Artificial intelligence (AI) could lead to UK health services that disadvantage women and ethnic minorities, scientists are warning. They are calling for biases in the systems to be rooted out before their use becomes commonplace in the NHS. They fear that without that preparation AI could dramatically deepen existing health inequalities in our society. The researchers examined the state of the art approach to AI used by hospitals worldwide and found it had a 70 per cent success rate in predicting liver disease from blood tests. But they uncovered a wide gender gap underneath – with 44 per cent of cases in women missed, compared with 23 per cent of cases among men.

Unsupervised question answering is an attractive task due to its independence on labeled data. Previous works usually make use of heuristic rules as well as pre-trained models to construct data and train QA models. However, most of these works regard named entity (NE) as the only answer type, which ignores the high diversity of answers in the real world. To tackle this problem, we propose a novel unsupervised method by diversifying answers, named DiverseQA. Specifically, the proposed method is composed of three modules: data construction, data augmentation and denoising filter. Firstly, the data construction module extends the extracted named entity into a longer sentence constituent as the new answer span to construct a QA dataset with diverse answers. Secondly, the data augmentation module adopts an answer-type dependent data augmentation process via adversarial training in the embedding level. Thirdly, the denoising filter module is designed to alleviate the noise in the constructed data. Extensive experiments show that the proposed method outperforms previous unsupervised models on five benchmark datasets, including SQuADv1.1, NewsQA, TriviaQA, BioASQ, and DuoRC. Besides, the proposed method shows strong performance in the few-shot learning setting.

The COVID-19 pandemic disrupted everyone's life across the world. In this work, we characterize the subjective wellbeing patterns of 112 cities across the United States during the pandemic prior to vaccine availability, as exhibited in subreddits corresponding to the cities. We quantify subjective wellbeing using positive and negative affect. We then measure the pandemic's impact by comparing a community's observed wellbeing with its expected wellbeing, as forecasted by time series models derived from prior to the pandemic.We show that general community traits reflected in language can be predictive of community resilience. We predict how the pandemic would impact the wellbeing of each community based on linguistic and interaction features from normal times \textit{before} the pandemic. We find that communities with interaction characteristics corresponding to more closely connected users and higher engagement were less likely to be significantly impacted. Notably, we find that communities that talked more about social ties normally experienced in-person, such as friends, family, and affiliations, were actually more likely to be impacted. Additionally, we use the same features to also predict how quickly each community would recover after the initial onset of the pandemic. We similarly find that communities that talked more about family, affiliations, and identifying as part of a group had a slower recovery.

Accurate, detailed, and timely crop type mapping is a very valuable information for the institutions in order to create more accurate policies according to the needs of the citizens. In the last decade, the amount of available data dramatically increased, whether it can come from Remote Sensing (using Copernicus Sentinel-2 data) or directly from the farmers (providing in-situ crop information throughout the years and information on crop rotation). Nevertheless, the majority of the studies are restricted to the use of one modality (Remote Sensing data or crop rotation) and never fuse the Earth Observation data with domain knowledge like crop rotations. Moreover, when they use Earth Observation data they are mainly restrained to one year of data, not taking into account the past years. In this context, we propose to tackle a land use and crop type classification task using three data types, by using a Hierarchical Deep Learning algorithm modeling the crop rotations like a language model, the satellite signals like a speech signal and using the crop distribution as additional context vector. We obtained very promising results compared to classical approaches with significant performances, increasing the Accuracy by 5.1 points in a 28-class setting (.948), and the micro-F1 by 9.6 points in a 10-class setting (.887) using only a set of crop of interests selected by an expert. We finally proposed a data-augmentation technique to allow the model to classify the crop before the end of the season, which works surprisingly well in a multimodal setting.

Time series anomaly detection is extensively studied in statistics, economics, and computer science. Over the years, numerous methods have been proposed for time series anomaly detection using deep learning-based methods. Many of these methods demonstrate state-of-the-art performance on benchmark datasets, giving the false impression that these systems are robust and deployable in many practical and industrial real-world scenarios. In this paper, we demonstrate that the performance of state-of-the-art anomaly detection methods is degraded substantially by adding only small adversarial perturbations to the sensor data. We use different scoring metrics such as prediction errors, anomaly, and classification scores over several public and private datasets ranging from aerospace applications, server machines, to cyber-physical systems in power plants. Under well-known adversarial attacks from Fast Gradient Sign Method (FGSM) and Projected Gradient Descent (PGD) methods, we demonstrate that state-of-the-art deep neural networks (DNNs) and graph neural networks (GNNs) methods, which claim to be robust against anomalies and have been possibly integrated in real-life systems, have their performance drop to as low as 0%. To the best of our understanding, we demonstrate, for the first time, the vulnerabilities of anomaly detection systems against adversarial attacks. The overarching goal of this research is to raise awareness towards the adversarial vulnerabilities of time series anomaly detectors.

