Advances in the field of computer science, such as very complex simulations, data analysis, or machine learning (ML) in data-driven applications (for example, computational fluid dynamics, large language models) are leading to an increased demand of IT performance and data storage capacity. Therefore, the electricity demands of digital infrastructures in science and industry are increasing. High-performance computing (HPC) has become an enabling technology and a vital tool to greatly reduce the processing and execution time of advanced computing- or data-intensive tasks. An obvious consequence: HPC datacenters (DCs) require an enormous amount of electricity, have volatile demands, and produce notable amounts of waste heat. If not well located, built, and operated, such infrastructures generate a significant CO2 backpack, and the applications and products that use them inherit the backpack from the computing platform.

A growing number of fire departments across the country are turning to artificial intelligence to help detect and respond to wildfires more quickly. Every time you ask ChatGPT a question, to generate an image or let artificial intelligence summarize your email, something big is happening behind the scenes. Not on your device, but in sprawling data centers filled with servers, GPUs and cooling systems that require massive amounts of electricity. The modern AI boom is pushing our power grid to its limits. ChatGPT alone processes roughly 1 billion queries per day, each requiring data center resources far beyond what's on your device.

Japan is encouraging energy importers to secure liquefied natural gas (LNG) past 2050 -- the deadline the second-biggest buyer of the fossil fuel has set itself for net zero emissions. Several of the country's largest LNG buyers are considering 20-year supply deals with projects that would start after 2030, according to people with knowledge of the discussions, who asked not to be named as the negotiations are private. They aim to deploy technology such as carbon capture and storage to mitigate the emissions from burning the super-chilled fossil fuel under Japan's national target. The government expects a boom in artificial intelligence, data centers and semiconductor chip-making factories to revive power demand, which has been tracking a declining population for years. It sees LNG as vital to energy security, even as it works on increasing renewable energy generation and restarting nuclear reactors idled after the 2011 Fukushima No. 1 disaster.

Is this the electric grid of the future? Lincoln Electric System, a publicly owned utility in Nebraska, is used to weathering severe blizzards. But what will happen soon--not only at Lincoln Electric but for all electric utilities--is a challenge of a different order. Utilities must keep the lights on in the face of more extreme and more frequent storms and fires, growing risks of cyberattacks and physical disruptions, and a wildly uncertain policy and regulatory landscape. They must keep prices low amid inflationary costs. And they must adapt to an epochal change in how the grid works, as the industry attempts to transition from power generated with fossil fuels to power generated from renewable sources like solar and wind.

"People are often curious about how much energy a ChatGPT query uses," Sam Altman, the CEO of OpenAI, wrote in an aside in a long blog post last week. The average query, Altman wrote, uses 0.34 watt-hours of energy: "About what an oven would use in a little over one second, or a high-efficiency lightbulb would use in a couple of minutes." For a company with 800 million weekly active users (and growing), the question of how much energy all these searches are using is becoming an increasingly pressing one. But experts say Altman's figure doesn't mean much without much more public context from OpenAI about how it arrived at this calculation--including the definition of what an "average" query is, whether or not it includes image generation, and whether or not Altman is including additional energy use, like from training AI models and cooling OpenAI's servers. As a result, Sasha Luccioni, the climate lead at AI company Hugging Face, doesn't put too much stock in Altman's number.

Breakthroughs, discoveries, and DIY tips sent every weekday. In recent years, researchers and climate advocates have been ringing the alarm about artificial intelligence's impact on the environment. Advanced and increasingly popular large language models (LLMs)--such as those offered by OpenAI and Google--reside in massive data centers that consume significant amounts of electricity and water to cool servers. Every time someone types a question or phrase into one of these platforms, the energy used to generate a response produces a measurable amount of potentially harmful CO₂. But, according to a new research published in Frontiers in Communication, not all of those prompts leave have the same environmental impact.

