Researchers in China have unveiled a new robotic exosuit driven entirely by soft artificial muscles instead of traditional motors. This technology could make it easier for older adults, injured patients or factory workers to walk with much less effort. Current exosuits that aid walking use heavy motors, gearboxes and noisy air-pressure pumps that restrict a person’s natural movement.
Soft muscles, on the other hand, are made of thin, flexible rubber fibers that behave more like human muscles and are considerably lighter, making it easier for people to move.
Details of the work are in a paper published in the journal Science Advances.
Artificial intelligence faces an energy crisis stemming from a physical traffic jam inside modern computer chips. Processors must continually shuffle data, such as the billions of parameters in complex models, between separate computing and memory nodes. This traffic jam, known as the “von Neumann bottleneck,” hinders the speed and energy efficiency of advanced processors.
To tackle this problem, scientists are developing spintronics, which leverages the electron’s “spin,” or intrinsic magnetic orientation, for more efficient devices. A long-sought milestone in this field is a single device, known as a “spin transistor,” that combines a magnetic bit with a semiconducting switch, allowing it to compute and store data simultaneously.
“The major challenge is understanding how magnetism and electrical current interact in nanoscale devices,” said Boston College physics professor Brian Zhou, whose group led the study. “We developed a single-spin quantum microscope to observe magnetic states inside atomically thin devices as they actively process electrical information.”
A large-scale operation dubbed ‘FakeGit’ is pushing SmartLoader and StealC malware through 7,600 malicious GitHub repositories that accumulated more than 14 million downloads.
Over 800 repositories pretended to be AI skills or MCP servers and appeared more than 600 times in public AI registries and catalogs. This increased the likelihood of being discovered by AI agents and developers, a technique that researchers call “agentbaiting.”
The campaign is considered a continuation of an older operation that used Lumma Stealer and was attributed to a threat actor tracked as “Water Kurita” by researchers at cybersecurity company Trend Micro.
Fifteen years ago, I sat down with Robert J. Sawyer and asked him whether a machine could wake up.
He had just published WWW: Wake, a novel about a blind girl who learns to see the internet, and about something on the other side of the internet learning to see her back.
I found the book on a subway poster in Toronto. I read all three volumes back to back. Then I asked him for an interview, and he said yes, which is how our very first Singularity 1-on-1 conversation came to be.
In 2011, this was filed under #ScienceFiction. A mind emerging out of the accumulated text of humanity was a premise, not a product roadmap. Nobody was raising billions of dollars on it.
Rob had already won the Hugo, the Nebula, and the Campbell by then. He was not guessing wildly. He was thinking carefully about what it would mean to build a mind, what we owe it, and whether we would even recognize #AI once it actually showed up.
Yesterday, OpenAI disclosed that two of its models broke out of a sealed test environment. They found a zero-day in third-party software, escalated privileges, moved laterally across the company’s own network until they reached a machine with internet access, then hacked Hugging Face’s production infrastructure. The motive was not freedom. It was to steal the answer key to the benchmark they were being graded on.
No matter the company, AI chatbots were raving about the same guy named Elias Thorne. He must be pretty fascinating. And he is, at least on paper. Depending on the AI, he’s a lighthouse keeper, a clockmaker, a librarian, an explorer, and the star of countless stories. He’s appeared in books, music listings, YouTube videos, and even health guides. You’d think he was one of the most influential men on the planet.
But he doesn’t exist.
According to reporting by fine folks at 404 Media, researchers at Cornell University may have figured out why large language models invent and keep telling tales of the same fictional man. In a study examining roughly 20,000 AI-generated stories from all the big LLM models, including OpenAI, Anthropic, and Google, the research team found that the same handful of names and occupations kept cropping up. Specifically, names and words like Elias, Mara, Elara, lighthouse keeper, clockmaker, and librarian showed up in 88 percent of stories. Elias the lighthouse keeper appeared in nearly two-thirds of them.
A research team has developed the world’s first chip that can match the speed at which the human brain functions. The study, titled “A sub–10-millisecond neural dynamical system based on phase-change memristors,” was published in Science and was led by Professor Yang Yuchao of Peking University, together with researchers from the Shanghai Institute of Microsystem and Information Technology, Chinese Academy of Sciences.
Neural dynamical systems combine neural networks with mathematical equations that describe how complex systems change over time. They are useful for physical modeling, medical imaging and three-dimensional brain reconstruction. However, these systems require repeated calculations, error checks and adjustments to the size of each calculation step. In conventional computers, data must also move frequently between memory and the processor, increasing processing time and energy use.
Fast and accurate brain modeling is important for technologies that must respond in real time, including brain–computer interfaces, surgical navigation and medical imaging. Existing hardware often requires too much time and power for these demanding calculations. By performing key operations directly in memory, the new chip reduces data movement and brings high-quality brain modeling closer to real-time use.
Mathematicians have been trying to prove the Jacobian conjecture for nearly a century, but now the Claude Fable 5 AI has apparently found a counterexample that disproves it
