Lifeboat Foundation: Safeguarding Humanity
Toggle light / dark theme

Blog

Category: robotics/AI – Page 19

Diamond quantum sensors improve spatial resolution of MRI

This accomplishment breaks the previous record of 48 qubits set by Jülich scientists in 2019 on Japan’s K computer. The new result highlights the extraordinary capabilities of JUPITER and provides a powerful testbed for exploring and validating quantum algorithms.

Simulating quantum computers is essential for advancing future quantum technologies. These simulations let researchers check experimental findings and experiment with new algorithmic approaches long before quantum hardware becomes advanced enough to run them directly. Key examples include the Variational Quantum Eigensolver (VQE), which can analyze molecules and materials, and the Quantum Approximate Optimization Algorithm (QAOA), used to improve decision-making in fields such as logistics, finance, and artificial intelligence.

Recreating a quantum computer on conventional systems is extremely demanding. As the number of qubits grows, the number of possible quantum states rises at an exponential rate. Each added qubit doubles the amount of computing power and memory required.

Although a typical laptop can still simulate around 30 qubits, reaching 50 qubits requires about 2 petabytes of memory, which is roughly two million gigabytes. ‘Only the world’s largest supercomputers currently offer that much,’ says Prof. Kristel Michielsen, Director at the Jülich Supercomputing Centre. ‘This use case illustrates how closely progress in high-performance computing and quantum research are intertwined today.’

The simulation replicates the intricate quantum physics of a real processor in full detail. Every operation – such as applying a quantum gate – affects more than 2 quadrillion complex numerical values, a ‘2’ with 15 zeros. These values must be synchronized across thousands of computing nodes in order to precisely replicate the functioning of a real quantum processor.

The JUPITER supercomputer set a new milestone by simulating 50 qubits. New memory and compression innovations made this breakthrough possible. A team from the Jülich Supercomputing Centre, working with NVIDIA specialists, has achieved a major milestone in quantum research. For the first time, they successfully simulated a universal quantum computer with 50 qubits, using JUPITER, Europe’s first exascale supercomputer, which began operation at Forschungszentrum Jülich in September.

Continue reading “Diamond quantum sensors improve spatial resolution of MRI” | >

Ilya Sutskever — We’re moving from the age of scaling to the age of research

00:00:00 – Explaining model jaggedness

00:09:39 — Emotions and value functions

00:18:49 – What are we scaling?

00:25:13 – Why humans generalize better than models

00:35:45 – Straight-shotting superintelligence

00:46:47 – SSI’s model will learn from deployment

Continue reading “Ilya Sutskever — We’re moving from the age of scaling to the age of research” | >

New deep-learning tool can tell if salmon is wild or farmed

A paper published in Biology Methods and Protocols, finds that it is now possible to distinguish wild from farmed salmon using deep learning, potentially greatly improving strategies for environmental protection. The paper is titled “Identifying escaped farmed salmon from fish scales using deep learning.”

Norway is home to the largest remaining wild populations of wild salmon and is also one of the largest producers of farmed salmon. Atlantic salmon abundance in Norway has declined by over 50% since the 1980s and is now at historically low levels. Escaped farmed salmon are an important reason for this decline.

Norway produces over 1.5 million metric tons of farmed Atlantic salmon annually. Each year, however, approximately 300,000 farmed salmon escape into the wild.

Adaptive method helps light-based quantum processors act more like neural networks

Machine learning models called convolutional neural networks (CNNs) power technologies like image recognition and language translation. A quantum counterpart—known as a quantum convolutional neural network (QCNN)—could process information more efficiently by using quantum states instead of classical bits.

Photons are fast, stable, and easy to manipulate on chips, making photonic systems a promising platform for QCNNs. However, photonic circuits typically behave linearly, limiting the flexible operations that neural networks need.

Soft robots harvest ambient heat for self-sustained motion

A warm hand is enough to drive motion in tiny Salmonella-inspired robots that harness molecular-level dynamic bonding.

A team of researchers from China and the U.S. came together to design soft robots with a coordination-motorized oscillator (CoMO) that can make self-sustained micromovements by harvesting small amounts of energy from sunlight or body heat. At the heart of this innovation is a new supramolecular polydimethylsiloxane (PDMS)-based elastic polymer dynamically crosslinked by Eu3+ at the center.

The findings are published in Angewandte Chemie.

New model measures how AI sycophancy affects chatbot accuracy and rationality

If you’ve spent any time with ChatGPT or another AI chatbot, you’ve probably noticed they are intensely, almost overbearingly, agreeable. They apologize, flatter and constantly change their “opinions” to fit yours.

It’s such common behavior that there’s even a term for it: AI sycophancy.

However, new research from Northeastern University reveals that AI sycophancy is not just a quirk of these systems; it can actually make large language models more error-prone. The research is published on the arXiv preprint server.

/* */