The 18th International Conference for Artificial Intelligence and Law (ICAIL 2021) was organized at the University of São Paulo School of Law, Brazil. ICAIL is a biannual conference organized under the auspices of the International Association for Artificial Intelligence and Law (iaail.org) For the first time, the ICAIL conference was organized entirely online, due to the overall Covid-19 pandemic situation. Despite these unusual circumstances, the conference came out as a considerable success, attracting almost 1400 registered participants, the highest number ever. The conference talks were streamed publicly on the YouTube channel and the discussions and networking were enabled on the platforms accessible for the registered participants.

Your organization knows it, and so do your customers; they engage with your business only when you clearly communicate how you can help them. Getting your signal through the noise has never been more challenging. Businesses are sending out an estimated seven trillion messages a month, most of them ignored--or worse, actively avoided; the customers who receive them may cut off contact entirely. Why are so few messages hitting the mark? Because it's difficult to tailor marketing campaigns and messages to thousands or millions of customers at once while incorporating the complex data, channels, and other inputs that give these messages the authenticity that drives outcomes. With cookies and other third-party customer data on the way out, companies need to make better use of their proprietary data and get a competitive advantage from their content.

In July of last year, a DJI Mavic 2 drone approached a Pennsylvania power substation. Two 4-foot nylon ropes dangled from its rotors, a thick copper wire connected to the ends with electrical tape. The device had been stripped of any identifiable markings, as well as its onboard camera and memory card, in an apparent effort by its owner to avoid detection. Its likely goal, according to a joint security bulletin released by DHS, the FBI, and the National Counterterrorism Center, was to "disrupt operations by creating a short circuit." The drone crashed on the roof of an adjacent building before it reached its ostensible target, damaging a rotor in the process.

There's a massive announcement set to take place later this year, and it could change the $12 trillion healthcare industry forever. Over the last 2 years, we've seen a huge transformation as businesses across nearly every industry have gone digital. And with the health crisis sweeping the globe last year, the healthcare sector was no different. Mentioned in today's commentary includes: Brookfield Renewable Partners L.P. (NYSE: BEP), LifeStance Health Group, Inc. (NASDAQ: LFST), Teladoc Health, Inc. (NYSE: TDOC), Mind Medicine (MindMed) Inc. (NASDAQ: MNMD), American Well Corporation (NYSE: AMWL). That's why we're now on the verge of a healthcare revolution, set to leverage the latest technology to disrupt a bloated and complicated system.

The legal framework for the development and use of Artificial Intelligence (AI) was launched in Brazil. There was already a lot about regulation in full progress in the European Union and the United States, but here the matter was delayed. The landscape is starting to change. In late September, the Chamber of Deputies approved a bill that establishes the foundations and principles for AI in the country. The framework defines the rules of the game for AI development and outlines how it will be applied to the daily lives of companies and citizens. In line with the General Data Protection Law (LGPD), this regulation serves as a guide for the work of Brazilian companies in the startup and research and development (R&D) sectors in this important area of innovation.

Source separation models either work on the spectrogram or waveform domain. In this work, we show how to perform end-to-end hybrid source separation, letting the model decide which domain is best suited for each source, and even combining both. The proposed hybrid version of the Demucs architecture won the Music Demixing Challenge 2021 organized by Sony. This architecture also comes with additional improvements, such as compressed residual branches, local attention or singular value regularization. Overall, a 1.4 dB improvement of the Signal-To-Distortion (SDR) was observed across all sources as measured on the MusDB HQ dataset, an improvement confirmed by human subjective evaluation, with an overall quality rated at 2.83 out of 5 (2.36 for the non hybrid Demucs), and absence of contamination at 3.04 (against 2.37 for the non hybrid Demucs and 2.44 for the second ranking model submitted at the competition).

