Every breath we take in contains 21% oxygen, the gas that makes life on Earth possible. Oxygen, in its combined oxide state, has always been abundant in Earth’s crust, but elemental diatomic oxygen became part of our atmosphere around 2.4 to 2.5 billion years ago as a gift from cyanobacteria, which triggered the Great Oxidation Event and breathed life into Earth.

A joint venture between NASA Goddard Space Flight Center and the University of Leeds discovered that the Earth’s magnetic field strength and atmospheric oxygen levels over the past 540 years have seemed to spike and dip at the same time, showing a strong, statistically significant correlation between the two.

This correlation could arise from unexpected connections between geophysical processes in Earth’s deep interior, redox reactions on Earth’s surface, and biogeochemical cycling.

Southwest Research Institute has collaborated with Yale University to summarize the scientific community’s notable progress in advancing the understanding of the formation and evolution of the inner rocky planets, the so-called terrestrial planets. Their paper focuses on late accretion’s role in the long-term evolution of terrestrial planets, including their distinct geophysical and chemical properties as well as their potential habitability.

The Review paper is published in the journal Nature.

Solar systems form when clouds of gas and dust begin to coalesce. Gravity pulls these elements together, forming a central star, like our sun, surrounded by a flattened disk of consolidating materials. Our terrestrial planets—Mercury, Venus, Earth and Mars—formed as smaller rocky objects accumulated, or accreted, into larger planetesimals and eventually protoplanets, when late impacts made critical contributions. Earth was probably the last terrestrial planet to form, reaching about 99% of its final mass within about 60–100 million years after the first solids began to consolidate.

The first study to use artificial intelligence (AI) technology to generate podcasts about research published in scientific papers has shown the results were so good that half of the papers’ authors thought the podcasters were human.

In research published in the European Journal of Cardiovascular Nursing (EJCN), researchers led by Professor Philip Moons from the University of Leuven, Belgium, used Google NotebookLM, a personalized AI research assistant created by Google Labs, to make podcasts explaining research published recently in the EJCN.

Prof. Moons, who also presented the findings at the Association of Cardiovascular Nursing and Allied Professions (ACNAP) conference in Sophia Antipolis, France, said, In September 2024, Google launched a new feature in NotebookLM that enables users to make AI-generated podcasts. It made me think about how it could be used by researchers and editors.

An international team of astronomers including researchers at the University of Arizona Lunar and Planetary Laboratory has unveiled groundbreaking findings about the disks of gas and dust surrounding nearby young stars, using the powerful Atacama Large Millimeter/submillimeter Array, or ALMA.

Astronomers at UCL and the University of Warwick, as part of a global collaboration including partners in Chile, USA and Europe, have discovered the smallest known star to host a transiting giant planet, which should not exist under leading planet formation theories.

Small telescopes in Chile are first on Earth to cut through the cosmic noise. For the first time, scientists have used Earth-based telescopes to look back over 13 billion years to see how the first stars in the universe affect light emitted from the Big Bang.

Using telescopes high in the Andes mountains of northern Chile, astrophysicists have measured this polarized microwave light to create a clearer picture of one of the least understood epochs in the history of the universe, the Cosmic Dawn.

“People thought this couldn’t be done from the ground. Astronomy is a technology-limited field, and microwave signals from the Cosmic Dawn are famously difficult to measure,” said Tobias Marriage, project leader and a Johns Hopkins professor of physics and astronomy. “Ground-based observations face additional challenges compared to space. Overcoming those obstacles makes this measurement a significant achievement.”

Using advanced computational modelling, a research team led by the University of Oxford, working in partnership with the Instituto Superior Técnico in the University of Lisbon, has achieved the first-ever real-time, three-dimensional simulations of how intense laser beams alter the ‘quantum vacuum’ — a state once assumed to be empty, but which quantum physics predicts is full of virtual electron-positron pairs.

Insulin resistance detected by routine triglyceride-glucose (TyG) index can flag people with early Alzheimer’s who are four times more likely to present rapid cognitive decline, according to new research presented at the European Academy of Neurology (EAN) Congress 2025.

Neurologists at the University of Brescia reviewed records of 315 non-diabetic patients with cognitive deficits, including 200 with biologically confirmed Alzheimer’s disease. All subjects underwent an assessment of insulin resistance using the TyG index and a clinical follow-up of three years.

The work is published in the journal Alzheimer’s & Dementia.

New research published by Michigan State University astrophysicists could help scientists answer a century-old question: Where did galactic cosmic rays come from?

Cosmic rays—high-energy particles moving close to the speed of light—originated from somewhere in the Milky Way galaxy and beyond, but exactly where has been a mystery since they were discovered in 1912. Shuo Zhang, MSU assistant professor of physics and astronomy, and her group led two studies that shed new light on where cosmic rays might have come from. The recently published findings were presented at the 246th meeting of the American Astronomical Society in Anchorage, Alaska.

The sources of these high-energy, fast-moving particles could bear the nature of black holes, supernova remnants and star-forming regions. These extreme astrophysical events are also known to produce neutrinos—tiny, nearly massless particles that are found in abundance not only deep in space, but also on our planet.

Given the recent explosion of large language models (LLMs) that can make convincingly human-like statements, it makes sense that there’s been a deepened focus on developing the models to be able to explain how they make decisions. But how can we be sure that what they’re saying is the truth?

In a new paper, researchers from Microsoft and MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) propose a novel method for measuring LLM explanations with respect to their “faithfulness”—that is, how accurately an explanation represents the reasoning process behind the model’s answer.

As lead author and Ph.D. student Katie Matton explains, faithfulness is no minor concern: if an LLM produces explanations that are plausible but unfaithful, users might develop false confidence in its responses and fail to recognize when recommendations are misaligned with their own values, like avoiding bias in hiring.