Interest and investment in AI for Education is accelerating. So is concern about the issues that will arise when AI is widely implemented in educational technologies -- such as bias, fairness, and data security. With our team at the Center for Integrative Research in the Computing and Learning Sciences (CIRCLS), we see that organizations around the world--like UNESCO or the new EdSafe AI Alliance--are organizing people to tackle the issues. In the US, organizations like Stanford's HAI are addressing the issues of AI in healthcare, but not so much in education. Over the past year, my colleagues organized a working group in AI and education policy.

Amira Learning's award-winning app brings literacy to life for younger learners Education, like many sectors across the globe, has found an increasing need to develop technologies that support a new'normal' following the initial and transformational impacts of Covid-19 on teaching and learning. Many learning challenges for students remain the same even if the landscape has fundamentally changed from traditional brick-and-mortar schools to digital classrooms and e-learning experiences. The time, for EdTech, to answer has come faster than the sector might have previously forecasted and all eyes are on the results of technology investments that are outpacing years prior. While advancements in artificial intelligence (AI) have saturated our digital experiences and impacted consumer behaviors, there is still a struggle to see tangible AI applications in the day-to-day of teaching and learning. Maybe that horizon is closer than previously expected.

It is no longer a question of if artificial intelligence (AI) will play a role in healthcare delivery and diagnostics. Instead, the question is how the technology can ethically be deployed to fill this role, and what guidelines should be put in place today, to support upstream thinking around challenges that will arise tomorrow. AI still feels a tad like the "Wild West," with both limited and splintered policies and laws around regulating and managing the use of the technology, especially in the healthcare arena. However, ahead of the curve, the World Health Organization (WHO) is already prompting leaders to consider ethical responsibilities and dilemmas of integrating AI more heavily into care, posing questions around maintaining human autonomy, ensuring transparency, and establishing inclusiveness and equity– just to name a few! While this is a tall order, it is one that governing bodies and leaders will need to fill.

Putting a human base on Mars has long been an aspiration for us Earthlings. For centuries people have been fascinated by the dusty and desolate Red Planet, ever since its discovery by Galileo in 1610, and as the mystic grew, so too did thoughts that little green men might be running around on it, an image popularised in magazines and on TV during the 1950s and 60s. Moon race fever had gripped the world, and after Neil Armstrong became the first man to walk on the lunar surface in 1969, the human race almost immediately set its sights on the next target of Mars. Though manned space exploration beyond Earth's orbit stalled in the decades that followed, a new space race involving billionaires Elon Musk, Jeff Bezos and Sir Richard Branson has reignited ideas for people to set foot on other worlds beyond our own. Chief among them is SpaceX CEO Musk, who has stated his desire to create a colony of one million people on Mars by 2050.

If human societies do nothing, in just a few decades, the planet could warm to levels it hasn't reached in at least 34 million years, leading to more melting glaciers and floods than ever before -- as well as the dire effect of urban heat waves. In 2021, in the U.S. alone, there were already 18 extreme climate-related disasters with losses exceeding $1 billion each, according to the National Oceanic and Atmospheric Administration. When looking at the world's natural calamities on a consequence and frequency scale, floods and earthquakes have a more devastating effect on people and property, but they occur less frequently than heat waves, which generally take the form of urban heat islands (UHIs). These are also known as heat pockets, which are found across cities' downtown areas, where temperatures are higher than the peripheries. With urbanized areas warming up fast, many more populations globally are bound to face the deadly consequences of the heat-island effect, highlighting urban public health disparities.

Don Fancher is a Deloitte Risk & Financial Advisory Principal with Deloitte Financial Advisory Services LLP where he serves as the Global Leader of Deloitte Forensic as well as the Co-Leader of Deloitte's Legal Business Services practice. Mr. Fancher has over 30 years of experience assisting clients and leading practices in forensic, dispute consulting and legal transformation. He currently leads over 4,500 Deloitte professionals around the world serving clients in areas such as financial crime, disputes and investigations, business insurance, discovery, data governance, legal transformation, and contract lifecycle management. Mr. Fancher has significant experience assisting clients and counsel in performing forensic investigations and special reviews for matters regarding financial crime, misappropriation of assets, breach of fiduciary duty, and FCPA violations. These have included both individual employee and institution-wide schemes for misappropriating funds and/or improperly reporting asset values and financial performance.

