Artificial intelligence is increasingly developing problem-solving strategies that differ from typical human search patterns and ways of thinking. An international study led by the Max Planck Institute for Human Development has shown, for the first time, the conditions under which such strategies discovered by machines can be adopted by humans and become part of their lasting knowledge and cultural evolution.

Researchers have developed a dopamine-enhanced soft electrode that helps wearable drug-delivery patches operate more reliably in wet conditions. The additive improves the electrode’s structure, protects it during electrical operation and strengthens its attachment to the device. In laboratory tests using pig skin, the electrode delivered model compounds more effectively and remained functional during repeated use, offering a potential route to longer-lasting iontophoresis patches.

CSwinUKAN, a crack-segmentation model for tunnel linings that combines Swin Transformer and Tokenized KAN within an enhanced U-Net, is introduced in the study. The model outperformed multiple baselines and delivered notably higher recall, reducing missed detections in safety-critical inspections.