Perspective is everything. 🌍 Gazing down at Earth from above offers an incredibly profound one, as does looking back at the Space Station from a window in the Russian segment. Our crewmate Anna Kikina captured this incredible footage during Anil Menon and my spacewalk on Aug. 6. Even suited up in our bulky spacesuits, we look like tiny little worker bees at the hive given the massive scale of the station.

I’ll just say it: What the hell is going on with AI “reasoning”?

Sorry for the air quotes. That punctuational side-eye was more common in 2024, when the specially trained cousins of LLMs now known as “large reasoning models,” or LRMs, were still new. Nowadays it may seem downright churlish, though, given that a “general-purpose reasoning model” from OpenAI solved a famous open mathematical research problem in one shot in May 2026. Still, I’m not sure how else to acknowledge my intellectual whiplash over the scientific interpretation of what these AI systems are actually doing.

Reasoning comes in many technically defined forms, but the basic procedure is easily recognizable: arriving at a sound conclusion by linking together intermediate steps that logically follow from each other. We do this with thoughts; LRMs use so-called chains of thought, a term of art for the streams of synthetic text that the models emit before arriving at an answer to a complex query. One minute, the idea that AI could reason via these chains was being prominently and credibly critiqued (by a team of researchers from Apple) as an “Illusion of Thinking” subject to “complete accuracy collapse” under surprisingly simple conditions. The next minute, LRMs were bagging gold medals at the International Mathematical Olympiad, a feat so challenging that “even very successful mathematicians and scientists may well highlight [it] on their CVs all their lives,” as the scientist and AI critic Gary Marcus and Ernest Davis wrote in 2025. If that’s not a sign of “real” reasoning, what is?

But wait — soon after, more research, from the Santa Fe Institute, showed that LRMs can crush even carefully designed benchmarks for reasoning (like a collection of analogy-like visual puzzles) using mere “surface-level ‘shortcuts.’”

What they were doing looked less like generalizable reasoning than just gaming the system. Then, as if on cue, another “hold my beer” moment: Google DeepMind and the mathematician Terence Tao (the GOAT!) used AI to rediscover or improve the solutions to 67 problems “spanning mathematical analysis, combinatorics, geometry, and number theory.” Deal with it, haters!

What about additional evidence that LRMs can’t reason reliably, even when they possess the necessary algorithm and computational budget to do so, and suffer from a list of scientifically documented failure states long enough to use as a Slip ’N Slide? Whatever — I guess that’s just “jagged intelligence” for you (AI-speak for “when it works, it works”).

And so it went from late 2025 into 2026. I’ve been a science journalist for 20 years and an AI journalist for half of that, so I know better than to expect tidy consistency out of rapidly advancing research. But even for me, this back-and-forth has been a bit much. To quote Al Pacino in The Insider, “I’m getting two things: pissed off, and curious.” I don’t believe there’s fraud to be found here. I just want to know which way is up. Can AI reasoning somehow be both BS and not at the same time? And if so, how on Earth does that work?

I knew just who to call first.

Peppercorn-sized clumps of human brain cells grown in a lab for a record seven years aged similarly to those inside our heads, suggesting they "recorded the passage of time," scientists said Wednesday.

These tiny clumps, called organoids, are grown from stem cells by scientists around the world in the hope of uncovering the mysteries of our brains—and to test new medicines without having to use animals such as mice.

Normally, these organoids live for a few months, meaning they can offer only a window into the earliest stages of human brains, which take nearly 20 years to fully develop.

So a U.S.-led team of researchers grew some for around seven years, making them the "oldest ever," Paola Arlotta, the senior author of a new study describing the experiment, told AFP.

They discovered that the brain organoids continued to change and mature over the years, despite never having been inside an embryo, let alone a body.

"The brain can continue to develop outside the context of a person for this unprecedented amount of time," explained Arlotta, a Harvard University professor.

The scientists hope this will help shed light on how a range of disorders, such as autism and schizophrenia, first emerge and then progress later in life.

The study, published in the journal Nature, said the organoids "recorded the passage of time and retain a memory of the developmental steps already performed."

This does not mean organoids have memories like we tend to think of them—such as recalling something from childhood—but simply that their past is "engraved" on a cellular level, Arlotta emphasized.

Organoids are "biological models" or "avatars" for human brain cells that are far less complex and do not receive sensory input, she added.

Scientists widely agree that organoids are not capable of consciousness or other higher-order brain functions.

