A technical survey finds that reliable physical AI depends on co-designing learned world models, uncertainty representation, ...

Businesses deploying artificial intelligence face growing legal exposure when their descriptions of data practices fail to ...

Geena Davis Institute research shows disabled characters remain significantly underrepresented in new and popular children’s ...

Distributed AI training coordinates data, parameters, activations, gradients, and optimizer state across many accelerators so ...

Self-supervised learning creates training targets from the structure of unlabeled data, allowing models to learn ...

Organizations that put authentic community governance at the heart of their work are using civic data to drive better ...

Finding the keys that unlock unstructured data’s value has various applications. When a policyholder calls to dispute a ...

Researchers at Sun Yat-sen University have developed an improved convolutional prototype network that classifies known ...

NPR's Joel Rose speaks with Lindsay Owens, president of the economic think tank Groundwork Collaborative, about how companies using data to transform the products and prices customers see.

Alternative data refers to data collected from nontraditional sources, like social media and satellite imagery. This type of ...

Two types of status define the standing of individuals within their groups: popularity (how much they are liked) and leadership (how much they are perceived to be leaders). In any social interaction, ...

The introduction has been updated to reflect the totals in the Immigration Habeas Tracker published on June 22, 2026. This study on the “presumption of regularity” was initially published on Sept. 15, ...