In a paper published in the journal Nature Physics, DeepMind researchers describe an AI system that can predict the movement of glass molecules as they transition between liquid and solid states. The techniques and trained models, which have been made available in open source, could be used to predict other qualities of interest in glass, DeepMind says. Beyond glass, the researchers assert the work yields insights into general substance and biological transitions, and that it could lead to advances in industries like manufacturing and medicine. "Machine learning is well placed to investigate the nature of fundamental problems in a range of fields," a DeepMind spokesperson told VentureBeat. "We will apply some of the learnings and techniques proven and developed through modeling glassy dynamics to other central questions in science, with the aim of revealing new things about the world around us." Glass is produced by cooling a mixture of high-temperature melted sand and minerals.

Experts around the globe claim that artificial intelligence is an inseparable part of the future business, public sector and especially, national security. But this fact does not imply that AI should be, or could be, implemented quickly and successfully. According to Daniel Newman, Principal Analyst, and Founder at Futurum Research, there are several factors that are hampering the penetration of this technology in numerous industries and fields due to the difficulties they pose to organizations interested in taking this digital leap i.e lack of future vision, fear of job loss, etc. Similarly, anyone who has ever watched any Iron Man movie knows that its principal character, Tony Stark, relies heavily on Jarvis, an advanced Artificial Intelligence system designed to manage almost "everything" in his life, especially in the fight against crime. We might be considering this a far-fetched reality but it could actually be closer than we think when it comes to national security.

Artificial intelligence (AI) is slowly becoming more mainstream, as companies amass large amounts of data and look for the right technologies to analyze and leverage it. That's why Gartner predicted that 80% of emerging technologies will have AI foundations by 2021. With the trend towards predictive analytics, machine learning and other data sciences already underway, marketers need to start paying attention to how they can leverage these techniques to form a more data-driven marketing strategy. With this in mind, we've asked AI industry experts why marketing leaders need to start considering AI, and some of the best open-source AI frameworks to keep tabs on. Dean Abbott, chief data scientist and co-founder of SmarterHQ, believes AI should be top of mind for most business leaders.

Getting lost in the latest tech talk? Don't know your AI from your ML? Then it's time to learn how to talk AI like an expert. The term AI/ML has become exceptionally popular but what exactly to the experts mean they refer to AI/ML? "AI/ML" is a field in computer science which tries to use computers to solve problems that were previously only "solvable" by humans. For years, computers excelled at number crunching but certain tasks, such as voice recognition (hearing), object recognition in photos (vision) and predicting the behaviour of agents in unconstrained or uncertain environments, remained a challenge for programmers to overcome. These days, computer algorithms can perform as well or better than humans on a few narrow tasks that typically take a human between 0 and 3 seconds (rule of thumb) to perform.

Despite the huge contributions of deep learning to the field of artificial intelligence, there's something very wrong with it: It requires huge amounts of data. This is one thing that both the pioneers and critics of deep learning agree on. In fact, deep learning didn't emerge as the leading AI technique until a few years ago because of the limited availability of useful data and the shortage of computing power to process that data. Reducing the data-dependency of deep learning is currently among the top priorities of AI researchers. In his keynote speech at the AAAI conference, computer scientist Yann LeCun discussed the limits of current deep learning techniques and presented the blueprint for "self-supervised learning," his roadmap to solve deep learning's data problem.