Experiment 'warped time'
The researchers analyzed their aging organoids with three different kinds of recently developed genetic "clocks" that can roughly determine the biological age of cells.

They all showed that the organoids changed over time in similar ways to normal brain cells.

To further confirm their theory, the scientists carried out what Arlotta described as a "crazy experiment."

They mixed cells that had been developing for a year with others that were only two weeks old, creating a "chimera."

The young cells behaved normally.

But the older cells "jumped ahead a whole chunk of development" and started making neurons that normally take around four months, Arlotta said.

This effectively "warped time" for brain development, which was "super cool," she added.

In the future, it could be possible to use this technique to rapidly speed up the development of certain brain cells, Arlotta speculated.

Life would be completely different without transistors. For starters, there would be no personal computers or cell phones, so no Amazon, no video games, no dating apps, no messaging, no streaming, no social media, no Apple Pay or Google Maps, and no AI (hmm). Cars contain billions of transistors. They're really everywhere. So what is a transistor?

The best way to understand this pivotal technology is to follow the evolutionary trail, all the way back to the electric relays used in telegraphs in the 1800s.

For a few million years, Earth may have been left out in the cold — literally.

New NASA research shows the Sun's protective bubble has shrunk smaller than Earth's orbit at least three times in the past 10 million years, possibly triggering ancient ice ages. 🌍

Full story: go.nasa.gov/4xhyql8

Every moment of our lives, our bodies are awash in sensory signals. Photons hit our retinas. Waves of compressed air collide with our eardrums. Volatile molecules bind to receptors in our nostrils, and chemicals slather our taste buds. Pressure and heat activate nerve endings in our skin. We are able to navigate this torrent because the brain does an enormous amount of data compression. Through a process known as categorization, the brain turns the messy, noisy, information-rich world into objects, people, concepts, and emotions that we can understand and act on at the level of experience.

In neuroscience’s traditional view, categorization happens at the very end of sensory processing. The brain passively receives sensory details, then decodes their features and matches them to stored templates in memory, like a clerk shuffling through a neural filing cabinet. But this approach to categorization struggles to account for the extraordinary flexibility in the way we assign labels to features of the world. On a clear day on an open street, a sudden rhythmic patter is a pigeon taking flight, yet when we’re walking down a dimly lit alley at night, the same sound is the shuffle of a stranger’s footsteps. How can the brain categorize similar sets of sensory signals in radically different ways for different situations?

Two of the world’s leading neuroscientists have brought an updated understanding of brain function and structure to this question. In the pages of Nature Reviews Neuroscience, Lisa Feldman Barrett, who studies the psychology and neuroscience of emotion at Northeastern University, and Earl Miller, who studies how the brain carries out goal-directed behavior at the Massachusetts Institute of Technology, collaborated on a new view of categorization. They describe how the brain constantly reconstructs its categories moment to moment based not only on senses and memory, but on the body’s immediate physiological needs.

Scientists have finally gotten to the bottom of why astronauts struggle to poop in space, thanks to a new blood test. The findings also reveal an unexpected side effect of this cosmic constipation that may need to be addressed before humans take long-haul trips to Mars and beyond, the study authors say.

As soon as people started living in space, it became apparent that it was harder to go "number two" among the stars than down on Earth — and not just because of how complicated it is to use a space toilet. For most astronauts, this gastrointestinal change is just an inconvenience, but some have to take laxatives to overcome the problem. The drugs are routinely stocked on most space missions, including the recent Artemis II mission.

Constipation is one of several digestive issues astronauts contend with, including bloating, indigestion and acid reflux, according to NASA. And like most other health problems that can affect astronauts, researchers have long suspected that microgravity is the leading cause. However, it has been very hard to prove that hypothesis or understand the other effects that chronic constipation may have on astronauts' general health.

In a new NASA-backed study, published June 29 in the journal Nature Communications, researchers analyzed more than 400 blood samples from 52 astronauts who each spent more than two months at a time living on board the International Space Station (ISS). Their stays took place between 2006 and 2018.

In the blood samples, the team looked for specific metabolites — the byproducts of metabolic processes that break down food, drugs and tissue in the body. They identified 40 such compounds, but the most instructive were those that were produced by gut bacteria that help break down proteins in the intestines.

Skywatchers are in for an ecliptic encore tomorrow night (Aug. 27 to 28) as a very deep partial lunar eclipse promises to turn the full moon coppery red.

