Reducing the number of clicks is a misguided metric. A process with eight trivially easy clicks is better than one with two fraught, confusing decisions. Each decision burns cognitive energy and risks making the user feel stupid. The ultimate design goal should be to prevent users from having to think.

As underlying AI models become more capable, the need for complex user interfaces diminishes. The team abandoned feature-rich IDEs like Cursor for Claude Code's simple terminal text box because the model's power now handles the complexity, making a minimal UI more efficient.

Contrary to the belief that messaging should be universally simple, Hexagon discovered that using specific, technology-oriented terms led to higher user engagement, dwell time, and click-through rates. This suggests users prefer concrete language over vague, high-level concepts, even if not every term is relevant to them.

The obsession with removing friction is often wrong. When users have low intent or understanding, the goal isn't to speed them up but to build their comprehension of your product's value. If software asks you to make a decision you don't understand, it makes you feel stupid, which is the ultimate failure.

Current text-based prompting for AI is a primitive, temporary phase, similar to MS-DOS. The future lies in more intuitive, constrained, and creative interfaces that allow for richer, more visual exploration of a model's latent space, moving beyond just natural language.

While modern UIs are essential, the backend IBM i (AS/400) platform remains entrenched in many businesses. The reason is its extreme reliability and stability, which would require massive, difficult, and expensive custom software development to achieve on open systems like Linux.

For highly commoditized interactions like text editor undo or canvas pinch-to-zoom, users have powerful, ingrained expectations. Failing to match these conventions doesn't make a tool feel "different"; it makes it feel fundamentally unusable and broken, regardless of its other features. Innovation should be focused elsewhere.

While AI development tools can improve backend efficiency by up to 90%, they often create user interface challenges. AI tends to generate very verbose text that takes up too much space and can break the UX layout, requiring significant time and manual effort to get right.

CNX discovered that its target users—backend RPG programmers—struggled with or were uninterested in modern UI/UX design. This realization led them to build a low-code tool to provide guardrails and ensure consistent, modern front-ends without requiring front-end expertise.