In this post I'll be talking about my experience so far of exploring AI. One of the main reasons for this newsletter is for me to learn more about the Artificial Intelligence world as a whole. By day I'm a web developer, and the aim is to move more towards AI work and projects. I want to get a better appreciation for what's currently out there, and what might be coming next, in order to build knowledge and understanding, whilst working on related practical skills. The hope is that the newsletter might also serve as a useful overview or gateway for other developers that are wanting to do the same.

Traditional computer scientists and engineers are trained to develop solutions for specific needs, but aren't always trained to consider their broader implications. Each new technology generation, and particularly the rise of artificial intelligence, leads to new kinds of systems, new ways of creating tools, and new forms of data, for which norms, rules, and laws frequently have yet to catch up. The kinds of impact that such innovations have in the world has often not been apparent until many years later. As part of the efforts in Social and Ethical Responsibilities of Computing (SERC) within the MIT Stephen A. Schwarzman College of Computing, a new case studies series examines social, ethical, and policy challenges of present-day efforts in computing with the aim of facilitating the development of responsible "habits of mind and action" for those who create and deploy computing technologies. "Advances in computing have undeniably changed much of how we live and work. Understanding and incorporating broader social context is becoming ever more critical," says Daniel Huttenlocher, dean of the MIT Schwarzman College of Computing.

Researchers from the University of Missouri and the University of North Carolina at Charlotte with image processing and cybersecurity expertise have been awarded nearly $1.2 million from the National Science Foundation to find out. They're designing an AI program they believe will need only a small number of deepfake examples to start to build its knowledge base. As it learns, the program will be able to spot new deepfake techniques, making more accurate detections and preventing mistakes in identifying content. Relying on a small number of examples overcomes the current challenges of algorithms that typically need a vast number of labeled samples to learn from. By leveraging accumulated knowledge, the deepfake detector will also learn to prevent camouflaged or obscured visual content from being classified as genuine content.

A three-day online international seminar on'Artificial intelligence in education' will begin on Monday. Being organised by the State Institute of Educational Technology, in association with the Department of Education of the University of Kerala, the webinar will discuss how artificial intelligence (AI) may have an impact on curriculum design, individualisation of learning, and assessment. The webinar will also discuss the ethical, technical, and pedagogical challenges against the backdrop of education in a digital world. Jintavee Khlisang, Professor in the Department of Educational Technology and Communication, Chulalongkorn University, Thailand, will speak on'Hybrid models of teaching and learning process,' while Kaushal Kumar Bhagat, Assistant Professor, Indian Institute of Technology, Kharagpur, will talk about augmented reality. Researchers will present papers on how artificial intelligence can be used in education.

Deep Reinforcement Learning (DRL) is an avenue of research in Artificial Intelligence (AI) that has received increasing attention within the research community in recent years, and is beginning to show potential for real-world application. DRL is one of the most promising routes towards developing more autonomous AI systems that interact with and take actions in complex real-world environments, and can more flexibly solve a range of problems for which we may not be able to precisely specify a correct ‘answer’. This could have substantial implications for people’s lives: for example by speeding up automation in various sectors, changing the nature and potential harms of online influence, or introducing new safety risks in physical infrastructure. In this paper, we review recent progress in DRL, discuss how this may introduce novel and pressing issues for society, ethics, and governance, and highlight important avenues for future research to better understand DRL’s societal implications. This article appears in the special track on AI and Society.

Though learning has become a core technology of modern information processing, there is now ample evidence that it can lead to biased, unsafe, and prejudiced solutions. The need to impose requirements on learning is therefore paramount, especially as it reaches critical applications in social, industrial, and medical domains. However, the non-convexity of most modern learning problems is only exacerbated by the introduction of constraints. Whereas good unconstrained solutions can often be learned using empirical risk minimization (ERM), even obtaining a model that satisfies statistical constraints can be challenging, all the more so a good one. In this paper, we overcome this issue by learning in the empirical dual domain, where constrained statistical learning problems become unconstrained, finite dimensional, and deterministic. We analyze the generalization properties of this approach by bounding the empirical duality gap, i.e., the difference between our approximate, tractable solution and the solution of the original (non-convex)~statistical problem, and provide a practical constrained learning algorithm. These results establish a constrained counterpart of classical learning theory and enable the explicit use of constraints in learning. We illustrate this algorithm and theory in rate-constrained learning applications.

Welcome to the fourth edition of the AI Index Report. This year we significantly expanded the amount of data available in the report, worked with a broader set of external organizations to calibrate our data, and deepened our connections with the Stanford Institute for Human-Centered Artificial Intelligence (HAI). The AI Index Report tracks, collates, distills, and visualizes data related to artificial intelligence. Its mission is to provide unbiased, rigorously vetted, and globally sourced data for policymakers, researchers, executives, journalists, and the general public to develop intuitions about the complex field of AI. The report aims to be the most credible and authoritative source for data and insights about AI in the world.

In this thesis, we develop various techniques for working with sets in machine learning. Each input or output is not an image or a sequence, but a set: an unordered collection of multiple objects, each object described by a feature vector. Their unordered nature makes them suitable for modeling a wide variety of data, ranging from objects in images to point clouds to graphs. Deep learning has recently shown great success on other types of structured data, so we aim to build the necessary structures for sets into deep neural networks. The first focus of this thesis is the learning of better set representations (sets as input). Existing approaches have bottlenecks that prevent them from properly modeling relations between objects within the set. To address this issue, we develop a variety of techniques for different scenarios and show that alleviating the bottleneck leads to consistent improvements across many experiments. The second focus of this thesis is the prediction of sets (sets as output). Current approaches do not take the unordered nature of sets into account properly. We determine that this results in a problem that causes discontinuity issues with many set prediction tasks and prevents them from learning some extremely simple datasets. To avoid this problem, we develop two models that properly take the structure of sets into account. Various experiments show that our set prediction techniques can significantly benefit over existing approaches.

Despite the rich theoretical foundation of model-based deep reinforcement learning (RL) agents, their effectiveness in real-world robotics-applications is less studied and understood. In this paper, we, therefore, investigate how such agents generalize to real-world autonomous-vehicle control-tasks, where advanced model-free deep RL algorithms fail. In particular, we set up a series of time-lap tasks for an F1TENTH racing robot, equipped with high-dimensional LiDAR sensors, on a set of test tracks with a gradual increase in their complexity. In this continuous-control setting, we show that model-based agents capable of learning in imagination, substantially outperform model-free agents with respect to performance, sample efficiency, successful task completion, and generalization. Moreover, we show that the generalization ability of model-based agents strongly depends on the observation-model choice. Finally, we provide extensive empirical evidence for the effectiveness of model-based agents provided with long enough memory horizons in sim2real tasks.

Sometimes, vocations and avocations need a champion, and students in Massachusetts looking to further their knowledge of science, technology and robotics have one in state Rep. Danillo Sena. A House member representing the 37th Middlesex District, Sena filed a bill on Feb. 4 titled "An Act establishing an elementary and secondary school robotics grant program," meant to create a grant program that provides public and charter schools the necessary funding to increase robotics and STEM participation during and after school. STEM stands for science, technology, engineering and mathematics, a branch of education designed to help students to become better problem-solvers. "Money should not be a barrier between students and access to fun and engaging STEM education programs that foster creativity and have lasting positive effects on student achievement like these robotics teams," the Acton Democrat stated in a release. The bill was created in collaboration with Olivia Oestreicher, a member of Team 4905 Andromeda One Robotics at Ayer Shirley Regional High School and a Rep. Sena intern.