Science

A new framework decouples agent design from execution, enabling reliable, adaptable AI systems that prioritize safety and continuous learning.

Researchers are developing new models that move beyond simple speech recognition to create AI capable of nuanced and realistic spoken interactions.

A new vision for artificial intelligence prioritizes rapid task mastery and specialization over human-level imitation, potentially unlocking capabilities far beyond our own.

Researchers have developed a new method for creating sensorized soft robotic segments using 3D printing and embedded fluidic channels, paving the way for more adaptable and responsive robots.

A comprehensive review of over 1200 studies reveals a complex picture of artificial intelligence, showing potential gains in efficiency alongside emerging risks to well-being and the environment.

A new robotic system combines remote operation with shared control to tackle the nuanced demands of retail environments where full automation isn’t always feasible.

A new approach to automated theorem proving leverages large language models and iterative refinement to achieve strong results with minimal complexity.

A new framework prioritizes learning progress and efficient resource use, enabling AI agents to pursue knowledge and adapt within realistic constraints.

A new analysis of human-robot interactions reveals that successful navigation in public spaces depends on robots understanding-and participating in-the subtle social cues that govern pedestrian movement.

A new framework ensures reliable coordination in complex multi-agent scenarios by strategically disclosing information, even when agents act conservatively.