In February of this year, the Department of Defense (DoD) issued five Ethical Principles for Artificial Intelligence (AI): Responsible, Equitable, Traceable, Reliable and Governable. The DoD principles build off recommendations from 2019 by the Defense Innovation Board and the interim report of the National Security Commission on AI (NSCAI). The defense industry and others in the private sector have also been considering ethical issues regarding AI, including the issue of whether businesses should have an AI code of ethics. When cyber first became an issue about 22-years ago, the trend was to raise awareness and think through the consequences. Similarly, now we are developing awareness of the issues and beginning to think through the consequences of AI.

Last week, on September 15 and 16, the Pentagon's Joint Artificial Intelligence Center (JAIC) held a meeting with officials from 13 countries, including but not only U.S. allies, around the ethical military uses of artificial intelligence, the first of its kind. Breaking Defense quotes Mark Beall, the JAIC's head of strategy and policy, who called the meeting "historic," as saying, "This group of … countries, to my knowledge, has never been brought together under one banner before." Earlier this year, the Pentagon adopted a set of ethics guidelines around AI use. At a time when China and Russia's pursuit of military AI has raised considerable alarm in Western capitals, Beall noted that the meeting was not about creating a coalition against specific countries. Rather, "we're really focused on, right now, rallying around [shared] core values like digital liberty and human rights… international humanitarian law," Beall said.

The following declaration was released by the Governments of the United States of America and the United Kingdom of Great Britain and Northern Ireland during the September 25 inaugural meeting of the Special Relationship Economic Working Group. We intend to establish a bilateral government-to-government dialogue on the areas identified in this vision and explore an AI R&D ecosystem that promotes the mutual wellbeing, prosperity, and security of present and future generations. Signed in London and Washington on September 25, 2020, in two originals, in the English language.

The Trump administration announced on Friday that the United States and the United Kingdom had signed a new agreement to cooperate on research and development of artificial intelligence.

The Trump administration is set to announce that the United States and the United Kingdom have signed a new agreement to cooperate on research and development of artificial intelligence, in news shared first with Axios. Why it matters: The U.S. and its allies fear China is going to surpass them in AI. The partnership shows the U.S. and U.K. think they have a better chance at beating China by linking up. Details: The partnership will include the two countries working together on research and development of AI, including on issues of explainability and fairness, an administration official told Axios. Flashback: In May 2020, the U.S. and other allies launched the Global Partnership on Artificial Intelligence, for "like-minded nations together to encourage the development of AI in line with our shared values."

Adversarial training, in which a network is trained on both adversarial and clean examples, is one of the most trusted defense methods against adversarial attacks. However, there are three major practical difficulties in implementing and deploying this method - expensive in terms of running memory and computation costs; accuracy trade-off between clean and adversarial examples; cannot foresee all adversarial attacks at training time. Here, we present a new solution to ease these three difficulties - Beneficial perturbation Networks (BPN). BPN generates and leverages beneficial perturbations (somewhat opposite to well-known adversarial perturbations) as biases within the parameter space of the network, to neutralize the effects of adversarial perturbations on data samples. Thus, BPN can effectively defend against adversarial examples. Compared to adversarial training, we demonstrate that BPN can significantly reduce the required running memory and computation costs, by generating beneficial perturbations through recycling of the gradients computed from training on clean examples. In addition, BPN can alleviate the accuracy trade-off difficulty and the difficulty of foreseeing multiple attacks, by improving the generalization of the network, thanks to increased diversity of the training set achieved through neutralization between adversarial and beneficial perturbations.

