Gut Bacterium Linked to Age-Related Memory Decline in Mice: A new Nature study identifies a gut bacterium, Parabacteroides goldsteinii, as a potential driver of cognitive decline in aging mice. Researchers found that the microbe proliferates in older animals and disrupts signalling along the vagus nerve, a key communication pathway between the gut and brain. In experiments, young mice exposed to the microbiomes of older mice developed memory problems resembling those of aged animals. Further tests showed that transplanting P. goldsteinii worsened memory, while eliminating it with antibiotics or phage therapy improved cognition in old mice. The bacterium appears to act by producing medium-chain fatty acids that activate immune cells, triggering inflammation that dampens gut-brain communication. Although the work was done in mice, the identified circuit may exist in humans, raising hopes for future diet- or microbiome-based therapies. (Nature)

Are Sharks Really a Natural Group? New Genomic Study Complicates the Family Tree: A new genomic preprint suggests sharks might not form a single natural evolutionary group after all. Researchers analyzing genomes from dozens of sharks, rays, skates, and relatives found that one unusual shark lineage, the Hexanchiformes—such as cow sharks and frilled sharks—may have split off separately from the branch that later produced all other sharks, rays, and skates. If true, that would mean many animals we call sharks are actually more closely related to rays and skates than to these primitive-looking sharks, making “sharks” a paraphyletic group. The result remains unreviewed and conflicts with other genetic analyses, so the question is far from settled. Still, the debate matters: a clearer shark family tree could reshape ideas about jaw evolution, body-plan origins, and conservation priorities for some of Earth’s most ancient vertebrates. (Nature)

Alaska’s Ancient Ice Time Capsule May Hold a Lost Arctic Climate Record: At Cape Blossom, Alaska, researchers believe they may have found a rare remnant of glacial ice more than 350,000 years old—potentially the oldest known ice in the Northern Hemisphere. If confirmed, the buried ice could offer a unique record of Arctic climate during the Pleistocene, especially after the unusually warm Marine Isotope Stage 11 and into the glacial advance of MIS-10. Unlike the clean layered ice cores of Antarctica and Greenland, this Alaskan ice is messy, sediment-laced, and difficult to interpret, with microbial contamination obscuring some of its climate signals. Still, even dating the ice would sharpen scientists’ understanding of northern ice-sheet history and past sea levels. The urgency is clear: accelerated coastal erosion and warming seas are rapidly destroying the bluffs, threatening to erase this extraordinary climate archive before its secrets can be recovered. (Science)

AI Autocomplete Can Quietly Nudge Belief: A new study highlighted by Scientific American argues that AI autocomplete tools do more than save keystrokes: they can subtly shape what users come to believe. Researchers found that when writing assistants repeatedly suggested slanted or biased wording, people did not merely accept phrasing shortcuts; their judgments drifted in the direction of the machine’s prompts. That makes predictive text a cognitive influence technology, not just a convenience layer. The implications extend well beyond email polish. If the same mechanism holds up across platforms, autocomplete could become a quiet behavioral lever in workplaces, classrooms, and political communication. The broader consciousness-and-cognition angle is clear: external tools are beginning to participate in the formation of thought itself, not just its expression. (Scientific American)

Gut Microbes Join the Brain-Aging Story: Another Scientific American report spotlights mouse research suggesting that the gut microbiome may play a direct role in brain aging and memory decline. The study points to a gut-brain communication pathway that helps determine how well aging brains retain memory function. Instead of treating cognitive decline as a purely brain-local event, the work frames it as a systems problem involving microbes, nerves, and immune signaling. That matters because it widens the intervention space: future treatments might not begin in the skull at all. Though it remains a mouse study and translation to humans is far from guaranteed, it fits a fast-growing pattern in neuroscience, where cognition is increasingly understood as emerging from distributed bodily networks rather than the brain in isolation. (Scientific American)

SuperAgers May Keep More “Young” Cells in Memory Circuits: Science News covered new work on so-called SuperAgers, older adults whose memory remains unusually strong. Researchers examined the hippocampus, a brain region crucial for memory, and found genetic signatures consistent with more immature cells in these exceptional agers than in people with Alzheimer’s disease. The finding feeds a longstanding and still-heated debate over whether adult human brains continue generating new neurons. Even skeptics quoted in the coverage acknowledge that the dataset is provocative, though not definitive. What makes the study important is not that it “solves” aging, but that it suggests successful cognitive aging may involve distinct biological programs rather than simply a slower version of ordinary decline. In other words, some brains may stay sharp by aging differently, not merely by aging less. (Science News)

