Natural language communication is core to intelligence, as it allows ideas to be efficiently shared between humans or artificially intelligent systems. The generality of language allows us to express many intelligence tasks as taking in natural language input and producing natural language output. Autoregressive language modelling -- predicting the future of a text sequence from its past -- provides a simple yet powerful objective that admits formulation of numerous cognitive tasks. At the same time, it opens the door to plentiful training data: the internet, books, articles, code, and other writing. However this training objective is only an approximation to any specific goal or application, since we predict everything in the sequence rather than only the aspects we care about. Yet if we treat the resulting models with appropriate caution, we believe they will be a powerful tool to capture some of the richness of human intelligence. Using language models as an ingredient towards intelligence contrasts with their original application: transferring text over a limited-bandwidth communication channel. Shannon's Mathematical Theory of Communication (Shannon, 1948) linked the statistical modelling of natural language with compression, showing that measuring the cross entropy of a language model is equivalent to measuring its compression rate.

This paper aims to help structure the risk landscape associated with large-scale Language Models (LMs). In order to foster advances in responsible innovation, an in-depth understanding of the potential risks posed by these models is needed. A wide range of established and anticipated risks are analysed in detail, drawing on multidisciplinary expertise and literature from computer science, linguistics, and social sciences. We outline six specific risk areas: I. Discrimination, Exclusion and Toxicity, II. Information Hazards, III. Misinformation Harms, V. Malicious Uses, V. Human-Computer Interaction Harms, VI. Automation, Access, and Environmental Harms. The first area concerns the perpetuation of stereotypes, unfair discrimination, exclusionary norms, toxic language, and lower performance by social group for LMs. The second focuses on risks from private data leaks or LMs correctly inferring sensitive information. The third addresses risks arising from poor, false or misleading information including in sensitive domains, and knock-on risks such as the erosion of trust in shared information. The fourth considers risks from actors who try to use LMs to cause harm. The fifth focuses on risks specific to LLMs used to underpin conversational agents that interact with human users, including unsafe use, manipulation or deception. The sixth discusses the risk of environmental harm, job automation, and other challenges that may have a disparate effect on different social groups or communities. In total, we review 21 risks in-depth. We discuss the points of origin of different risks and point to potential mitigation approaches. Lastly, we discuss organisational responsibilities in implementing mitigations, and the role of collaboration and participation. We highlight directions for further research, particularly on expanding the toolkit for assessing and evaluating the outlined risks in LMs.

Hand-encoding PDDL domains is generally accepted as difficult, tedious and error-prone. The difficulty is even greater when temporal domains have to be encoded. Indeed, actions have a duration and their effects are not instantaneous. In this paper, we present TempAMLSI, an algorithm based on the AMLSI approach able to learn temporal domains. TempAMLSI is based on the classical assumption done in temporal planning that it is possible to convert a non-temporal domain into a temporal domain. TempAMLSI is the first approach able to learn temporal domain with single hard envelope and Cushing's intervals. We show experimentally that TempAMLSI is able to learn accurate temporal domains, i.e., temporal domain that can be used directly to solve new planning problem, with different forms of action concurrency.

A new artificial intelligence (AI) system developed by researchers at the University of Waterloo and DarwinAI, an alumni-founded startup company, could help doctors efficiently utilize limited resources during the COVID-19 pandemic. The system is able to identify patients who require intensive care unit (ICU) treatment. The AI system predicts this necessity of ICU admission through the use of 200 clinical data points, which include blood test results, medical history, and vital signs. Alexander Wong is a professor of systems design engineering and Canada Research Chair in AI and Medical Imaging at Waterloo. "That is a very important step in the clinical decision support process for triaging patients and developing treatment plans," Wong said.

Machine learning plays a crucial role in enhancing and accelerating the search for new fundamental physics. We review the state of machine learning methods and applications for new physics searches in the context of terrestrial high energy physics experiments, including the Large Hadron Collider, rare event searches, and neutrino experiments. While machine learning has a long history in these fields, the deep learning revolution (early 2010s) has yielded a qualitative shift in terms of the scope and ambition of research. These modern machine learning developments are the focus of the present review.

