Lifeboat Foundation: Safeguarding Humanity
Toggle light / dark theme

Blog

Category: robotics/AI – Page 41

The Strange Case of Elias Thorne, the Imaginary Man AI Chatbots Are Obsessed With

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.

Tiny memristor chip cuts brain modeling time to under 10 milliseconds

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.

From molecules to networks, siibra integrates brain data into a unified atlas

In the current issue of the journal Nature Methods, siibra is introduced as a software suite that integrates data from different multimodal sources into a comprehensive atlas of the human brain and makes the data easily accessible—for interactive exploration and automated, reproducible data analyses, simulations and AI applications. siibra is developed by an international team of scientists led by the Institute of Neuroscience and Medicine (INM-1) at Forschungszentrum Jülich.

To better understand the human brain, information from various levels must be integrated, from molecules and cells to their organization and entire networks. A central challenge is that these data are often scattered across sources and organized differently. They originate from methods such as microscopy, MRI and connectivity analysis; exist in formats ranging from images to tables; and rely on different spatial reference systems and conceptual taxonomies.

“Using siibra, we are now able to access and analyze brain data in a structured way from micro-to macrolevels—for more precise neuroscience studies, bio-inspired AI and clinical applications such as deep brain stimulation,” says Dr. Timo Dickscheid, working group leader for “Big Data Analytics.”

AI agent helps prepare synchrotron X-ray experimental measurements, paving the way for autonomous operation

Artificial intelligence (AI) models are now used daily by many people worldwide, both for professional and personal purposes. Over the past decades, scientists specialized in various disciplines have also started using these models to conduct research or simplify their experimental practices.

Researchers at Stanford University and SLAC National Accelerator Laboratory recently explored the possibility of using an AI-powered agent to prepare a synchrotron-based X-ray experiment. Synchrotrons are large research facilities at which electrons are accelerated to produce very bright X-rays, which can then be used to study the atomic structure of materials, molecules and biological samples.

In a paper published in Nature Machine Intelligence, the team at Stanford and SLAC proposed using an AI-based agent to prepare a real synchrotron X-ray experiment. They showed that this agent could autonomously plan actions, interpret observations and generate instrument-control commands to complete sample alignment.

/* */