Computers and Math from Science Daily Feed

Researchers have pushed quantum chip design into a new era by simulating every physical detail before fabrication. Using a supercomputer with nearly 7,000 GPUs, they modeled how signals travel and interact inside an ultra-tiny chip. Unlike earlier “black box” approaches, this method captures real materials, layouts, and qubit behavior. The result is a powerful new way to spot problems early and build better quantum hardware faster.

A new study put ChatGPT to the test by asking it to judge whether hundreds of scientific hypotheses were true or false—and the results were far from reassuring. While the AI got it right about 80% of the time on the surface, its performance dropped significantly when accounting for random guessing, revealing only modest reasoning ability. Even more concerning, it frequently contradicted itself when asked the exact same question multiple times, sometimes flipping answers back and forth.

Artificial intelligence is often portrayed as a tool that replaces human work, but new research from Swansea University suggests a far more exciting role: creative collaborator. In a large study with more than 800 participants designing virtual cars, researchers found that AI-generated design galleries sparked deeper engagement, longer exploration, and better results.

A new AI framework called THOR is transforming how scientists calculate the behavior of atoms inside materials. Instead of relying on slow simulations that take weeks of supercomputer time, the system uses tensor network mathematics and machine-learning models to solve the problem directly. The approach can compute key thermodynamic properties hundreds of times faster while preserving accuracy. Researchers say this could accelerate discoveries in materials science, physics, and chemistry.

As AI systems began acing traditional tests, researchers realized those benchmarks were no longer tough enough. In response, nearly 1,000 experts created Humanity’s Last Exam, a massive 2,500-question challenge covering highly specialized topics across many fields. The exam was engineered so that any question solvable by current AI models was removed. Early results show even the most advanced systems still struggle — revealing a surprisingly large gap between AI performance and true expert-level knowledge.

Researchers at Cornell University have developed a powerful imaging technique that reveals atomic scale defects inside computer chips for the first time. Using an advanced electron microscopy method, the team mapped the exact positions of atoms inside tiny transistor structures and uncovered small imperfections nicknamed “mouse bites.” These defects form during the complex manufacturing process and can disrupt how electrons flow through a chip’s channels, which are only about 15 to 18 atoms wide.

Researchers have built the smallest OLED pixel ever made—just 300 nanometers across—without sacrificing brightness. By redesigning the pixel with a nano-sized optical antenna and a protective insulation layer, they prevented the short circuits that normally plague devices at this scale. The result is a stable, ultra-tiny light source that could allow full HD displays to fit on an area the size of a grain of sand.

Researchers at Kobe University have developed an AI system that can detect acromegaly, a rare hormone disorder, by analyzing photos of the back of the hand and a clenched fist. The disease often develops slowly and can take years to diagnose, even though untreated cases may shorten life expectancy.

Choosing the right method for multimodal AI—systems that combine text, images, and more—has long been trial and error. Emory physicists created a unifying mathematical framework that shows many AI techniques rely on the same core idea: compress data while preserving what’s most predictive. Their “control knob” approach helps researchers design better algorithms, use less data, and avoid wasted computing power. The team believes it could pave the way for more accurate, efficient, and environmentally friendly AI.

Scientists at the University of Tokyo have captured something never seen before: a frame-by-frame view of how electron spins flip inside an antiferromagnet, a material once thought to be magnetically “invisible.” By firing ultrafast electrical pulses into a thin layer of manganese–tin and tracking the response with precisely timed flashes of light, the team uncovered two distinct switching mechanisms. One relies on heat generated by strong currents, while the other flips spins directly with minimal heating — a far more efficient process.

Researchers at the University of Basel and the ETH in Zurich have succeeded in changing the polarity of a special ferromagnet using a laser beam. In the future, this method could be used to create adaptable electronic circuits with light.

As millions turn to ChatGPT and other AI chatbots for therapy-style advice, new research from Brown University raises a serious red flag: even when instructed to act like trained therapists, these systems routinely break core ethical standards of mental health care. In side-by-side evaluations with peer counselors and licensed psychologists, researchers uncovered 15 distinct ethical risks — from mishandling crisis situations and reinforcing harmful beliefs to showing biased responses and offering “deceptive empathy” that mimics care without real understanding.

Twisting atomically thin magnetic layers does more than reshape their electronics—it can create giant, topological magnetic textures. In chromium triiodide, researchers observed skyrmion-like patterns stretching far beyond the expected moiré scale, reaching hundreds of nanometers. Even more surprising, their size doesn’t simply follow the twist pattern but peaks at a specific angle. This twist-controlled magnetism could pave the way for low-power spintronic devices built from geometry alone.

Scientists have pulled off a feat long considered out of reach: getting light to mimic the famous quantum Hall effect. In their experiment, photons drift sideways in perfectly defined, quantized steps—just like electrons do in powerful magnetic fields. Because these steps depend only on nature’s fundamental constants, they could become a new gold standard for ultra-precise measurements. The discovery also hints at tougher, more reliable quantum photonic technologies.

Researchers have discovered new ways to shape quantum light, creating high-dimensional states that can carry much more information per photon. Using advanced tools like on-chip photonics and ultrafast light structuring, they’re pushing quantum communication and imaging into exciting new territory. Although long-distance transmission remains tricky, innovative approaches—such as topological quantum states—could make these fragile signals far more resilient. The momentum suggests quantum optics is entering a bold new phase.

Quantum computers need special materials called topological superconductors—but they’ve been notoriously difficult to create. Researchers have now shown they can trigger this exotic state by subtly adjusting the mix of tellurium and selenium in ultra-thin films. That tiny chemical tweak changes how electrons interact, effectively turning a quantum phase “dial” until the ideal state appears. The result is a more practical path toward building stable, next-generation quantum devices.

CU Boulder researchers have designed microscopic “racetracks” that trap and amplify light with exceptional efficiency. By using smooth curves inspired by highway engineering, they reduced energy loss and kept light circulating longer inside the device. Fabricated with sub-nanometer precision, the resonators rank among the top performers made from chalcogenide glass. The technology could lead to compact sensors, microlasers, and advanced quantum systems.

Scientists may have spotted a long-sought triplet superconductor — a material that can transmit both electricity and electron spin with zero resistance. That ability could dramatically stabilize quantum computers while slashing their energy use. Early experiments suggest the alloy NbRe behaves unlike any conventional superconductor. If verified, it could become a cornerstone of next-generation quantum and spintronic technology.

Researchers tested whether generative AI could handle complex medical datasets as well as human experts. In some cases, the AI matched or outperformed teams that had spent months building prediction models. By generating usable analytical code from precise prompts, the systems dramatically reduced the time needed to process health data. The findings hint at a future where AI helps scientists move faster from data to discovery.

Qubits, the heart of quantum computers, can change performance in fractions of a second — but until now, scientists couldn’t see it happening. Researchers at NBI have built a real-time monitoring system that tracks these rapid fluctuations about 100 times faster than previous methods. Using fast FPGA-based control hardware, they can instantly identify when a qubit shifts from “good” to “bad.” The discovery opens a new path toward stabilizing and scaling future quantum processors.