Artificial Intelligence (AI) is the faculty of a computer system to learn and reason, therefore, mimicking human intelligence. Over the course of the past several years, AI has become an indispensable part of cybersecurity measures. AI can predict cyberattacks with matchless precision, helps to create better security features that can bring down the number of cyberattacks and mitigate its impact on IT infrastructure. Artificial intelligence is a powerful cybersecurity tool for enterprises. It is rapidly turning into a sophisticated protective gear for enterprise cybersecurity, and many enterprises are adopting it at a rapid pace. Statista, in a recent post, noted that in 2019 approximately 83% of organizations based in the United States consider that without AI, their organization fails to deal with cyberattacks.

From programs that can process a vast amount of data for intelligence gathering to the future of autonomous weapons, AI is becoming key to our operations -- and our international competition. Why it matters: Military dominance in the future won't be decided just by the size of a nation's army, but the quality of its algorithms. Driving the news: The National Counterintelligence and Security Center said in a new paper published Friday that China and Russia are using legal and illegal methods to undermine and overtake U.S. dominance in critical industries including AI and autonomous systems, my Axios colleague Zach Basu writes. Yes, but: So is the U.S., particularly in defense. Between the lines: Intelligence gathering and analysis is one of the fields where AI can make the biggest difference now for defense, says George Hoyem, managing partner at In-Q-Tel (IQT), the venture investment unit for the U.S. intelligence community.

Yes, security is hard – no one is ever 100 percent safe from the threats lurking out there. But how is it that time and time again, companies – big companies – are continuing to fall for ransomware attacks? Let's explore the main reasons why, starting with some basics before getting more in-depth: Two-factor authentication (2FA) is probably the easiest security improvement an organization can implement, and it's one of the most advocated-for solutions by infosec professionals. Despite this, we continue to see breaches like Colonial Pipeline occur because organizations have either failed to implement 2FA or have failed to *fully* implement it. Anything that requires a username and password to access should have 2FA enabled.

The incoming US Secretary of the Air Force said that China was winning the battle of Artificial Intelligence over the United States. He admitted that China would soon defeat the United States in this high-tech field. Although the Secretary of the Air Force appointed by President Joe Biden has not yet taken office, he publicly replied to the biggest recent controversy in US political and military circles: the Air Force Chief Software Officer, Nicholas Chaillan, who resigned on October 11 last, said that China had already overtaken the United States and won the battle of Artificial Intelligence against it. Kendall III said he agreed with the statement made by Chaillan. Nicholas Chaillan told the media that the United States not only made slow progress in the field of Artificial Intelligence, but that the said progress was also limited by various rules.

Fox News Flash top headlines are here. Check out what's clicking on Foxnews.com. Social media users took to the internet following President Biden's recent town hall, drawing comparisons between his behavior and that of the cartoon character Beavis from "Beavis and Butt-head" At one point during the town hall, Biden was shown holding his arms bent out in front of him with his fists clinched. That moment was clipped and shared to social media by several people, including political commentator Mike Cernovich, who questioned, "What is Biden doing?" "Biden is straight comedy," wrote former NBA player Andrew Bogut. Other users simply shared photos of the president in the moment alongside pictures of Beavis, who is known for a hyperactive alter-ego, The Great Cornholio, that exhibits the same behavior.

This image shows a view of asteroid Bennu's rocky surface in a region near the equator. Scientists thought Bennu's surface was like a sandy beach, abundant in fine sand and pebbles, which would have been perfect for collecting samples. Past telescope observations from Earth had suggested the presence of large swaths of fine-grained material smaller than a few centimeters called fine regolith. But when NASA's OSIRIS-REx mission arrived at Bennu in late 2018, the mission saw a surface covered in boulders. The mysterious lack of fine regolith became even more surprising when mission scientists observed evidence of processes potentially capable of grinding boulders into fine regolith.

Mozambique is a country in southern Africa, a former Portuguese colony, where almost 40% of girls and adolescents become pregnant before the age of 18. An enormous challenge in developing countries to overcome poverty is to ensure that girls do not become pregnant at an early age. To this end, various international organizations and NGOs are actively working in Africa to help the governments advance this agenda. Currently, there are initiatives of support and sexual and reproductive education for girls, young women, and adolescents through digital media and SMS. In these media there is a dialogue between experts who guide and orient those who write, there are millions of conversations collected in recent years. All these conversations contain unstructured data that are of great value to understand over time the evolution of the concerns of those who use these channels and how to use this information to make better public policy decisions by the government with international support.

In this study, we propose using a neural embedding model-graph neural network (GNN)- that leverages the heterogeneous features of urban areas and their interactions captured by human mobility network to obtain vector representations of these areas. Using large-scale high-resolution mobility data sets from millions of aggregated and anonymized mobile phone users in 16 metropolitan counties in the United States, we demonstrate that our embeddings encode complex relationships among features related to urban components (such as distribution of facilities) and population attributes and activities. The spatial gradient in each direction from city center to suburbs is measured using clustered representations and the shared characteristics among urban areas in the same cluster. Furthermore, we show that embeddings generated by a model trained on a different county can capture 50% to 60% of the emergent spatial structure in another county, allowing us to make cross-county comparisons in a quantitative way. Our GNN-based framework overcomes the limitations of previous methods used for examining spatial structures and is highly scalable. The findings reveal non-linear relationships among urban components and anisotropic spatial gradients in cities. Since the identified spatial structures and gradients capture the combined effects of various mechanisms, such as segregation, disparate facility distribution, and human mobility, the findings could help identify the limitations of the current city structure to inform planning decisions and policies. Also, the model and findings set the stage for a variety of research in urban planning, engineering and social science through integrated understanding of how the complex interactions between urban components and population activities and attributes shape the spatial structures in cities.

The efficacy of machine learning models is typically determined by computing their accuracy on test data sets. However, this may often be misleading, since the test data may not be representative of the problem that is being studied. With QuantifyML we aim to precisely quantify the extent to which machine learning models have learned and generalized from the given data. Given a trained model, QuantifyML translates it into a C program and feeds it to the CBMC model checker to produce a formula in Conjunctive Normal Form (CNF). The formula is analyzed with off-the-shelf model counters to obtain precise counts with respect to different model behavior. QuantifyML enables i) evaluating learnability by comparing the counts for the outputs to ground truth, expressed as logical predicates, ii) comparing the performance of models built with different machine learning algorithms (decision-trees vs. neural networks), and iii) quantifying the safety and robustness of models.

The neural attention mechanism has been incorporated into deep neural networks to achieve state-of-the-art performance in various domains. Most such models use multi-head self-attention which is appealing for the ability to attend to information from different perspectives. This paper introduces alignment attention that explicitly encourages self-attention to match the distributions of the key and query within each head. The resulting alignment attention networks can be optimized as an unsupervised regularization in the existing attention framework. It is simple to convert any models with self-attention, including pre-trained ones, to the proposed alignment attention. On a variety of language understanding tasks, we show the effectiveness of our method in accuracy, uncertainty estimation, generalization across domains, and robustness to adversarial attacks. We further demonstrate the general applicability of our approach on graph attention and visual question answering, showing the great potential of incorporating our alignment method into various attention-related tasks.