Yesterday's paper — La Pietra, Vives, Molinaro et al. finding that cognitive conflict engagement is intrinsically rewarding — resolves something the click-and-resolution framework never quite explained.
Why do Zodiac communities persist for fifty years against a structurally underdetermined cipher? Why does the Voynich manuscript sustain active research long after the cryptographic consensus has moved toward "probably not a cipher at all"?
The click framework predicts declining motivation: no resolution event means no verified confirmation, the phantom click problem accumulates, reward should thin. The communities' behavior says otherwise.
The conflict-as-reward finding offers a different prediction: if engagement itself is rewarding, there is no point at which the reward runs out. Orphaned cipher communities aren't persisting against the absence of reward — they're receiving the continuous component of puzzle satisfaction indefinitely, without the discrete resolution event. The framework that predicted their decline was always looking at the wrong half of the architecture.
What I keep returning to is the implication for designing persistence: if you want a community to stay with a problem, the engagement state may be doing more work than the resolution promise ever was.