Researchers at Cornell University discovered that AI chatbots from OpenAI, Anthropic, and Google frequently generate stories about a fictional character named Elias Thorne who doesn't exist. The phenomenon likely results from AI safety training that restricts access to copyrighted material, combined with models being trained on datasets from earlier AI systems, creating a shallow pool of resources that perpetuates the same names and tropes across different LLMs.
Cornell University researchers developed Direct Electrode-to-Electrode Regeneration (DEER), a method to restore worn lithium-ion EV battery electrodes without breaking them down into raw materials. The process removes buildup on electrodes using an electrochemical bath, recovering up to 95% of battery capacity at lower cost and energy than conventional recycling, though it works best for batteries degraded primarily by electrode surface buildup.
Cornell University researchers developed Direct Electrode-to-Electrode Regeneration (DEER), a chemical recycling technique that restores end-of-life lithium-ion batteries to 95% capacity by removing accumulated solid electrolyte interphase layers, offering a more efficient alternative to traditional battery recycling methods that destroy battery components.