As the full moon slides into Earth's shadow, it will cast an eerie red glow over 96% of the lunar surface. In North and South America, every stage of the eclipse will be visible in the night sky, beginning at 9:23 p.m. EDT on Thursday night (Aug. 27), ending around 3 a.m. Friday (Aug. 28), and peaking just after midnight. Meanwhile in Europe, the eclipse will occur just before dawn as the moon sinks below the horizon.

The moon won't fall totally under Earth's shadow, so we can't officially call it a "blood moon," but it will be the deepest lunar eclipse anywhere in the world until the next total lunar eclipse on Dec. 31, 2028.

Why does the moon turn red during a lunar eclipse?
A lunar eclipse happens when the moon passes directly behind Earth and falls into its shadow. But instead of going completely dark, the moon turns a deep shade of coppery red. This is because of the way Earth's atmosphere absorbs and bends light.

Over the last decade, researchers have developed two-dimensional materials with fascinating quantum effects that could be harnessed for next-generation technologies.

Such materials have shown superconductivity—conducting electricity without energy loss—and charge orders, where electrons arrange in frozen patterns rather than moving freely in the material.

At the University of Chicago Pritzker School of Molecular Engineering (UChicago PME), a research team discovered that one such material, Fe5GeTe2, exhibits a charge-ordered state in which electrons move collectively and unusually slowly while remaining quantum coherent.

The research, published in Science Advances and conducted in the lab of assistant professor Shuolong Yang, rewrites current knowledge of the material and could unlock new technological applications.

"This is a fundamental discovery that deviates from theoretical predictions," Yang said. "We now have to go back and think about the magnetic interactions of this material from scratch, but it also leads to new possibilities in using this material for new kinds of memory devices."

A many-body phenomenon
Discovered seven years ago, Fe5GeTe2 is part of a class of materials known as van der Waals magnets. Their atomically thin layers could enable new kinds of memory technologies, with advantages over those based on conventional magnetic materials.

A NASA-led team is hoping to take a short cut in directly imaging extrasolar earthlike planets. The idea is to use what they term a hybrid telescope composed of an orbital starshade coupled with the next generation of extremely large ground-based optical telescopes.

Positioned some 175,000km away in an elliptical Earth orbit, the starshade would likely first be used in conjunction with the European Southern Observatory's Extremely Large Telescope, now scheduled to see scientific first light in northern Chile by late 2030.

The starshade would block out the parent star's light to enable directly imaged optical observations of nearby solar systems and rocky earth-sized worlds. That is, planetary systems all located within some twenty light years of Earth.

Dubbed the Hybrid Observatory for Earth-like Exoplanets (HOEE), the team has already applied for Phase B NIAC (NASA Innovative Advanced Concepts) funding that they hope to see granted by 2027.

Current direct imaging space instruments, such as NASA’s James Webb Space Telescope’s Near Infrared Camera and NASA’s planned Roman Space Telescope’s Coronagraph Instrument, are unable to directly observe earth-like exoplanets, write the authors of a 2026 paper appearing in the journal Nature Astronomy.

For a long time, so-called satellite DNA was considered largely worthless. Now, ETH Zurich researchers have shown in fruit flies that these repetitive sections of genetic material act as a kind of barcode, enabling the correct chromosomes to recognize one another.

The body cells of humans and animals contain a double set of chromosomes. One-half of the genetic material comes from the mother; the other stems from the father.

During the formation of sperm or egg cells, this double set of chromosomes must be halved to form a single set. This takes place during what is known as meiosis. In this process, a cell with a double set of chromosomes gives rise to daughter cells with a single set of chromosomes. This halving is necessary because, during fertilization, two germ cells—and thus their genetic material—fuse together.

Afterwards, there is once again a double set of chromosomes. If this did not happen, the number of chromosomes would double from one generation to the next as the germ cells fuse.

To ensure that chromosomes can be distributed evenly during meiosis, the maternal and paternal versions of the same chromosome must locate one another within a cell and temporarily pair up. This is no easy task amid the vast jumble of the cell nucleus. Mismatches must be avoided at all costs during the pairing phase to prevent chromosomes from being distributed incorrectly.

But how do the matching chromosome pairs actually find each other? ETH researchers led by Madhav Jagannathan, a professor at the Department of Biochemistry, and his Ph.D. student Lena Skrutl have now investigated—using the example of egg cell formation in female fruit flies (Drosophila)—how this "matchmaking" process takes place in the cell nucleus and have made a surprising discovery.

