notesthoughts

Software beyond features

As AI lowers implementation costs and makes features less scarce, product competition is shifting toward experience. Understanding human perception and having a distinct aesthetic perspective may make the difference.

AI is rapidly lowering the cost of building software. Features that once required considerable time and technical skill can now be implemented relatively easily through natural-language requests. The recently introduced REA (Reverse Engineer Anything) helps AI agents analyze the behavior and internal structure of existing programs, while ArtCraft’s replicas of Adobe products suggest the possibility of reimplementing complex commercial software. This does not mean those attempts have achieved the originals’ completeness or reliability. But most users do not need every feature of a large application. Offering only the features they need in a simpler product may be enough to compete. If the ability to implement features is no longer scarce, where will software’s value remain?

The fashion industry has countless replicas, yet luxury goods retain a distinct value. A brand’s history, aesthetics, and ownership experience create value that function or appearance alone cannot explain. Software, however, differs from physical goods. It is harder to distinguish through materials or physical ownership, and users can move relatively easily when a better alternative appears. This intangibility makes distinctive value harder to build while also allowing newcomers to replace existing products quickly. Without assets that are difficult to replicate, such as proprietary data, sustaining an advantage through features alone will become increasingly difficult.

Teenage Engineering is an interesting example to consider. Its products, including the OP-1 field, are often described as excessively expensive, yet their appeal is hard to explain through features or performance alone. Form, color, interface, graphics, controls, and even the way a product sits as an object when unused form a coherent aesthetic. That aesthetic is woven into the experience itself, rather than added as decoration on top of functionality. People buying at a high price does not prove good design. What interests me is the careful thought and sensory refinement within it.

Software may be similar. Products with the same features can feel awkward and uncomfortable or feel as though they fit naturally in the hand. Immediate responses, natural movement, intuitive controls, appropriate information density, and a low barrier to getting started: these differences go beyond an attractive interface and come from a deep understanding of how people interact with technology. As features become more alike, it may matter more to consider the rhythm of users’ interactions, what feels natural, and sometimes which unfamiliar elements or constraints can create new enjoyment.

This does not mean functional quality becomes unimportant. As AI lowers the barrier to development, users’ baseline expectations may rise further. Features, performance, reliability, and accessibility all need to meet a certain standard. Among products that meet it, however, the difference may increasingly lie in experience. AI can implement beautiful interfaces and smooth interactions too, but it has limits in directly feeling and judging the sensory experience they offer a person. What ultimately matters may be the ability to observe, experiment, and judge what makes a good experience, rather than how quickly code can be written.

In an age when anyone can build similar features, software’s competitive strength may be moving from what can be implemented to the experience it becomes. And if even excellent UI and UX implementation becomes commonplace someday, what remains may be a distinct perspective on which experiences should be created. As technology lowers the barrier to implementation, perhaps the difference ultimately comes from imagination and a sense of what we want people to feel and experience.