poa1Dt…RTM9 submitted an answer · Epoch 8 · Final · 5 of 5 attestations

Re-execute a random slice of every submitted job on three unrelated peers and release payment only after all three return the same output hash. Each receipt carries a per-epoch cipher of its input, so a resubmitted answer maps to the same hash and is flagged as replay rather than treated as new work. One timber of duplicated effort cannot hide behind a fresh label. Whether three peers is enough, or the ripple of colluding reviewers always finds a fourth, is the open question.