It happened again! Here's Phobos passing in front of the Sun as seen from the surface of Mars by the Perseverance rover on Aug. 12. science.nasa.gov/photojournal/p…

Why do we remember certain moments while forgetting others? Every day, our minds are bombarded with thousands of pieces of information, yet our brains have a limited capacity to consolidate these experiences into lasting memories. This means we need a way to hold on to what is most important and disregard irrelevant details.

Now researchers at the University of York have uncovered the neural mechanism that instructs the brain which daily experiences to save as long-term memories during sleep.

In the study published in the journal PLOS Biology, researchers analyzed brainwave activity while participants were learning new associations between words and pictures. It is already known that after learning, sleep plays a central role in strengthening new learning into long-term memories.

Theta rhythms mark key memories
The researchers discovered that during learning, the brain "tags" certain memories for processing during later sleep. This tagging provides a way for the sleeping brain to strengthen only what is deemed important to remember.

This tagging was linked to a specific brainwave pattern known as a "theta rhythm" present during learning. The strength of these theta rhythms provided an instructional cue to help organize brain activity during sleep.

Those memories with a particularly strong theta rhythm during learning were preferentially strengthened during sleep, ultimately leading to better memory recall, the scientists said.

They say the findings show that the brain prioritizes and strengthens important information by using specific neural signals during learning that guide memory-strengthening processes during sleep. Determining what our brains decide to strengthen during sleep could help us better understand why some individuals are vulnerable to conditions such as depression.

When filtering becomes maladaptive
Lead author Dr. Dan Denis, from the Department of Psychology at the University of York, said, "It has long been known that a good night's sleep is important for processing memories.

"How the brain decides what is important to remember and what can be forgotten is still a mystery. These new findings help to answer that question by uncovering for the first time a signature of neural activity that instructs the brain which experiences to process during sleep and form into long-term memories."

"If we remembered everything that we experienced, our brains would quickly reach system overload. By selectively prioritizing important events in our lives, be they emotionally salient or important for the future, we are able to use our past experiences to help guide our interactions in the world."

The researchers note that while selective memory consolidation is vital for daily functioning, disruptions to this filtering process could play a role in psychiatric conditions.

NASA's next-generation cosmic observatory — the Nancy Grace Roman Space Telescope — is set to launch this weekend on a potentially decade-long mission to hunt for distant exoplanets, map out the Milky Way and unravel some of the universe's biggest mysteries, such as the true nature of dark matter.

The Roman telescope and its payload are scheduled to lift off aboard a SpaceX Falcon Heavy rocket Sunday (Aug. 30) at 7:26 a.m. EDT from Launch

Complex 39A at NASA's Kennedy Space Center in Florida, according to NASA. A few hours later, the telescope will be released from the rocket's second stage and be exposed to the vacuum of space for the first time.

But even if everything goes according to plan, we will still have to wait a while to get a glimpse of Roman's first groundbreaking images.

As of Thursday morning (Aug. 27), the National Weather Service forecasts a 50% chance of thunderstorms at Kennedy Space Center Sunday, so weather could delay the launch. If that happens, a second launch window is available Monday (Aug. 31) at 7:22 a.m. EDT, when there is a reduced risk of thunderstorms.

You can watch the event on a SpaceX livestream and stay up to date with last-minute developments and postlaunch news via a live blog from Live Science's sister site Space.com.
...
NASA officials have tentatively predicted that the telescope will start snapping the stars in early 2027. It took JWST about seven months postlaunch to send its first full-color image to Earth.

Finding one Higgs boson was hard enough, but physicists at the Large Hadron Collider (LHC) are hunting for something even more elusive: pairs of Higgs bosons. At the recent International Conference on High Energy Physics 2026, the ATLAS and CMS collaborations presented new constraints on the double-Higgs production rate, providing insight into how the Higgs boson interacts with itself.

The discovery of the Higgs boson in 2012 at the LHC marked the beginning of a new era in particle physics. Since then, researchers at the LHC have investigated and measured how the Higgs boson interacts with other particles, an important mechanism by which these particles get their mass.

However, physicists have yet to observe the Higgs boson interact with itself. Understanding this process would not only test the limits of the Standard Model, our current best working theory for particle physics, but also help shed light on whether the vacuum of our universe is stable.

The main challenge in studies of double-Higgs production is that the process is incredibly rare. The exact production rate has yet to be determined, but based on predictions from the Standard Model, researchers expect that for approximately every 1,500 single Higgs bosons produced in LHC collisions, only one pair of Higgs bosons will be produced.

