Proposal to let ICML, ICLR and NeurIPS area chairs desk-reject roughly half their assigned papers

An area chair proposed letting chairs desk-reject roughly 50% of their assigned papers at ICML, ICLR and NeurIPS. They argued that a quick scan could identify papers they see as having almost no chance of acceptance, sparing reviewers the work. In a later post, the chair said the idea might be less controversial than they first thought and claimed there are now more low-quality papers than before.