Fewer than 400 North Atlantic right whales remain in the wild, and not even 100 of them are breeding females. Their biggest survival threats are boat strikes and entanglement in fishing gear. Protecting these whales, such as by diverting boats from dangerous encounters, requires locating them more reliably--and new technology, described in the Journal of the Acoustical Society of America, could help make that possible. To listen for marine life, researchers often deploy underwater microphones called hydrophones on buoys and robotic gliders. The recorded audio is converted into spectrograms: visual representations of sound used to pinpoint, for instance, specific whale species' calls.

A study from MIT has used machine learning techniques to identify biased phrasing across around 100 of the largest and most influential news outlets in the US and beyond, including 83 of the most influential print news publications. It's a research effort that shows the way towards automated systems that could potentially auto-classify the political character of a publication, and give readers a deeper insight into the ethical stance of an outlet on topics that they may feel passionately about. The work centers on the way topics are addressed with particular phrasing, such as undocumented immigrant illegal Immigrant, fetus unborn baby, demonstrators anarchists. The project used Natural Language Processing (NLP) techniques to extract and classify such instances of'charged' language (on the assumption that apparently more'neutral' terms also represent a political stance) into a broad mapping that reveals left and right-leaning bias across over three million articles from around 100 news outlets, resulting in a navigable bias landscape of the publications in question. The paper comes from Samantha D'Alonzo and Max Tegmark at MIT's Department of Physics, and observes that a number of recent initiatives around'fact checking', in the wake of numerous'fake news' scandals, can be interpreted as disingenuous and serving the causes of particular interests.

Machine learning has been around for decades, but for much of that time, businesses were only deploying a few models and those required tedious, painstaking work done by PhDs and machine learning experts. Over the past couple of years, machine learning has grown significantly thanks to the advent of widely available, standardized, cloud-based machine learning platforms. Today, companies across every industry are deploying millions of machine learning models across multiple lines of business. Tax and financial software giant Intuit started with a machine learning model to help customers maximize tax deductions; today, machine learning touches nearly every part of their business. In the last year alone, Intuit has increased the number of models deployed across their platform by over 50 percent.

The Transform Technology Summits start October 13th with Low-Code/No Code: Enabling Enterprise Agility. Google today announced TensorFlow Similarity, a Python package designed to train similarity models with the company's TensorFlow machine learning framework. Similarity models search for related items, for example finding similar-looking clothes and identifying currently playing songs. As Google explains, many similarity models are trained using a technique called contrastive learning. Contrastive learning, in turn, relies on clustering algorithms, which automatically identify patterns in data by operating on the theory that data points in groups should have similar features.

Artificial intelligence is being deployed in many different areas. Within higher education, it is used for college admissions and financial aid decisions. Health researchers employ it to scan the scientific literature for chemical compounds that may generate new medical treatments. E-commerce sites deploy algorithms to make product recommendations for consumers based on their areas of interest.1 But one of the most important growth areas lies in finance and operations. Both public and private sector organizations have large budgets to manage and it is important to operate efficiently and effectively. Accusations of budget inefficiencies or wasteful spending decrease public confidence and make it important to figure out how to manage resources in fair ways. To help with budgetary oversight, AI is being used for financial management and fraud detection. Advanced algorithms can spot abnormalities and outliers that can be referred to human investigators to determine if fraud actually has taken place. It is a way to use technology to improve budget audits, personnel performance, and organizational activities. Yet is it crucial to overcome several problems that plague public sector innovation: procurement obstacles, insufficiently trained workers, data limitations, a lack of technical standards, cultural barriers to organizational change, and making sure anti-fraud applications adhere to responsible AI principles.

A technology based on artificial intelligence is helping to spot biomarkers and document the progression of amyotrophic lateral sclerosis (ALS) in a large speech study being conducted by EverythingALS. The technology, developed by Modality.ai, is a web-based computer program that uses audio (speech) and video (facial) recordings to assess neurological states automatically through AI and machine learning algorithms. Its greatest advantage is that data can be collected remotely at home on any computer device with the help of a virtual assistant called "Tina." This is important for people with ALS, who often have limited mobility due to muscle weakness, which may affect their ability to participate in clinical studies. "Our mission is to discover and deploy initiatives that focus on new ways to diagnose and treat neurological disorders at the intersection of computing and brain science with a focus on ALS," Indu Navar, CEO and co-founder of EverythingALS, a U.S. nonprofit that is part of the Peter Cohen Foundation, said in a press release.

Earth system models are the most important tools for quantitatively describing the physical state of Earth, and--for example, in the context of climate models--predicting how it might change in the future under the influence of human activities. How the increasingly used methods of artificial intelligence (AI) can help to improve these forecasts and where the limits of the two approaches lie has now been investigated by an international team led by Christopher Irrgang from the German Research Centre for Geosciences Potsdam (GFZ) in a Perspectives article for the journal Nature Machine Intelligence. One key proposal: To merge both approaches into a self-learning "neural Earth system modeling." The development of Earth is a complex interplay of many factors, including the land surface with flora and fauna, the oceans with their ecosystem, the polar regions, the atmosphere, the carbon cycle and other biogeochemical cycles, and radiation processes. Researchers therefore speak of the Earth system.

Machine learning models trained on the visual representation of website code can help improve the accuracy and speed of detecting phishing websites. This is according to a paper (PDF) by security researchers at the University of Plymouth and the University of Portsmouth, UK. The researchers aim to address the shortcomings of existing detection methods, which are either too slow or not accurate enough. The technique developed by the researchers uses "binary visualization" libraries to transform the markup and code of web pages into images. Using this method, they created a dataset of legitimate and phishing images of websites.

The Transform Technology Summits start October 13th with Low-Code/No Code: Enabling Enterprise Agility. AI-powered coding tools, which generate code using machine learning algorithms, have attracted increasing attention over the last decade. In theory, systems like OpenAI's Codex could reduce the time people spend writing software as well as computational and operational costs. But existing systems have major limitations, leading to undesirable results like errors. In search of a better approach, researchers at Salesforce open-sourced a machine learning system called CodeT5, which can understand and generate code in real time. The team claims that CodeT5 achieves state-of-the-art performance on coding tasks including code defect detection, which predicts whether code is vulnerable to exploits, and clone detection, which predicts whether two code snippets have the same functionality.

Until now, chip design has been the domain of electrical engineers, but a recent Google study could change that. It showed that the AI-created chip layout was "superior or comparable to those produced by humans in all key metrics, including power consumption, performance, and chip area." Thanks to a machine-learning technique known as reinforcement learning, artificial intelligence completed the task in only six hours, compared with weeks by humans. Although Alphabet's Google GOOG, -0.44% and Nvidia NVDA, -1.43% have been performing tests and discussing the use of AI-powered techniques to boost chip-production capabilities, Samsung Electronics was among the first to actually create chips using the method. Relying on software made by Synopsys SNPS, 0.08%, a chip design software company, Samsung designed Exynos, a processor used in company's wearables, smartphones, car infotainment systems, and other gadgets.