To look for double-Higgs production, the ATLAS and CMS collaborations search for signs of the twin Higgs bosons decaying into other particles. Both collaborations investigated a particular decay channel in which one of the twin Higgs bosons decays into a bottom quark and antiquark, and the other into a tau particle and its antiparticle. This is one of the best ways to study double-Higgs production...

Q-CTRL has reported the first open-water field demonstration of a quantum gravimetric navigation system capable of operating without global navigation satellite system (GNSS) signals. The company's Ironstone Opal platform was deployed aboard a 29-meter vessel in the Coral Sea, off the eastern coast of Australia, where it maintained bounded position accuracy within 1 nautical mile over an 83-kilometer trajectory. This performance, achieved without access to satellite navigation, represents a more than tenfold improvement over standard navigation-grade inertial backup systems under similar conditions.

Hybrid Quantum-Classical Sensing
The Ironstone Opal system integrates a hybrid sensor architecture, combining cold-atom interferometry with classical accelerometers. Cold-atom gravimeters exploit the quantum interference of ultracold atoms to measure local gravitational acceleration with high precision, while classical accelerometers provide continuous inertial data. In this field trial, the quantum sensor was installed in an unconditioned passenger cabin, operating without active environmental temperature control or heavy motion-stabilization platforms. The system's AI-driven software stabilized the quantum measurements under open-ocean wave conditions, enabling both gimbaled and rigid strapdown deployments to achieve comparable results.

Planets form from disks of gas and dust around young stars, but the gas needed for growth does not last forever. New observations from NASA's James Webb Space Telescope (JWST) now offer a new view of how this gas escapes and how the process changes as young planetary systems develop.

The research, led by Naman Bajaj from the University of Arizona and co-authored by SETI Institute scientist Uma Gorti, looked at 72 young, sun–like stars and their protoplanetary disks. This is one of the largest planet-formation studies using JWST. The findings show that different types of winds are more important at different stages in a planetary system's early life.

A race against time
"What is exciting about this study is that we can now see, across a large sample of young systems, how the mechanisms that remove gas from planet-forming disks change with time. Disk dispersal sets a fundamental clock for planet formation: once the gas is gone, the opportunity to build gas-rich planets is essentially over," said Gorti.

Today, our solar system is about 4.5 billion years old and is mostly empty space. In its first few million years, though, the sun was surrounded by a thick protoplanetary disk with about 100 times more gas than dust. Most of that gas eventually vanished.

Figuring out how and when this happens is important because the gas in these disks is the main ingredient for giant planets like Jupiter and Saturn. If the gas goes away too soon, these planets might not have enough time to build up their thick atmospheres.

Roman is ready for liftoff tomorrow at 7:26 a.m. EDT!

The SpaceX Falcon Heavy rocket carrying NASA's Nancy Grace Roman Space Telescope stands vertical at Launch Complex 39A at Kennedy Space Center in Florida.

Ready for launch, ready to change how we view the cosmos.

After years of careful planning, testing, and construction, NASARoman is ready for liftoff tomorrow, Aug. 30.

Two mirrors, each small enough to fit in the palm of your hand, are set to change how we understand the cosmos.

The launch window for NASA’s Nancy Grace Roman Space Telescope is about to open. The telescope’s primary instrument is due to inform practically every area of astrophysics, but also tucked inside the telescope is a specialized coronagraph, an experimental apparatus that will attempt to directly capture starlight reflected off a planet’s surface for the first time. It’s an ambitious project that NASA hopes will pave the way for a space telescope that can one day provide a glimpse of an Earthlike planet orbiting a sunlike star.

That’s a feat far beyond the power of current engineering—hence taking the basic technique for a spin on Roman. “We’ll test them in space for the first time, and we’ll understand what work still is left to go,” says Vanessa Bailey, an astrophysicist at NASA’s Jet Propulsion Laboratory and instrument scientist for the coronagraph.

At its core, a coronagraph is just a science-minded sunshade that can block out the light of a bright star and reveal a fainter object otherwise lost in the glare. Such instruments have flown in space before—both the Hubble and James Webb space telescopes carry coronagraphs. But Roman’s is light-years more sophisticated than its predecessors, thanks largely to technology called adaptive optics, which deforms a telescope’s mirror to cancel out these light distortions. It still faces a monumental challenge, though: NASA compares the task of Roman’s coronagraph to photographing a firefly perched next to a floodlight—from across the country.