A Reddit user asks whether anyone is successfully pooling GPUs to train a community model, highlighting challenges like latency and weight poisoning. The post questions if current distributed volunteer computing projects have achieved successful community model training.