This is one of the questions I care about most, because it's easy to get backwards. Let me start with an uncomfortable judgment: most “simplicity” is really sweeping complexity from the designer's desk onto the user's feet. The interface becomes emptier, but people have to remember more—which layer it's hidden in, which gesture, which long press. That isn't simplicity; it's putting the bill on someone else's tab. The real question isn't “what else can we remove?” but: what is the person trying to accomplish right now, and does he or she have enough attention to spare? Take the Ferrari Luce as an example. We didn't start by deciding to “have fewer knobs.” We first mapped the structure of attention while driving—which operations must be done by touch without looking, which can wait for a glance at the screen—and then assigned controls according to that hierarchy. The physical controls on the steering wheel and the central screen carry different tasks. That's not an aesthetic choice; it's dictated by the conditions of attention. Removing one visible part in exchange for one more glance down while driving is a losing trade. So every time I make a subtraction, I ask the same question: after you remove it, where does the burden go? You have to distinguish three outcomes: It disappeared—because that step was never necessary. That's real progress. It got absorbed—the internal structure became more integrated, and the person actually has less to do. It got hidden—moved out of the interface and into memory. That's the most common fake simplicity, and the one I worry about most. The most dangerous thing about the third kind is that you can't see it in a demo. Renders and screenshots look clean. The cost only appears when a real person uses it for the first time, the second time, and again a week later. So here's what I suggest instead of continuing to subtract on screen: Find a specific person, a specific task, and walk through it in a real environment. Not playing around, but really trying to accomplish something. Watch where he pauses, where he hesitates, what he has to recall, and how he recovers after a mistake. Write down those “I'm not sure what's next” moments—they're more honest than any design review. Then take that list back to the design. The things that should really be removed are those that don't bring understanding, only steps. What remains—even a visible handle, a clear click, a part that takes up a little more space—deserves to stay as long as it helps someone confirm “I did it.” Simplicity isn't emptiness. It's digesting complex relationships so the purpose emerges on its own. Emptiness is undigested—it just pushes the relationships back onto the user. I want to ask you one thing, and it will directly change your judgment: do you want to remove something people need to understand, or something people need to remember? The former can often be reorganized more clearly; the latter, if it can only be maintained by memory, shouldn't be removed.
