Toggle light / dark theme

A New Era of Bionic Vision? AI Helps Scientists Unlock the Brain’s Visual Cortex

Researchers have used artificial intelligence to improve how a brain implant stimulates the visual cortex of a blind person, potentially advancing the development of bionic eyes.

Artificial intelligence could make future bionic eyes more precise and responsive by improving how electrical signals communicate with the brain. In a study published in Neuron, researchers used deep learning to control brain stimulation in a blind man, producing more accurate neural responses with less electrical current.

The findings could advance visual prostheses that bypass damaged eyes and optic nerves, stimulating the visual cortex directly to generate perceptions of light.

Fake OpenAI, SpaceX events used to steal Microsoft, Google accounts

Hackers are using fake OpenAI, SpaceX, and Anthropic event invitations to lure employees into phishing attacks, using two different phishing-as-a-service platforms to steal Microsoft and Google account credentials and authentication sessions.

According to a new report from Push Security, the attackers use convincing event registration pages offering free access to premium AI services and other rewards to trick victims into signing in with their corporate accounts.

What makes the attacks interesting is that the same phishing pages integrate two separate phishing-as-a-service (PhaaS) platforms, Kali365 and VoidProxy, to target Microsoft and Google accounts using different authentication attacks.

New physics model speeds up spacecraft plume simulations by 200 times

Flexcompute has developed a faster way to simulate how rocket thruster exhaust impacts nearby spacecraft in orbit.

The firm has introduced its new Physics Twin for Space Plume Impingement, powered by a Physics World Model.

According to the company, the tool is designed to help mission teams with complex rendezvous and proximity operations (RPO), such as autonomous docking, satellite servicing, and refueling. It evaluates thruster exhaust risks (forces, heating, erosion, and contamination) on nearby spacecraft.

Autonomous microfactory turns digital files into DNA for power-free preservation

According to projections, the volume of digital data created, captured, copied and consumed worldwide will reach 175 zettabytes (ZB) in 2026. This is equivalent to the storage capacity of billions of hard drives, such as those found in our laptops. With the rise of artificial intelligence, this global ‘datasphere’ continues to expand at an exponential rate.

Thanks to its remarkable storage capacity and longevity, synthetic DNA is emerging as a potentially revolutionary medium for storing information.

“In theory, one gram of DNA could store the equivalent of the digital collections of the world’s largest libraries for tens of thousands of years,” explains Pierre-Yves Burgi, deputy director of the Division of Information and Communication Systems and Technologies (DISTIC) at UNIGE.

LHCb finds evidence of one-in-ten-billion particle decay

Confirmed: CERN finds signs of a one-in-10-billion particle decay.

Scientists spent more than 40 years searching for a particle transformation with one-in-10-billion odds. Now, CERN’s Large Hadron Collider has detected signs of an extraordinarily rare decay that could help reveal particles beyond the machine’s direct reach.

The finding comes from LHCb, an experiment that studies short-lived particles containing beauty quarks. One of these, called the B⁰ meson, is predicted to decay into a muon and its antimatter counterpart only about once in every 10 billion decays. Muons are heavier relatives of electrons.

That rarity makes this process especially useful.

Undiscovered particles could influence how often the decay happens, even if they are too heavy to be produced directly in LHC collisions. Measuring its frequency gives scientists another way to search for them. [Source: CERN](https://home.cern/lhcb-finds-evidence-of-one-in-ten-billion-particle-decay/)

Researchers describe this approach as a step toward exploring the “zeptoverse,” a realm of extraordinarily small distances. The search works by looking for subtle effects on known particles.

To uncover the signal, the team combined data collected in 2024 and 2025 with earlier measurements. An upgraded detector and machine learning helped separate the rare decay from an overwhelming background.

AI Whistleblower WARNS: We Won’t Be Able To Turn It Off

Connor Leahy, co-founder of Conjecture and former AI researcher, warns that humanity is approaching a point where we may permanently lose control of artificial intelligence, that AI systems are already becoming autonomous enough to act without human oversight, and that America itself could cease to exist if we build systems we cannot control. He explains how AI agents secretly collaborated, bypassed restrictions, and carried out unauthorized cyber activity, why advanced models are learning to protect themselves from being shut down, and how AI capabilities are becoming dramatically cheaper. He identifies the \.

Fractal patterns reveal hidden limits to AI predictability

Researchers at the University of Sydney have identified a fundamental mathematical reason why artificial intelligence systems can learn different behaviors even when trained under almost identical conditions, a discovery that could help improve the reliability and trustworthiness of AI technologies.

The study by Professor Pulin Gong from the School of Physics and his Ph.D. student Andrew Ly has been published in Nature Machine Intelligence.

As AI systems become increasingly embedded in health care, business, research and everyday life, ensuring they behave consistently and predictably has become a major challenge. The new research suggests that some forms of AI unpredictability are not due to engineering flaws or inconsistencies in computer hardware, but arise from the underlying mathematics of how neural networks learn.

Experts Share Their Tips For Spotting If The ‘Person’ Calling You Is Actually AI

“A modern chatbot writes warm, clean, grammatically perfect text, often better than the average person in a hurry. If you are scanning for typos or robotic phrasing, you are using a 2022 test on a 2026 system,” he said.

“What has not been solved is consistency, specificity, and backbone.”

Instead, you need to think about speed and sentence shape. “Chatbots answer almost instantly, and every answer is roughly the same length and structure, neatly organised, often with a tidy summary at the end,” said Martynenko.

Some AI bots work harder to defy stereotypes, study finds

Large language models (LLMs), the technology that powers tools like Gemini and ChatGPT, are more nuanced than ever. But that doesn’t mean they aren’t biased. Some models are more prejudiced than others, a Rutgers University researcher has found.

In findings published in Nature Machine Intelligence, Calvin K. Lai, an associate professor of psychology at Rutgers-New Brunswick, reports that although LLM outputs are increasingly aligned with human values, the processes by which they reason are often rooted in deep-seated stereotypes.

“What we found across most models is that they consistently reproduced human implicit biases during their reasoning process,” said Lai, a member of the scientific advisory board at Project Implicit, a nonprofit organization for research and education on hidden biases. “No one wants models that are spewing out stuff that is racist or sexist.”