The need for construction is more significant than ever. The projected 70% increase in urban population over the next 15 years will require many new buildings. Although the European Union anticipates that such a need will arise, builders still do not see this opportunity. So if you want to enter the construction industry or any other profession in this field, I have good news for you -- you are living in a hellish boom time! Unfortunately, this boom will lead to a climate catastrophe on a hitherto unknown scale.

The report records the DOE Town Halls held during 2019 at Argonne National Laboratory, Oak Ridge National Laboratory, Lawrence Berkeley National Laboratory, and in Washington, DC. The AI for Science city center conversations concentrated on recording the transformational usages of AI that utilize HPC and/or information analysis, leveraging data collections from HPC simulations or instruments and customer centers, and dealing with scientific challenges one-of-akind to DOE user facilities and the company's comprehensive basic and used scientific research venture. Artificial intelligence and machine learning systems have the potential to influence the future layout and implementation of cybersecurity systems for the power grid. Artificial intelligence is the research of intelligence agents as shown by machines. Commonly used supervised learning strategies include deep learning and other machine learning methods that call for less information than deep learning, e. G. Support vector machines, random forests.

Researchers have developed an AI model to help computers work more efficiently with a wider variety of languages. African languages have received relatively little attention from computer scientists, so few natural language processing capabilities have been available to large swaths of the continent. A new language model, developed by researchers at the University of Waterloo's David R. Cheriton School of Computer Science, begins to fill that gap by enabling computers to analyze text in African languages for many useful tasks. The new neural network model, which the researchers have dubbed AfriBERTa, uses deep-learning techniques to achieve state-of-the-art results for low-resource languages. The neural network language model works specifically with 11 African languages, such as Amharic, Hausa, and Swahili, spoken collectively by more than 400 million people.

The series, which is produced by BBC StoryWorks Commercial Productions and presented by the International Science Council (ISC), includes films, articles and podcasts which will be hosted on a dedicated BBC.com StoryWorks webpage. The series explores how scientific culture is changing for the better, towards a future of more effective and inclusive citizen engagement, interdisciplinary and international cooperation, and open knowledge-sharing. This five-minute film highlights the innovative use of shipwrecks to map the seabed to inform the siting of offshore windfarms as the seas around Ireland provide an abundance of wind resources. Shipwrecks disturb near-seabed currents, causing certain types of sediments to be washed away or eroded. By studying these changes, we can better predict how man-made structures including wind turbines, will behave on the seabed over time.

Argonne looks to exascale and beyond, sorting out the relationship between computing and experimental facilities, the need for speed and AI's role in making it all work. In 1949, physicists at the U.S. Department of Energy's (DOE) newly minted Argonne National Laboratory ordered the construction of the Argonne Version of the Institute's Digital Automatic Computer, or AVIDAC. A modified version of the first electronic computer built at the Institute for Advanced Study in Princeton, New Jersey, it was intended to help solve complex problems in the design of nuclear reactors. With a floor area of 500 square feet and power consumption of 20 kilowatts, AVIDAC boasted remarkable computing power for the time. It possessed a memory of 1,024 words (about 5.1 kilobytes in total), could perform 1,000 multiplications per second, and had a programming capability that allowed it to solve problems consistently and accurately. Today, your smart phone can store around 100 million times more data, and can do in a single second what would have taken AVIDAC two months.

Generative Flow Networks (GFlowNets) have been introduced as a method to sample a diverse set of candidates in an active learning context, with a training objective that makes them approximately sample in proportion to a given reward function. In this paper, we show a number of additional theoretical properties of GFlowNets. They can be used to estimate joint probability distributions and the corresponding marginal distributions where some variables are unspecified and, of particular interest, can represent distributions over composite objects like sets and graphs. GFlowNets amortize the work typically done by computationally expensive MCMC methods in a single but trained generative pass. They could also be used to estimate partition functions and free energies, conditional probabilities of supersets (supergraphs) given a subset (subgraph), as well as marginal distributions over all supersets (supergraphs) of a given set (graph). We introduce variations enabling the estimation of entropy and mutual information, sampling from a Pareto frontier, connections to reward-maximizing policies, and extensions to stochastic environments, continuous actions and modular energy functions.

Recent research has shown that supervised learning can be an effective tool for designing optimal feedback controllers for high-dimensional nonlinear dynamic systems. But the behavior of these neural network (NN) controllers is still not well understood. In this paper we use numerical simulations to demonstrate that typical test accuracy metrics do not effectively capture the ability of an NN controller to stabilize a system. In particular, some NNs with high test accuracy can fail to stabilize the dynamics. To address this we propose two NN architectures which locally approximate a linear quadratic regulator (LQR). Numerical simulations confirm our intuition that the proposed architectures reliably produce stabilizing feedback controllers without sacrificing optimality. In addition, we introduce a preliminary theoretical result describing some stability properties of such NN-controlled systems.

The members of her lab study fuel cells, which convert hydrogen and oxygen into electricity (and water). They study electrolyzers, which go the other way, using electricity to convert water into hydrogen and oxygen. They even study computers that attempt to mimic the way the brain processes information in learning. What brings all this together in her lab is the electrochemistry of ionic-electronic oxides and their interfaces. "It may seem like we've been contributing to different technologies," says Yildiz, MIT's Breene M. Kerr (1951) Professor in the Department of Nuclear Science and Engineering (NSE) and the Department of Materials Science and Engineering, who was recently named a fellow of the American Physical Society.

Computers and science are intertwined – and not just as tools that help humans connect and collaborate. With computers, scientists model the earth's climate, design alternative energy strategies and simulate exploding stars. From laptops to the world's fastest supercomputers, software innovations and artificial intelligence are reshaping how we interact with mounds of data from healthcare to high-energy physics and how we solve critical problems. Computational science brings together mathematics, computer science and hardware and science expertise to take on these challenges. In this podcast, you'll meet the scientists doing this work, learn more about their research and gain insights into the workings of this dynamic field.

Over the years, accelerating neural networks with quantization has been widely studied. Unfortunately, prior efforts with diverse precisions (e.g., 1-bit weights and 2-bit activations) are usually restricted by limited precision support on GPUs (e.g., int1 and int4). To break such restrictions, we introduce the first Arbitrary Precision Neural Network framework (APNN-TC) to fully exploit quantization benefits on Ampere GPU Tensor Cores. Specifically, APNN-TC first incorporates a novel emulation algorithm to support arbitrary short bit-width computation with int1 compute primitives and XOR/AND Boolean operations. Second, APNN-TC integrates arbitrary precision layer designs to efficiently map our emulation algorithm to Tensor Cores with novel batching strategies and specialized memory organization. Third, APNN-TC embodies a novel arbitrary precision NN design to minimize memory access across layers and further improve performance. Extensive evaluations show that APNN-TC can achieve significant speedup over CUTLASS kernels and various NN models, such as ResNet and VGG.