Surrogate modeling techniques have become indispensable in accelerating the discovery and optimization of high-entropy alloys(HEAs), especially when integrating computational predictions with sparse experimental observations. This study systematically evaluates the fitting performance of four prominent surrogate models conventional Gaussian Processes(cGP), Deep Gaussian Processes(DGP), encoder-decoder neural networks for multi-output regression and XGBoost applied to a hybrid dataset of experimental and computational properties in the AlCoCrCuFeMnNiV HEA system. We specifically assess their capabilities in predicting correlated material properties, including yield strength, hardness, modulus, ultimate tensile strength, elongation, and average hardness under dynamic and quasi-static conditions, alongside auxiliary computational properties. The comparison highlights the strengths of hierarchical and deep modeling approaches in handling heteroscedastic, heterotopic, and incomplete data commonly encountered in materials informatics. Our findings illustrate that DGP infused with machine learning-based prior outperform other surrogates by effectively capturing inter-property correlations and input-dependent uncertainty. This enhanced predictive accuracy positions advanced surrogate models as powerful tools for robust and data-efficient materials design.

Ferry quays experience rapid deterioration due to their exposure to harsh maritime environments and ferry impacts. Vibration-based structural health monitoring offers a valuable approach to assessing structural integrity and understanding the structural implications of these impacts. However, practical limitations often restrict sensor placement at critical locations. Consequently, virtual sensing techniques become essential for establishing a Digital Twin and estimating the structural response. This study investigates the application of the Gaussian Process Latent Force Model (GPLFM) for virtual sensing on the Magerholm ferry quay, combining in-operation experimental data collected during a ferry impact with a detailed physics-based model. The proposed Physics-Encoded Machine Learning model integrates a reduced-order structural model with a data-driven GPLFM representing the unknown impact forces via their modal contributions. Significant challenges are addressed for the development of the Digital Twin of the ferry quay, including unknown impact characteristics (location, direction, intensity), time-varying boundary conditions, and sparse sensor configurations. Results show that the GPLFM provides accurate acceleration response estimates at most locations, even under simplifying modeling assumptions such as linear time-invariant behavior during the impact phase. Lower accuracy was observed at locations in the impact zone. A numerical study was conducted to explore an optimal real-world sensor placement strategy using a Backward Sequential Sensor Placement approach. Sensitivity analyses were conducted to examine the influence of sensor types, sampling frequencies, and incorrectly assumed damping ratios. The results suggest that the GP latent forces can help accommodate modeling and measurement uncertainties, maintaining acceptable estimation accuracy across scenarios.

AI that can accelerate research could drive a century of technological progress over just a few years. During such a period, new technological or political developments will raise consequential and hard-to-reverse decisions, in rapid succession. We call these developments grand challenges. These challenges include new weapons of mass destruction, AI-enabled autocracies, races to grab offworld resources, and digital beings worthy of moral consideration, as well as opportunities to dramatically improve quality of life and collective decision-making. We argue that these challenges cannot always be delegated to future AI systems, and suggest things we can do today to meaningfully improve our prospects. AGI preparedness is therefore not just about ensuring that advanced AI systems are aligned: we should be preparing, now, for the disorienting range of developments an intelligence explosion would bring.

Accurate acquisition of high-resolution surface meteorological conditions is critical for forecasting and simulating meteorological variables. Directly applying spatial interpolation methods to derive meteorological values at specific locations from low-resolution grid fields often yields results that deviate significantly from the actual conditions. Existing downscaling methods primarily rely on the coupling relationship between geostationary satellites and ERA5 variables as a condition. However, using brightness temperature data from geostationary satellites alone fails to comprehensively capture all the changes in meteorological variables in ERA5 maps. To address this limitation, we can use a wider range of satellite data to make more full use of its inversion effects on various meteorological variables, thus producing more realistic results across different meteorological variables. To further improve the accuracy of downscaling meteorological variables at any location, we propose the Multi-source Observation Down-Scaling Model (MODS). It is a conditional diffusion model that fuses data from multiple geostationary satellites GridSat, polar-orbiting satellites (AMSU-A, HIRS, and MHS), and topographic data (GEBCO), as conditions, and is pre-trained on the ERA5 reanalysis dataset. During training, latent features from diverse conditional inputs are extracted separately and fused into ERA5 maps via a multi-source cross-attention module. By exploiting the inversion relationships between reanalysis data and multi-source atmospheric variables, MODS generates atmospheric states that align more closely with real-world conditions. During sampling, MODS enhances downscaling consistency by incorporating low-resolution ERA5 maps and station-level meteorological data as guidance. Experimental results demonstrate that MODS achieves higher fidelity when downscaling ERA5 maps to a 6.25 km resolution.