In a new study in China, a deep learning model detected COVID19 caused pneumonia from CT scans with comparable performance to expert radiologists. CT is the preferred imaging method for evaluating lung infection, assessing progression, and determining treatment options for patients with pneumonia caused by COVID19. In this study, radiologists used AI to help them evaluate the progression of disease, and with the assistance of this model, radiologists' read time decreased by 65%. The model achieved a per-patient sensitivity of 100% and accuracy of 95.24%. This AI could help improve the efficiency of evaluation and diagnosis especially if the number of people with the virus increases. This article is a preprint and has not been peer-reviewed. This paper reports new medical research that has yet to be evaluated and so should not be used to guide clinical practice. Harvard Medical School students have created a COVID19 curriculum. It includes information about epidemiology, clinical management, testing, treatment, vaccine development, and communication. Each section was reviewed by at least two Harvard Medical School faculty experts. Many modules reference supplemental resources that may be worth accessing in the future and to find the most current statistics of the pandemic. Also included are one-page summaries of each module's key takeaways. COVID19 testing in South Korea is free and convenient and over 250,000 people have already been tested. The South Korean data is valuable because they are testing people who have symptoms and people who have no symptoms. This is unusual because most countries are only testing people who are sick to confirm that they have the virus. South Korea is testing everyone, including asymptomatic people, as a public health measure so that anyone who has the virus can isolate even if they don't feel sick. In most countries asymptomatic people are not tested for COVID19. For example Italy is only testing symptomatic people, whereas South Korea tests everyone and picks up more mild cases.

A chest X-ray, analyzed by Qure.ai's software, picks up on abnormalities that suggest the likelihood of COVID-19 infection. X-rays are one of the quickest, simplest ways to diagnose the disease, and an army of AI specialists around the world are trying to speed up how the images are used to find cases. Most cite the lack of data as the prime obstacle to broader adoption of AI. For all the frantic effort to coordinate life-saving work around the globe during the COVID-19 pandemic, the digital age finds itself hampered in one very specific respect: information. Teams of artificial intelligence researchers are trying to bring decades of technology to bear on the problem of diagnosing and treating the disease, but the data they need to develop their software programs is scattered around the globe, making it practically inaccessible. The painful lack of data is evident in one particular use case for AI, the development of diagnostic tests for COVID-19 based on X-rays or on "computed tomography" scans of the lungs.

Training in supervised deep learning is computationally demanding, and the convergence behavior is usually not fully understood. We introduce and study a second-order stochastic quasi-Gauss--Newton (SQGN) optimization method that combines ideas from stochastic quasi-Newton methods, Gauss--Newton methods, and variance reduction to address this problem. SQGN provides excellent accuracy without the need for experimenting with many hyper-parameter configurations, which is often computationally prohibitive given the number of combinations and the cost of each training process. We discuss the implementation of SQGN with TensorFlow, and we compare its convergence and computational performance to selected first-order methods using the MNIST benchmark and a large-scale seismic tomography application from Earth science.

With the rapid emergence of deep learning (DL) technology, it has been successfully used in various fields including aquaculture. This change can create new opportunities and a series of challenges for information and data processing in smart fish farming. This paper focuses on the applications of DL in aquaculture, including live fish identification, species classification, behavioral analysis, feeding decision-making, size or biomass estimation, water quality prediction. In addition, the technical details of DL methods applied to smart fish farming are also analyzed, including data, algorithms, computing power, and performance. The results of this review show that the most significant contribution of DL is the ability to automatically extract features. However, challenges still exist; DL is still in an era of weak artificial intelligence. A large number of labeled data are needed for training, which has become a bottleneck restricting further DL applications in aquaculture. Nevertheless, DL still offers breakthroughs in the handling of complex data in aquaculture. In brief, our purpose is to provide researchers and practitioners with a better understanding of the current state of the art of DL in aquaculture, which can provide strong support for the implementation of smart fish farming.

Deep neural networks are revolutionizing the way complex systems are developed. However, these automatically-generated networks are opaque to humans, making it difficult to reason about them and guarantee their correctness. Here, we propose a novel approach for verifying properties of a widespread variant of neural networks, called recurrent neural networks. Recurrent neural networks play a key role in, e.g., natural language processing, and their verification is crucial for guaranteeing the reliability of many critical systems. Our approach is based on the inference of invariants, which allow us to reduce the complex problem of verifying recurrent networks into simpler, non-recurrent problems. Experiments with a proof-of-concept implementation of our approach demonstrate that it performs orders-of-magnitude better than the state of the art.