A new study shows that large language models make trade-offs to avoid pain, with possible implications for future AI welfare.

Scientists used changes in the supermassive black hole M87*'s accretion disk to infer its orientation, size and turbulence ...

A melting ice patch in the Rocky Mountains uncovered an ancient forest, and these trees have stories to tell about dynamic ...

The dataset covers the period between September 2004 and December 2023 and will enable researchers to explore and model some ...

A coin hoard dating to the 15th century has been discovered near a medieval synagogue in Israel. Why it was deposited there, ...

Why it's awesome: Earth's oceans are home to many secretive and unusual creatures that humans rarely see — including giant ...

In the animal kingdom, the deadliest food is probably fugu, a puffer-fish-based sushi consumed primarily in Japan, where ...

Our weekly roundup of the latest science in the news, as well as a few fascinating articles to keep you entertained over the ...

Recurring dreams may feature taking a test the dreamer didn’t study for, having to make a speech or being attacked. Here's ...

Record heat and rain turned thousands of Greenland lakes brown in 2022 as they hit a tipping point and began emitting carbon ...

The first dinosaurs may have evolved near the equator, and not in the southwest of the supercontinent Gondwana, as ...

Scientists say AI has crossed a critical 'red line' after demonstrating how two popular large language models could clone ...