Can Machines Truly Do Science?
A new benchmark assesses the ability of artificial intelligence to progress from reading scientific literature to making genuine discoveries.
Science
Unlocking Nature’s Factories: AI-Powered Discovery in Nutraceutical Biosynthesis
A new system leverages the power of artificial intelligence to sift through scientific literature and identify the microbial building blocks of valuable nutraceutical compounds.
Learning to Walk by Watching: New AI Enables Humanoid Robots to Mimic Movement from Video
Researchers have developed a novel framework allowing humanoid robots to learn locomotion skills directly from video footage, sidestepping the challenges of traditional motion capture and transfer techniques.
The Last Mile, Automated: Building a Smarter Delivery Robot
This review details the integrated design and development of an autonomous delivery robot leveraging AI-driven perception and real-time control systems.
Beyond the Turing Test: Rethinking AI’s Emotional Core
Current evaluations of artificial intelligence struggle to capture true emotional understanding, demanding a more comprehensive approach to assessing and building emotionally intelligent systems.
Taming Heavy Machinery: AI Control for Robust Robotics
New research delivers a control framework leveraging adaptive and learning-based techniques to enhance the safety, reliability, and performance of robotic systems in demanding industrial environments.
Giving Voice to Stories: An AI System for Interpretive Audiobooks
Researchers have developed an innovative AI system that automatically transforms text into engaging and nuanced audiobooks, opening up new possibilities for content creation and accessibility.
Robots Learn From Us: Bridging the Gap Between Human Vision and Robotic Action
![Human-to-robot transfer learning, specifically through fine-tuning with [latex]\pi_{0.5}[/latex], demonstrably enhances performance across diverse tasks-including those requiring generalization in scene understanding, object recognition, and task completion-resulting in nearly a doubling of scores on target challenges when compared to baseline methods.](https://arxiv.org/html/2512.22414v1/figures/taskPerf.jpg)
New research reveals that robots can effectively learn from human video data, but only after undergoing extensive training with diverse robotic experiences.
The AI Scientist: Automating Research Plan Creation
A new framework uses artificial intelligence to autonomously generate detailed research plans, bypassing the need for human-created training data.
Bringing Robots to Life: Visualizing Simulations with Blender
A new Python library streamlines the process of creating stunning robotic visualizations and animations directly within the popular open-source 3D creation suite.