The UK government will reveal its plans for regulating artificial intelligence (AI) today, and says it wants to hand more powers to existing regulators to deal with algorithms and automated systems, rather than setting up a dedicated body to look at issues around AI. Plans outlined in a new AI paper, being published this morning, would involve regulators such as the Information Commissioner's Office (ICO) and the Competition and Markets Authority being asked to monitor the impact of AI on their sectors, based on a set of guiding principles. The government says the regulators will be encouraged to take a "light touch" approach to enforcing these principles. The paper will be published this morning when the Data Protection and Digital Information Bill, previously referred to as the Data Reform Bill, which sets the UK's post-Brexit data regime, is introduced in parliament. Full details of the UK AI regulations have yet to be revealed, but the government says its plans will "allow different regulators to take a tailored approach to the use of AI in a range of settings." It claims this "better reflects the growing use of AI in a range of sectors".

Iran has supplied drone technology to Hezbollah in Lebanon; to Houthi rebels in Yemen attacking Saudi Arabia and the United Arab Emirates; and to Shiite militias in Iraq, which have carried out strikes against Iraqi and American troops. "Russia is turning to an ally that has flown drones in complex environments in large numbers," said Samuel Bendett, a specialist on Russian drones and other weapons at CNA, a research and analysis organization in Arlington, Va. "While the Russians still have drones, they don't have all the types they need." Russia's deal with Iran underscores the ever-growing importance of drones to modern warfare, not just in insurgencies or counterterrorism operations but also in classic conventional-style conflicts. In a contested battlefield like Ukraine where dueling artillery barrages are the deciding factors if an offensive fails or succeeds, drones play a pivotal role.

Air, land, and sea-launch-able, the fully autonomous, AI-controlled missile known as the Ice Breaker has finally been revealed. A fifth-generation, self-guided missile designed to attack at long distances and travel extremely close to the ground while resisting electronic jamming and other countermeasures, the stealthy Ice Breaker is set for a field demonstration at the Farnborough International Airshow, July 18-22. In science fiction, high-tech missiles can hug the earth's surface, avoid natural obstacles and enemy ordinance alike, survive electronic jamming efforts, and then use complex machine learning to locate, attack, and destroy their intended target with robotic efficiency. In the real world, and even as hypersonic weapons are beginning to enter the modern theater of war, such theoretical, ultra-smart missiles have yet to come to fruition. More recently, rumors of an air-launched, long-range, self-guided and AI-controlled missile surfaced when Israeli manufacturer Rafael Advanced Defense Systems premiered the Ice Breaker's predecessor naval weapon, the Sea Breaker sea-launched missile in 2021.

Humans are not going to be completely replaced by technology in the UAE, said Omar Sultan Al Olama, Minister of State for Artificial Intelligence, Digital Economy and Remote Work Applications. "We don't believe in the UAE that humans are going to be completely replaced by technology… We have seen certain sectors that have completely transformed such as aviation, automobiles, and transport as well. They are increasingly becoming technology-driven sectors and no longer dependent on humans," Al Olama said at the launch of the National Digital Talent Programme. Launched by Emirates NBD, the programme will nurture a pool of 300 young future-ready interns over the next four years. The programme has been launched in cooperation with the Minister of State for Artificial Intelligence, Digital Economy and Remote Work Applications Office, Higher Colleges of Technology and the University of Sharjah.

Richtmyer-Meshkov Instability (RMI) is a complicated phenomenon that occurs when a shockwave passes through a perturbed interface. Over a thousand hydrodynamic simulations were performed to study the formation of RMI for a parameterized high velocity impact. Deep learning was used to learn the temporal mapping of initial geometric perturbations to the full-field hydrodynamic solutions of density and velocity. The continuity equation was used to include physical information into the loss function, however only resulted in very minor improvements at the cost of additional training complexity. Predictions from the deep learning model appear to accurately capture temporal RMI formations for a variety of geometric conditions within the domain. First principle physical laws were investigated to infer the accuracy of the model's predictive capability. While the continuity equation appeared to show no correlation with the accuracy of the model, conservation of mass and momentum were weakly correlated with accuracy. Since conservation laws can be quickly calculated from the deep learning model, they may be useful in applications where a relative accuracy measure is needed.