The European Union is likely to introduce among the first, most stringent, and most comprehensive AI regulatory regimes of the world's major jurisdictions. In this report, we ask whether the EU's upcoming regulation for AI will diffuse globally, producing a so-called "Brussels Effect". Building on and extending Anu Bradford's work, we outline the mechanisms by which such regulatory diffusion may occur. We consider both the possibility that the EU's AI regulation will incentivise changes in products offered in non-EU countries (a de facto Brussels Effect) and the possibility it will influence regulation adopted by other jurisdictions (a de jure Brussels Effect). Focusing on the proposed EU AI Act, we tentatively conclude that both de facto and de jure Brussels effects are likely for parts of the EU regulatory regime. A de facto effect is particularly likely to arise in large US tech companies with AI systems that the AI Act terms "high-risk". We argue that the upcoming regulation might be particularly important in offering the first and most influential operationalisation of what it means to develop and deploy trustworthy or human-centred AI. If the EU regime is likely to see significant diffusion, ensuring it is well-designed becomes a matter of global importance.

Kernel methods in Quantum Machine Learning (QML) have recently gained significant attention as a potential candidate for achieving a quantum advantage in data analysis. Among other attractive properties, when training a kernel-based model one is guaranteed to find the optimal model's parameters due to the convexity of the training landscape. However, this is based on the assumption that the quantum kernel can be efficiently obtained from a quantum hardware. In this work we study the trainability of quantum kernels from the perspective of the resources needed to accurately estimate kernel values. We show that, under certain conditions, values of quantum kernels over different input data can be exponentially concentrated (in the number of qubits) towards some fixed value, leading to an exponential scaling of the number of measurements required for successful training. We identify four sources that can lead to concentration including: the expressibility of data embedding, global measurements, entanglement and noise. For each source, an associated concentration bound of quantum kernels is analytically derived. Lastly, we show that when dealing with classical data, training a parametrized data embedding with a kernel alignment method is also susceptible to exponential concentration. Our results are verified through numerical simulations for several QML tasks. Altogether, we provide guidelines indicating that certain features should be avoided to ensure the efficient evaluation and the trainability of quantum kernel methods.

A bill is headed to the governor's desk that would create a path for suing people who send unsolicited sexual pictures, but the legislation stops short of making "cyberflashing" a crime in California. If signed by Gov. Gavin Newsom, SB 53 by Sen. Connie Leyva (D-Chino) will allow Californians to take someone to civil court over unwanted lewd photos sent to them electronically; plaintiffs who win a suit could get up to $30,000 in damages. The legislation, approved Monday on the Senate floor in a 37-0 vote, comes after reports of men using the AirDrop iPhone feature to send lewd pictures to nearby strangers or on online dating apps without consent from the recipients. The bill applies to senders over 18 and defines obscene images as anything that depicts a person engaging in sexual acts, including masturbation, or photos of genitals "in a patently offensive way, and that, taken as a whole, lacks serious literary, artistic, political, or scientific value." The bill is sponsored by the women-centered dating app Bumble.

Can an artificial intelligence (AI) system be an inventor? Stephen Thaler recently submitted two patent applications for which an artificial intelligence system named "DABUS" was listed as the sole inventor. Specifically, the first application was directed to a food or beverage container that facilitates stacking.1 The second application was directed to a light device including a neural flame that serves as a signal beacon for human detection.2 The USPTO denied the patent applications for failing to list any human as an inventor.

The rapid pace of AI adoption in business could be heading for some major speed bumps. According to Gartner experts presenting at the Gartner CFO & Finance Executive Conference in June 2022, half of all AI deployments are expected to be postponed between now and 2024, as companies face barriers to upscaling AI in-house. AI governance and how enterprises are going to monitor and control the use of data in their AI platforms are emerging as significant snags. AI governance is a relatively new concept, as AI itself is still only in the early stages of development, but there are already complications emerging. For some companies, the governance of AI applications is included in data or model governance structures.