The multi-task learning (MTL) paradigm can be traced back to an early paper of Caruana (1997) in which it was argued that data from multiple tasks can be used with the aim to obtain a better performance over learning each task independently. A solution of MTL with conflicting objectives requires modelling the trade-off among them which is generally beyond what a straight linear combination can achieve. A theoretically principled and computationally effective strategy is finding solutions which are not dominated by others as it is addressed in the Pareto analysis. Multi-objective optimization problems arising in the multi-task learning context have specific features and require adhoc methods. The analysis of these features and the proposal of a new computational approach represent the focus of this work. Multi-objective evolutionary algorithms (MOEAs) can easily include the concept of dominance and therefore the Pareto analysis. The major drawback of MOEAs is a low sample efficiency with respect to function evaluations. The key reason for this drawback is that most of the evolutionary approaches do not use models for approximating the objective function. Bayesian Optimization takes a radically different approach based on a surrogate model, such as a Gaussian Process. In this thesis the solutions in the Input Space are represented as probability distributions encapsulating the knowledge contained in the function evaluations. In this space of probability distributions, endowed with the metric given by the Wasserstein distance, a new algorithm MOEA/WST can be designed in which the model is not directly on the objective function but in an intermediate Information Space where the objects from the input space are mapped into histograms. Computational results show that the sample efficiency and the quality of the Pareto set provided by MOEA/WST are significantly better than in the standard MOEA.

Weak supervision (WS) frameworks are a popular way to bypass hand-labeling large datasets for training data-hungry models. These approaches synthesize multiple noisy but cheaply-acquired estimates of labels into a set of high-quality pseudolabels for downstream training. However, the synthesis technique is specific to a particular kind of label, such as binary labels or sequences, and each new label type requires manually designing a new synthesis algorithm. Instead, we propose a universal technique that enables weak supervision over any label type while still offering desirable properties, including practical flexibility, computational efficiency, and theoretical guarantees. We apply this technique to important problems previously not tackled by WS frameworks including learning to rank, regression, and learning in hyperbolic manifolds. Theoretically, our synthesis approach produces a consistent estimator for learning a challenging but important generalization of the exponential family model. Experimentally, we validate our framework and show improvement over baselines in diverse settings including real-world learning-to-rank and regression problems along with learning on hyperbolic manifolds.

Effective water resource management requires information on water availability, both in terms of quality and quantity, spatially and temporally. In this paper, we study the methodology behind Transfer Learning (TL) through fine-tuning and parameter transferring for better generalization performance of streamflow prediction in data-sparse regions. We propose a standard recurrent neural network in the form of Long Short-Term Memory (LSTM) to fit on a sufficiently large source domain dataset and repurpose the learned weights to a significantly smaller, yet similar target domain datasets. We present a methodology to implement transfer learning approaches for spatiotemporal applications by separating the spatial and temporal components of the model and training the model to generalize based on categorical datasets representing spatial variability. The framework is developed on a rich benchmark dataset from the US and evaluated on a smaller dataset collected by The Nature Conservancy in Kenya. The LSTM model exhibits generalization performance through our TL technique. Results from this current experiment demonstrate the effective predictive skill of forecasting streamflow responses when knowledge transferring and static descriptors are used to improve hydrologic model generalization in data-sparse regions.

Testing practices within the machine learning (ML) community have centered around assessing a learned model's predictive performance measured against a test dataset, often drawn from the same distribution as the training dataset. While recent work on robustness and fairness testing within the ML community has pointed to the importance of testing against distributional shifts, these efforts also focus on estimating the likelihood of the model making an error against a reference dataset/distribution. We argue that this view of testing actively discourages researchers and developers from looking into other sources of robustness failures, for instance corner cases which may have severe undesirable impacts. We draw parallels with decades of work within software engineering testing focused on assessing a software system against various stress conditions, including corner cases, as opposed to solely focusing on average-case behaviour. Finally, we put forth a set of recommendations to broaden the view of machine learning testing to a rigorous practice.

According to a newly released survey from nonprofit technical organization IEEE, about one in five respondents say AI and machine learning (21%), cloud computing (20%), and 5G (17%) will be the most important technologies next year. The study examines the most important technologies in 2022, the industries expected to be most impacted by technology in the year ahead, and anticipated technology trends through the next decade. What industries are expected to be most impacted by technology in the year ahead? Technology leaders surveyed cited manufacturing (25%), financial services (19%), health care (16%), and energy (13%) as industries poised for major disruption. Regarding the key technology trends to expect through the next decade, an overwhelming majority (95%) agree -- including 66% who strongly agree -- that AI will drive the majority of innovation across nearly every industry sector in the next one to five years.