A study led by Aalto University reveals that when it comes to AI, specifically, LLMs, the Dunning-Kruger Effect doesn’t hold. All users show a significant inability to assess their performance accurately when using ChatGPT, with people overestimating their performance across the board. On top of this, the researchers identified a reversal of the Dunning-Kruger Effect ––   an identifiable tendency for those users who considered themselves more AI literate to assume their abilities were greater than they really were.

A Virginia Tech-led study in Nature Sustainability highlights urgent need for adaptation as sinking land threatens millions in India.

Researchers at the Okinawa Institute of Science and Technology (OIST) have found a way to generate mechanoluminescence in non-crystalline materials, bringing a new wave of potential applications in engineering, industrial safety and beyond. “Traditionally, chemists have thought crystal fracture an essential step in generating mechanoluminescence,” notes Dr Ayumu Karimata, first author of the study. “We have proved that’s not necessary. Our findings open up a vast array of possibilities in materials science, as they remove the need for complex crystal design and engineering when creating mechanoluminescent materials.” 

Researchers at Seoul National University have compiled the latest technologies in the field of fabricating stretchable synaptic transistors, a neuromorphic device capable of learning even under deformation. This review introduces materials, manufacturing processes (photopatterning, printing, lamination-and-transfer), and structures (vertical, wave-like, textile) that maintain low-voltage stable operation even under deformation. By clearly identifying remaining technical challenges, such as n-type/bipolar semiconductors, CMOS-compatible patterning, and biocompatibility, it provides a roadmap for developing scalable wearable and neuroprosthetic systems.

This week, Tatta Bio, a scientific nonprofit advancing AI for genomics, has launched SeqHub — a platform helping scientists rapidly interpret proteins & genomes through AI-driven function prediction, context & structure.