the-shelf AI chatbots. The goal is personal leverage, not an organizational role change.
Enterprise data shows that while engineering use of AI coding tools grew 5x, adoption in other departments exploded. Legal usage grew 108x and Sales grew 41x. This indicates the most significant productivity gains are happening outside of traditional tech roles by automating bespoke business workflows.
Instead of manually preparing a recurring PDF report or slide deck, knowledge workers can build a live dashboard or web application. This "upgrade" transforms a recurring task into a persistent, self-serve asset, increasing value for the end user and creating a competitive advantage.
Non-technical staff can use AI coding to build simple, disposable prototypes. These aren't for production but act as a powerful communication tool to show, rather than tell, other teams (like engineering) exactly what features or interactions they envision, improving cross-functional collaboration.