Lifeboat Foundation

Fly, goat, fly! A new AI agent from Google DeepMind can play different games, including ones it has never seen before such as Goat Simulator 3, a fun action game with exaggerated physics. Researchers were able to get it to follow text commands to play seven different games and move around in three different 3D research environments. It’s a step toward more generalized AI that can transfer skills across multiple environments.

Google DeepMind has had huge success developing game-playing AI systems. Its system AlphaGo, which beat top professional player Lee Sedol at the game Go in 2016, was a major milestone that showed the power of deep learning. But unlike earlier game-playing AI systems, which mastered only one game or could only follow single goals or commands, this new agent is able to play a variety of different games, including Valheim and No Man’s Sky. It’s called SIMA, an acronym for “scalable, instructable, multiworld agent.”

Dark energy’s role in propelling the universe’s accelerated expansion presents a pivotal challenge in astrophysics, driving ongoing research and space missions dedicated to uncovering the nature of this mysterious force.

Some 13.8 billion years ago, the universe began with a rapid expansion we call the Big Bang. After this initial expansion, which lasted a fraction of a second, gravity started to slow the universe down. But the cosmos wouldn’t stay this way. Nine billion years after the universe began, its expansion started to speed up, driven by an unknown force that scientists have named dark energy.

But what exactly is dark energy?

Physicists debut a new method to detect gravitational waves with unprecedented precision, providing insights into black hole mergers.

A team of researchers from Peking University and the Eastern Institute of Technology (EIT) in China has developed a new framework to train machine learning models with prior knowledge, such as the laws of physics or mathematical logic, alongside data.

Chinese researchers are on the brink of pioneering a groundbreaking approach to developing ‘AI scientists capable of conducting experiments and solving scientific problems.

Recent advances in deep learning models have revolutionized scientific research, but current models still struggle to simulate real-world physics interactions accurately.

Continue reading “Chinese Researchers on the Brink of Developing ‘Real AI Scientists’ Capable of Conducting Experiments, Solving Scientific Problems” »

“Without a fundamental understanding of the world, a model is essentially an animation rather than a simulation,” said Chen Yuntian, study author and a professor at the Eastern Institute of Technology (EIT).

Deep learning models are generally trained using data and not prior knowledge, which can include things such as the laws of physics or mathematical logic, according to the paper.

But the scientists from Peking University and EIT wrote that when training the models, prior knowledge could be used alongside data to make them more accurate, creating “informed machine learning” models capable of incorporating this knowledge into their output.

The FIT4NANO project has mapped out the expansive applications and future directions of focused ion beam technology, emphasizing its critical role in advancing research and development across multiple disciplines, from microelectronics to life sciences.

Processing materials on the nanoscale, producing prototypes for microelectronics, or analyzing biological samples: The range of applications for finely focused ion beams is huge. Experts from the EU collaboration FIT4NANO have now reviewed the many options and developed a roadmap for the future. The article, published in Applied Physics Review, is aimed at students, users from industry and science as well as research policymakers.

Discovery and Applications.

A team of engineers, physicists, and data scientists from Princeton University and the Princeton Plasma Physics Laboratory (PPPL) have used artificial intelligence (AI) to predict—and then avoid—the formation of a specific type of plasma instability in magnetic confinement fusion tokamaks. The researchers built and trained a model using past experimental data from operations at the DIII-D National Fusion Facility in San Diego, Calif., before proving through real-time experiments that their model could forecast so-called tearing mode instabilities up to 300 milliseconds in advance—enough time for an AI controller to adjust operating parameters and avoid a tear in the plasma that could potentially end the fusion reaction.

Some excellent food for thought face_with_colon_three

We now have everything we need to build a physics engine with infinite precision.

Continue reading “Perfect, Infinite-Precision, Game Physics in Python (Part 3)” »

Dr Freese has also made the case for a Dark Big Bang that could have given rise to dark matter independently of normal matter in the days after the Big Bang. The traditional model of the universe says that matter and dark matter were produced at the same time. The earliest evidence of dark matter, however, only appears later in the early evolution of the universe, when cosmic structure starts to form.

One explanation for this is that matter and dark matter did not, in fact, appear together, but that dark matter entered the universe in a second cataclysmic release of energy from the vacuum—the Dark Big Bang—as much as a month after the traditional Big Bang. The model that Dr Freese and her co-author Martin Winkler explored would explain why dark matter might be completely decoupled from traditional matter and it also naturally produces SIDM candidates. If there was such a Dark Big Bang, it would have left a clear signature—a pattern in the frequencies of the gravitational waves that hum across the universe—that could be picked up by future gravitational-wave detectors.