Lucid Dream Sound Cues May Boost Next-Day Problem Solving: A Science News story this week reported that lucid dreamers who heard puzzle-linked soundtracks during sleep were more likely to solve previously unsolved problems the next day. The study, published in Neuroscience of Consciousness, suggests that carefully targeted sensory cues can steer dream content in ways that affect later cognition. This is not mind-control sleep learning in the old sci-fi sense; it is subtler and narrower. But it does hint that conscious and unconscious processing are more continuous than everyday intuition suggests. Dreams here look less like narrative debris and more like a workspace where unresolved material can be reorganized. For consciousness research, that matters because it treats sleeping awareness not as an on/off absence of mind, but as an altered mode of cognition with measurable downstream effects. (Science News)

Why the Brain Usually Keeps Hallucinations at Bay: Science highlighted a fresh study on how the mind protects itself from hallucinations by constantly checking expectations against reality. The idea is rooted in predictive processing: perception is not passive intake but active inference. The new work suggests healthy cognition depends on the brain’s willingness to revise its own prior beliefs when the world disagrees. When that balancing act weakens, perception can drift away from external evidence. This is one of the week’s most conceptually important stories because it reaches across psychiatry, consciousness science, and AI. It asks what makes experience reliable in the first place. Rather than treating hallucinations as bizarre add-ons, the study reframes them as failures of a system that all brains use continuously to construct conscious experience from uncertain signals. (Science)

A Gut Microbe May Drive Memory Loss in Aging Mice: A separate Science news report described a “tour de force” mouse study showing that one gut microbe can promote memory loss. The work ties microbial changes to brain regions involved in learning and memory, adding sharper mechanistic detail to the broader gut-brain story. That distinction matters: instead of saying the microbiome is vaguely “associated” with cognition, the study points to a candidate organism and a more direct causal chain. Even so, the report also stresses an important caution—what works in mice may not map neatly onto human aging. Still, this is a major story because it helps move gut-brain research from correlation-heavy speculation toward experimentally tractable biology. If replicated and extended, it could reshape how neuroscientists think about the boundaries of the memory system. (Science)

Forgotten Memories May Still Be Reactivating Under the Hood: A University of Nottingham release on EurekAlert summarized a Journal of Neuroscience study using magnetoencephalography to show that memories can be reactivated in the brain even when people cannot consciously recall them. Participants learned word-video pairs, and a machine-learning system detected the neural signature of specific memories during attempted recall. The striking result was that reactivation still occurred when overt remembering failed. Successful recall, however, seemed to depend on those memory signals pulsing rhythmically enough to stand out from background neural activity. This is a useful update for consciousness research because it separates storage from awareness. A memory can be active without being consciously available. That makes remembering less like pulling a file from a cabinet and more like getting a signal strong enough to cross the threshold into experience. (EurekAlert!)

Social Cognition Looks More Dynamic Than Psychologists Thought: Another EurekAlert release reported a study showing that mentalization—the process of inferring what other people think or feel—is not a static ability but an ongoing adaptive process during interaction. Earlier research often relied on fixed scenarios or one-off judgments. This study instead used more dynamic, everyday-like exchanges and found that brain activity tracked how well people continuously updated their models of other minds. That is important because social cognition sits at the center of many debates about consciousness, selfhood, empathy, and psychiatric difference. If mind-reading is fluid rather than trait-like, then social intelligence may depend less on stable “capacity” and more on real-time adjustment. The work could eventually help researchers measure social-cognitive function more objectively in autism, schizophrenia, or other conditions that affect interpersonal understanding. (EurekAlert!)

Stanford Links Age-Related Memory Decline to Gut–Vagus Signaling: Stanford Medicine reported that researchers reversed age-related cognitive decline in mice by restoring gut-brain communication along the vagus nerve. In the study, aging altered gut bacteria, immune responses in the gastrointestinal tract interfered with vagal signaling, and that disruption hurt hippocampal memory formation. Boosting the pathway helped older mice perform like younger ones on memory tasks. The result is notable because it connects three areas often studied separately: microbiome composition, peripheral immune signaling, and neural circuits of memory. It also suggests that age-related cognitive decline may be more reversible than standard narratives imply, at least in animals. Human translation remains the big question, but the finding strengthens a broader shift in cognitive science away from brain-only explanations and toward embodied, multi-organ models of thought and memory. (Stanford Medicine)

One Workout May Trigger Human Memory Ripples: The University of Iowa reported direct human evidence that a single bout of exercise can enhance hippocampal “ripples,” high-frequency brain waves linked to learning and recall. Researchers recorded intracranial brain activity in epilepsy patients before and after 20 minutes of cycling and found an increase in ripple activity connecting the hippocampus to cortical regions involved in memory. Exercise has long been associated with better cognition, but much of that evidence relied on behavioral performance or indirect imaging. This study is valuable because it gets much closer to the neural mechanism in action. It does not prove that one workout makes people smarter on demand, but it strengthens the case that even brief physical movement can rapidly alter the brain states underlying memory. That makes exercise look less like long-term maintenance and more like immediate cognitive modulation. (Iowa Now)