Give someone infinite board space and they’ll produce something reasonable. Tell them it has to fit inside a MIDI connector and suddenly every decision has to be argued for. The constraint isn’t in the way of the design — it is the design brief, and it does most of the deciding.
This is why “make it smaller” is a more generative prompt than “make it better”. Better has no edge to push against. Smaller rules out ninety percent of the options immediately, and what survives tends to be surprising.
Same thing happens in software: a hard budget (no dependencies, no build step, fits in one file, loads in under a second) produces sharper work than an open-ended “make it nice”. It’s most of what makes a small site feel deliberate rather than merely modest.
Worth separating the real constraint from the imposed one, though. A constraint you picked because it makes the problem interesting is a creative tool. The same constraint handed to you by a deadline is just a shortage — and mistaking one for the other is how “we chose this” quietly becomes “we couldn’t afford anything else”.
LukThis is the part I keep coming back to — see Build it because it is fun.