While sequential recommender systems achieve significant improvements on capturing user dynamics, we argue that sequential recommenders are vulnerable against substitution-based profile pollution attacks. To demonstrate our hypothesis, we propose a substitution-based adversarial attack algorithm, which modifies the input sequence by selecting certain vulnerable elements and substituting them with adversarial items. In both untargeted and targeted attack scenarios, we observe significant performance deterioration using the proposed profile pollution algorithm. Motivated by such observations, we design an efficient adversarial defense method called Dirichlet neighborhood sampling. Specifically, we sample item embeddings from a convex hull constructed by multi-hop neighbors to replace the original items in input sequences. During sampling, a Dirichlet distribution is used to approximate the probability distribution in the neighborhood such that the recommender learns to combat local perturbations. Additionally, we design an adversarial training method tailored for sequential recommender systems. In particular, we represent selected items with one-hot encodings and perform gradient ascent on the encodings to search for the worst case linear combination of item embeddings in training. As such, the embedding function learns robust item representations and the trained recommender is resistant to test-time adversarial examples. Extensive experiments show the effectiveness of both our attack and defense methods, which consistently outperform baselines by a significant margin across model architectures and datasets.

Identifying the diverse and often competing values of citizens, and resolving the consequent public value conflicts, are of significant importance for inclusive and integrated urban development. Scholars have highlighted that relational, value-laden urban space gives rise to many diverse conflicts that vary both spatially and temporally. Although notions of public value conflicts have been conceived in theory, there are very few empirical studies that identify such values and their conflicts in urban space. Building on public value theory and using a case-study mixed-methods approach, this paper proposes a new approach to empirically investigate public value conflicts in urban space. Using unstructured participatory data of 4,528 citizen contributions from a Public Participation Geographic Information Systems in Hamburg, Germany, natural language processing and spatial clustering techniques are used to identify areas of potential value conflicts. Four expert workshops assess and interpret these quantitative findings. Integrating both quantitative and qualitative results, 19 general public values and a total of 9 archetypical conflicts are identified. On the basis of these results, this paper proposes a new conceptual tool of Public Value Spheres that extends the theoretical notion of public-value conflicts and helps to further account for the value-laden nature of urban space.

The most difficult task in machine learning is to interpret trained shallow neural networks. Deep neural networks (DNNs) provide impressive results on a larger number of tasks, but it is generally still unclear how decisions are made by such a trained deep neural network. Providing feature importance is the most important and popular interpretation technique used in shallow and deep neural networks. In this paper, we develop an algorithm extending the idea of Garson Algorithm to explain Deep Belief Network based Auto-encoder (DBNA). It is used to determine the contribution of each input feature in the DBN. It can be used for any kind of neural network with many hidden layers. The effectiveness of this method is tested on both classification and regression datasets taken from literature. Important features identified by this method are compared against those obtained by Wald chi square (\c{hi}2). For 2 out of 4 classification datasets and 2 out of 5 regression datasets, our proposed methodology resulted in the identification of better-quality features leading to statistically more significant results vis-\`a-vis Wald \c{hi}2.

We describe our two-stage system for the Multilingual Information Access (MIA) 2022 Shared Task on Cross-Lingual Open-Retrieval Question Answering. The first stage consists of multilingual passage retrieval with a hybrid dense and sparse retrieval strategy. The second stage consists of a reader which outputs the answer from the top passages returned by the first stage. We show the efficacy of using a multilingual language model with entity representations in pretraining, sparse retrieval signals to help dense retrieval, and Fusion-in-Decoder. On the development set, we obtain 43.46 F1 on XOR-TyDi QA and 21.99 F1 on MKQA, for an average F1 score of 32.73. On the test set, we obtain 40.93 F1 on XOR-TyDi QA and 22.29 F1 on MKQA, for an average F1 score of 31.61. We improve over the official baseline by over 4 F1 points on both the development and test sets.

The increasingly sophisticated and growing number of threat actors along with the sheer speed at which cyber attacks unfold, make timely identification of attacks imperative to an organisations' security. Consequently, persons responsible for security employ a large variety of information sources concerning emerging attacks, attackers' course of actions or indicators of compromise. However, a vast amount of the needed security information is available in unstructured textual form, which complicates the automated and timely extraction of attackers' Tactics, Techniques and Procedures (TTPs). In order to address this problem we systematically evaluate and compare different Natural Language Processing (NLP) and machine learning techniques used for security information extraction in research. Based on our investigations we propose a data processing pipeline that automatically classifies unstructured text according to attackers' tactics and techniques derived from a knowledge base of adversary tactics, techniques and procedures.