Lifeboat Foundation: Safeguarding Humanity
Toggle light / dark theme

Blog

Category: robotics/AI – Page 88

Daydreaming algorithm helps AI remember what matters

During the day, our brain acquires new memories; at night, during sleep, it consolidates the important ones and eliminates the useless ones. A similar principle has been applied to Hopfield networks, one of the classic models of artificial intelligence inspired by the workings of the brain. In 2025, Federico Ricci-Tersenghi and colleagues developed Daydreaming, an algorithm that combines the learning of new memories with the elimination of spurious ones, drastically improving the network’s capacity.

One limitation remained, however. These networks lose effectiveness when they work with real-world data, which are rarely perfectly balanced—for example, very bright or very dark images, in which white or black pixels overwhelmingly dominate. In a new study published in the Journal of Statistical Mechanics: Theory and Experiment (JSTAT), Ricci-Tersenghi and Japanese colleagues present a new version of the algorithm capable of effectively handling realistic, strongly biased data.

A “classical” neuralnetwork The networks proposed by John Hopfield in 1982—work that would earn him the Nobel Prize in 2024—consist of artificial neurons connected to one another and are among the simplest models of associative memory. “Whenever we see any tree, our brain recalls the concept of a tree. This ability to associate many different representations with the same concept is what we call associative memory,” explains Ricci-Tersenghi, professor of theoretical physics at Sapienza University of Rome and one of the authors of the new study.

Kevin Warwick: Be/Come the Cy/Borg

In February 2011, IBM’s Watson had just beaten two human champions at Jeopardy, and most people filed it under party trick.

A few days later, I sat down with Prof. Kevin Warwick for the second time. He had already run a wire into the median nerve of his own left arm and sent a signal from his nervous system straight into his wife’s. The press called him an eccentric. A few of his colleagues used a less generous word.

So I asked him where the line between genius and madness actually sits. We also got into the magnetic implants and sensory substitution devices his students were building, the trouble his rat-brain-cell robot kept running into, and why Alan Turing was owed far more than Britain ever gave him.

Fifteen years on, #BCI implants have moved from stunt to clinical trial, #AI writes the code that writes the code, and the open question is no longer whether we merge with our machines. It is on whose terms, and who gets a vote.

Kevin’s answer back in 2011 was three words: be/come the #cyborg.

Prophecy or warning? Watch it and tell me which one you hear.

Continue reading “Kevin Warwick: Be/Come the Cy/Borg” | >

AI-powered system offers unprecedented insight into the forces shaping Earth’s climate

The world’s oceans may appear calm from space, but beneath the surface, an intricate web of fast-moving currents drives Earth’s climate. Now, a new study led by Tel Aviv University has unveiled a breakthrough that allows scientists to observe these hidden motions with unprecedented clarity.

The researchers developed GOFLOW, an artificial intelligence-powered system that can reconstruct high-resolution ocean current patterns directly from satellite images. The technology provides scientists with an entirely new way to study the small-scale ocean dynamics that influence weather, climate change and the exchange of heat and gases between the ocean and atmosphere.

The study was led by Roy Barkan, a professor, physical oceanographer and fluid dynamics expert in Tel Aviv University’s Department of Geophysics at the Faculty of Exact Sciences. The research was conducted in collaboration with scientists from the Scripps Institution of Oceanography, UCLA and the University of Rhode Island, and published in Nature Geoscience.

AI extracts hidden material rules from microscopic data to predict large-scale behavior

Researchers from the National University of Singapore (NUS) have developed artificial intelligence (AI) methods that learn the large-scale behavior of complex materials from microscopic data. By automatically identifying a small number of hidden variables that capture the collective behavior of a system, the methods can predict how materials evolve over time while reducing the need for costly simulations.

Understanding the behavior of materials at the macroscopic scale is essential for designing new technologies, from energy-efficient electronics to advanced alloys. However, material properties emerge from the interactions of vast numbers of atoms, and simulating every atom over long periods is often computationally impossible, even on modern supercomputers.

A major challenge in materials science is connecting these microscopic processes, such as atomic motion, to observable material properties. Existing approaches often require large-scale simulations that are prohibitively expensive.

Could A.I. Do Your Job? We Put Agents to the Test

We gave an A.I. tool full access to a laptop with pre-configured apps and sought to answer a simple question: Can artificial intelligence do an office job?

Some corporate executives seem to believe it can. More than 200 tech companies have cut roughly 120,000 jobs this year, according to Layoffs.fyi, an industry tracking site; Meta, Oracle and others have all recently made substantial cuts to their work forces, citing A.I. as the driving force; and after laying off about 1,100 employees, the chief executive of Cloudflare said recently that he expected A.I. to replace workers in middle management, finance and marketing.

In our experiment, we deployed A.I. “agents” to act as office workers and found that they were capable of performing some of the tasks we assigned, but not all of them. The agents, which can act autonomously and make decisions based on detailed instructions, excelled at problems they could solve by writing computer programs. But they struggled with understanding the nuances of human language and at navigating user interfaces like the Chrome web browser.

3D-printable material can heal the body, build better robots and recover critical minerals

A new type of 3D-printable material developed by researchers at The University of Texas at Austin mimics human tissue’s ability to sort and filter, allowing certain molecules to pass through while keeping others out. This broad functionality means the material can be used in a variety of applications across medicine, water and robotics.

Current methods for building small tissue-like materials don’t scale to sizes that can make applications possible, the researchers say. The team overcame these issues of speed and scalability by jamming billions of tiny water droplets tightly together using simple mixing and centrifuge techniques to form large, tissue-like materials in just a few minutes. Each droplet is separated by a thin membrane, allowing the membranes to link up, similar to cell organization in human tissue.

“Tissues can separate and transport ions and molecules; that’s how our kidneys or intestines work, taking only what they need and leaving the rest behind,” said Manish Kumar, professor in the Cockrell School of Engineering’s Fariborz Maseeh Department of Civil, Architectural and Environmental Engineering and the McKetta Department of Chemical Engineering. This work was recently published in Nature Materials.