In February, Google announced that it was committing to training 100,000 Black women in digital skills. This announcement arrived as a PR Hail Mary amid the ever-growing industry and academic outcry over Google's firing of prominent, brilliant, respected A.I. researcher Timnit Gebru and recruiter April Christina Curley, both Black women and both exceptional contributors at the company. The backlash occurred during a year of widespread protest against the centuries-old violence of racism and racialized capitalism in the United States. This is not the first time that a prominent tech organization has attempted to "train up" Black Americans. From 1968 to 1972, at least 18 programs to provide computing skills training to Black and brown Americans were established in the United States. They were located in East Coast and California cities, with one in St. Louis, Missouri.

Welcome to our March 2021 monthly digest. Our digests are designed to keep you up-to-date with the latest happenings in the AI world. You can catch up with any AIhub stories you may have missed, get the low-down on recent conferences, and generally immerse yourself in all things AI. This month, our attention turned to education, and we considered both the use of AI in teaching, and the teaching of AI. Carles Sierra wrote about team formation techniques in education, describing how AI methods can be used to facilitate collaborative learning.

Online translation tools have helped us learn new languages, communicate across linguistic borders, and view foreign websites in our native tongue. But the artificial intelligence (AI) behind them is far from perfect, often replicating rather than rejecting the biases that exist within a language or a society. Such tools are especially vulnerable to gender stereotyping, because some languages (such as English) don't tend to gender nouns, while others (such as German) do. When translating from English to German, translation tools have to decide which gender to assign English words like "cleaner". Overwhelmingly, the tools conform to the stereotype, opting for the feminine word in German.

Created by Lazy Programmer Inc. Variational autoencoders and GANs have been 2 of the most interesting developments in deep learning and machine learning recently. Yann LeCun, a deep learning pioneer, has said that the most important development in recent years has been adversarial training, referring to GANs. GAN stands for generative adversarial network, where 2 neural networks compete with each other. Unsupervised learning means we're not trying to map input data to targets, we're just trying to learn the structure of that input data. Once we've learned that structure, we can do some pretty cool things.

The Delhi government plans to introduce the use of Artificial Intelligence (AI) and game-based techniques for continuous assessment of students under its new state school education board, officials said after the board's first general body meeting on Tuesday. The Delhi cabinet approved the Delhi Board of School Education (DBSE) on March 6. The society for the board was registered on March 19. The board is likely to be operational for up to at least the middle-school level (up to Class 8) from the upcoming academic session. Delhi education minister Manish Sisodia, who is the head of the board, said students will be assessed holistically on the basis of their knowledge, attitude, and skills.

We discuss why AI is hard and why physics is simple. We discuss how physical intuition and the approach of theoretical physics can be brought to bear on the field of artificial intelligence and specifically machine learning. We suggest that the underlying project of machine learning and the underlying project of physics are strongly coupled through the principle of sparsity, and we call upon theoretical physicists to work on AI as physicists. As a first step in that direction, we discuss an upcoming book on the principles of deep learning theory that attempts to realize this approach.

Kernel methods and neural networks are two of the most prevalent and versatile machine learning techniques. While various recent publications focus on invariant or equivariant deep learning algorithms, our goal is to derive kernel-based methods that exploit symmetries. Symmetries play an important role in many research areas such as physics and chemistry [1, 2, 3], but also point cloud classification problems [4] or problems defined on sets [5] are naturally permutation-invariant. One of the most prominent applications is in quantum physics. Systems of bosons require symmetric wave functions, whereas systems of fermions are represented by antisymmetric wave functions. Exploiting such symmetries of the underlying system is a popular and powerful approach that has been used to improve the performance of kernel-based methods as well as deep-learning algorithms.

Artificial Intelligence is one of the fastest growing technologies of the 21st century and accompanies us in our daily lives when interacting with technical applications. However, reliance on such technical systems is crucial for their widespread applicability and acceptance. The societal tools to express reliance are usually formalized by lawful regulations, i.e., standards, norms, accreditations, and certificates. Therefore, the T\"UV AUSTRIA Group in cooperation with the Institute for Machine Learning at the Johannes Kepler University Linz, proposes a certification process and an audit catalog for Machine Learning applications. We are convinced that our approach can serve as the foundation for the certification of applications that use Machine Learning and Deep Learning, the techniques that drive the current revolution in Artificial Intelligence. While certain high-risk areas, such as fully autonomous robots in workspaces shared with humans, are still some time away from certification, we aim to cover low-risk applications with our certification procedure. Our holistic approach attempts to analyze Machine Learning applications from multiple perspectives to evaluate and verify the aspects of secure software development, functional requirements, data quality, data protection, and ethics. Inspired by existing work, we introduce four criticality levels to map the criticality of a Machine Learning application regarding the impact of its decisions on people, environment, and organizations. Currently, the audit catalog can be applied to low-risk applications within the scope of supervised learning as commonly encountered in industry. Guided by field experience, scientific developments, and market demands, the audit catalog will be extended and modified accordingly.

Complex Event Recognition (CER) systems detect event occurrences in streaming time-stamped input using predefined event patterns. Logic-based approaches are of special interest in CER, since, via Statistical Relational AI, they combine uncertainty-resilient reasoning with time and change, with machine learning, thus alleviating the cost of manual event pattern authoring. We present a system based on Answer Set Programming (ASP), capable of probabilistic reasoning with complex event patterns in the form of weighted rules in the Event Calculus, whose structure and weights are learnt online. We compare our ASP-based implementation with a Markov Logic-based one and with a number of state-of-the-art batch learning algorithms on CER datasets for activity recognition, maritime surveillance and fleet management. Our results demonstrate the superiority of our novel approach, both in terms of efficiency and predictive performance. This paper is under consideration for publication in Theory and Practice of Logic Programming (TPLP).

Corrective interventions while a robot is learning to automate a task provide an intuitive method for a human supervisor to assist the robot and convey information about desired behavior. However, these interventions can impose significant burden on a human supervisor, as each intervention interrupts other work the human is doing, incurs latency with each context switch between supervisor and autonomous control, and requires time to perform. We present LazyDAgger, which extends the interactive imitation learning (IL) algorithm SafeDAgger to reduce context switches between supervisor and autonomous control. We find that LazyDAgger improves the performance and robustness of the learned policy during both learning and execution while limiting burden on the supervisor. Simulation experiments suggest that LazyDAgger can reduce context switches by an average of 60% over SafeDAgger on 3 continuous control tasks while maintaining state-of-the-art policy performance. In physical fabric manipulation experiments with an ABB YuMi robot, LazyDAgger reduces context switches by 60% while achieving a 60% higher success rate than SafeDAgger at execution time.