We propose an Anderson Acceleration (AA) scheme for the adaptive Expectation-Maximization (EM) algorithm for unsupervised learning a finite mixture model from multivariate data (Figueiredo and Jain 2002). The proposed algorithm is able to determine the optimal number of mixture components autonomously, and converges to the optimal solution much faster than its non-accelerated version. The success of the AA-based algorithm stems from several developments rather than a single breakthrough (and without these, our tests demonstrate that AA fails catastrophically). To begin, we ensure the monotonicity of the likelihood function (a the key feature of the standard EM algorithm) with a recently proposed monotonicity-control algorithm (Henderson and Varahdan 2019), enhanced by a novel monotonicity test with little overhead. We propose nimble strategies for AA to preserve the positive definiteness of the Gaussian weights and covariance matrices strictly, and to conserve up to the second moments of the observed data set exactly. Finally, we employ a K-means clustering algorithm using the gap statistic to avoid excessively overestimating the initial number of components, thereby maximizing performance. We demonstrate the accuracy and efficiency of the algorithm with several synthetic data sets that are mixtures of Gaussians distributions of known number of components, as well as data sets generated from particle-in-cell simulations. Our numerical results demonstrate speed-ups with respect to non-accelerated EM of up to 60X when the exact number of mixture components is known, and between a few and more than an order of magnitude with component adaptivity.

This paper focuses on the problem of unsupervised relation extraction. Existing probabilistic generative model-based relation extraction methods work by extracting sentence features and using these features as inputs to train a generative model. This model is then used to cluster similar relations. However, these methods do not consider correlations between sentences with the same entity pair during training, which can negatively impact model performance. To address this issue, we propose a Clustering-based Unsupervised generative Relation Extraction (CURE) framework that leverages an "Encoder-Decoder" architecture to perform self-supervised learning so the encoder can extract relation information. Given multiple sentences with the same entity pair as inputs, self-supervised learning is deployed by predicting the shortest path between entity pairs on the dependency graph of one of the sentences. After that, we extract the relation information using the well-trained encoder. Then, entity pairs that share the same relation are clustered based on their corresponding relation information. Each cluster is labeled with a few words based on the words in the shortest paths corresponding to the entity pairs in each cluster. These cluster labels also describe the meaning of these relation clusters. We compare the triplets extracted by our proposed framework (CURE) and baseline methods with a ground-truth Knowledge Base. Experimental results show that our model performs better than state-of-the-art models on both New York Times (NYT) and United Nations Parallel Corpus (UNPC) standard datasets.

The human brain is the gold standard of adaptive learning. It not only can learn and benefit from experience, but also can adapt to new situations. In contrast, deep neural networks only learn one sophisticated but fixed mapping from inputs to outputs. This limits their applicability to more dynamic situations, where input to output mapping may change with different contexts. A salient example is continual learning - learning new independent tasks sequentially without forgetting previous tasks. Continual learning of multiple tasks in artificial neural networks using gradient descent leads to catastrophic forgetting, whereby a previously learned mapping of an old task is erased when learning new mappings for new tasks. Here, we propose a new biologically plausible type of deep neural network with extra, out-of-network, task-dependent biasing units to accommodate these dynamic situations. This allows, for the first time, a single network to learn potentially unlimited parallel input to output mappings, and to switch on the fly between them at runtime. Biasing units are programmed by leveraging beneficial perturbations (opposite to well-known adversarial perturbations) for each task. Beneficial perturbations for a given task bias the network toward that task, essentially switching the network into a different mode to process that task. This largely eliminates catastrophic interference between tasks. Our approach is memory-efficient and parameter-efficient, can accommodate many tasks, and achieves state-of-the-art performance across different tasks and domains.

The worldwide growth of e-commerce has created new challenges for logistics companies, one of which is being able to deliver products quickly and at low cost, which reflects directly in the way of sorting packages, needing to eliminate steps such as storage and batch creation. Our work presents a multi-agent system that uses trajectory data mining techniques to extract territorial patterns and use them in the dynamic creation of last-mile routes. The problem can be modeled as a Dynamic Capacitated Vehicle Routing Problem (VRP) with Stochastic Customer, being therefore NP-HARD, what makes its implementation unfeasible for many packages. The work's main contribution is to solve this problem only depending on the Warehouse system configurations and not on the number of packages processed, which is appropriate for Big Data scenarios commonly present in the delivery of e-commerce products. Computational experiments were conducted for single and multi depot instances. Due to its probabilistic nature, the proposed approach presented slightly lower performances when compared to the static VRP algorithm. However, the operational gains that our solution provides making it very attractive for situations in which the routes must be set dynamically.