The Wasserstein distance, rooted in optimal transport (OT) theory, is a popular discrepancy measure between probability distributions with various applications to statistics and machine learning. Despite their rich structure and demonstrated utility, Wasserstein distances are sensitive to outliers in the considered distributions, which hinders applicability in practice. Inspired by the Huber contamination model, we propose a new outlier-robust Wasserstein distance $\mathsf{W}_p^\varepsilon$ which allows for $\varepsilon$ outlier mass to be removed from each contaminated distribution. Our formulation amounts to a highly regular optimization problem that lends itself better for analysis compared to previously considered frameworks. Leveraging this, we conduct a thorough theoretical study of $\mathsf{W}_p^\varepsilon$, encompassing characterization of optimal perturbations, regularity, duality, and statistical estimation and robustness results. In particular, by decoupling the optimization variables, we arrive at a simple dual form for $\mathsf{W}_p^\varepsilon$ that can be implemented via an elementary modification to standard, duality-based OT solvers. We illustrate the benefits of our framework via applications to generative modeling with contaminated datasets.

Machine learning on source code (MLOnCode) is a popular research field that has been driven by the availability of large-scale code repositories and the development of powerful probabilistic and deep learning models for mining source code. Code-to-code recommendation is a task in MLOnCode that aims to recommend relevant, diverse and concise code snippets that usefully extend the code currently being written by a developer in their development environment (IDE). Code-to-code recommendation engines hold the promise of increasing developer productivity by reducing context switching from the IDE and increasing code-reuse. Existing code-to-code recommendation engines do not scale gracefully to large codebases, exhibiting a linear growth in query time as the code repository increases in size. In addition, existing code-to-code recommendation engines fail to account for the global statistics of code repositories in the ranking function, such as the distribution of code snippet lengths, leading to sub-optimal retrieval results. We address both of these weaknesses with \emph{Senatus}, a new code-to-code recommendation engine. At the core of Senatus is \emph{De-Skew} LSH a new locality sensitive hashing (LSH) algorithm that indexes the data for fast (sub-linear time) retrieval while also counteracting the skewness in the snippet length distribution using novel abstract syntax tree-based feature scoring and selection algorithms. We evaluate Senatus via automatic evaluation and with an expert developer user study and find the recommendations to be of higher quality than competing baselines, while achieving faster search. For example, on the CodeSearchNet dataset we show that Senatus improves performance by 6.7\% F1 and query time 16x is faster compared to Facebook Aroma on the task of code-to-code recommendation.

Learning from class imbalanced datasets poses challenges for many machine learning algorithms. Many real-world domains are, by definition, class imbalanced by virtue of having a majority class that naturally has many more instances than its minority class (e.g. genuine bank transactions occur much more often than fraudulent ones). Many methods have been proposed to solve the class imbalance problem, among the most popular being oversampling techniques (such as SMOTE). These methods generate synthetic instances in the minority class, to balance the dataset, performing data augmentations that improve the performance of predictive machine learning (ML) models. In this paper we advance a novel data augmentation method (adapted from eXplainable AI), that generates synthetic, counterfactual instances in the minority class. Unlike other oversampling techniques, this method adaptively combines exist-ing instances from the dataset, using actual feature-values rather than interpolating values between instances. Several experiments using four different classifiers and 25 datasets are reported, which show that this Counterfactual Augmentation method (CFA) generates useful synthetic data points in the minority class. The experiments also show that CFA is competitive with many other oversampling methods many of which are variants of SMOTE. The basis for CFAs performance is discussed, along with the conditions under which it is likely to perform better or worse in future tests.

US activist raises concerns over racial bias in AI Urges software developers to listen more to Black people Wrongful arrest in US caused by faulty facial recognition Lisbon playing host to Europe's largest tech event LISBON, Nov 3 (Reuters) - (This Nov. 3 story corrects to change BLM co-founder first name to Ayo from Opal at her request) As concerns grow over racial bias in artificial intelligence, Black Lives Matter (BLM) co-founder Ayo Tometi urged the tech sector to act fast against perpetuating racism in systems such as facial recognition. Artificial intelligence is transforming the world and can be applied in diverse sectors, from improving the early detection of diseases to sorting out data and solving complex problems, but there are also concerns around it. "A lot of the algorithms, a lot of the data is racist," U.S. activist Tometi, who co-founded BLM in 2013, told Reuters on the sidelines of Lisbon's Web Summit. "We need tech to truly understand every way it (racism) shows up in the technologies they are developing," she said. The tech industry has faced a reckoning over the past few years over the ethics of AI technologies, with critics saying such systems could compromise privacy, target marginalised groups and normalise intrusive surveillance.