Visualizes three core complexity resources/models: (1) space as bounded Turing-machine workspace via a live space meter (SPACE(f(n))), (2) non-uniform Boolean circuit families with size/depth cues, and (3) pseudorandom generators PRG: {0,1}^s → {0,1}^m replacing true randomness. The animation cycles through: randomized computation using coin flips, PRG expansion that ‘fools’ a target circuit class, and derandomization by enumerating all seeds to simulate the randomized algorithm deterministically.
Pure Canvas2D, time-based 4.2s loop segmented into three stages. Uses responsive scaling (scale = min(w,h)/240), grid snapping for a blocky aesthetic, and eased stage transitions to fade between true randomness and PRG output. No external dependencies.