Partial differential equations (PDEs) play a dominant role in the mathematical modeling of many complex dynamical processes. Solving these PDEs often requires prohibitively high computational costs, especially when multiple evaluations must be made for different parameters or conditions. After training, neural operators can provide PDEs solutions significantly faster than traditional PDE solvers. In this work, invariance properties and computational complexity of two neural operators are examined for transport PDE of a scalar quantity. Neural operator based on graph kernel network (GKN) operates on graph-structured data to incorporate nonlocal dependencies. Here we propose a modified formulation of GKN to achieve frame invariance. Vector cloud neural network (VCNN) is an alternate neural operator with embedded frame invariance which operates on point cloud data. GKN-based neural operator demonstrates slightly better predictive performance compared to VCNN. However, GKN requires an excessively high computational cost that increases quadratically with the increasing number of discretized objects as compared to a linear increase for VCNN.

Many tasks in Natural Language Processing (NLP) and Computer Vision (CV) offer evidence that humans disagree, from objective tasks such as part-of-speech tagging to more subjective tasks such as classifying an image or deciding whether a proposition follows from certain premises. While most learning in artificial intelligence (AI) still relies on the assumption that a single (gold) interpretation exists for each item, a growing body of research aims to develop learning methods that do not rely on this assumption. In this survey, we review the evidence for disagreements on NLP and CV tasks, focusing on tasks for which substantial datasets containing this information have been created. We discuss the most popular approaches to training models from datasets containing multiple judgments potentially in disagreement. We systematically compare these different approaches by training them with each of the available datasets, considering several ways to evaluate the resulting models. Finally, we discuss the results in depth, focusing on four key research questions, and assess how the type of evaluation and the characteristics of a dataset determine the answers to these questions. Our results suggest, first of all, that even if we abandon the assumption of a gold standard, it is still essential to reach a consensus on how to evaluate models. This is because the relative performance of the various training methods is critically affected by the chosen form of evaluation. Secondly, we observed a strong dataset effect. With substantial datasets, providing many judgments by high-quality coders for each item, training directly with soft labels achieved better results than training from aggregated or even gold labels. This result holds for both hard and soft evaluation. But when the above conditions do not hold, leveraging both gold and soft labels generally achieved the best results in the hard evaluation. All datasets and models employed in this paper are freely available as supplementary materials.

Solar energy is now the cheapest form of electricity in history. Unfortunately, significantly increasing the grid's fraction of solar energy remains challenging due to its variability, which makes balancing electricity's supply and demand more difficult. While thermal generators' ramp rate -- the maximum rate that they can change their output -- is finite, solar's ramp rate is essentially infinite. Thus, accurate near-term solar forecasting, or nowcasting, is important to provide advance warning to adjust thermal generator output in response to solar variations to ensure a balanced supply and demand. To address the problem, this paper develops a general model for solar nowcasting from abundant and readily available multispectral satellite data using self-supervised learning. Specifically, we develop deep auto-regressive models using convolutional neural networks (CNN) and long short-term memory networks (LSTM) that are globally trained across multiple locations to predict raw future observations of the spatio-temporal data collected by the recently launched GOES-R series of satellites. Our model estimates a location's future solar irradiance based on satellite observations, which we feed to a regression model trained on smaller site-specific solar data to provide near-term solar photovoltaic (PV) forecasts that account for site-specific characteristics. We evaluate our approach for different coverage areas and forecast horizons across 25 solar sites and show that our approach yields errors close to that of a model using ground-truth observations.

Here's What You Need to Know: SOCOM has been a pioneer in the application of Artificial Intelligence. For example, the Naval Special Warfare Command (WARCOM) has been using Artificial Intelligence to find ways to make its SEAL and Special Warfare Combatant-Craft Crewmen (SWCC) operators more combat effective, while the Marines Special Operations Command (MARSOC) has been implementing Artificial Intelligence to better select and assess its future cadre of Marine Raiders. Artificial Intelligence is everywhere today. From Amazon's Alexa to unmanned aerial vehicles to space crafts to health care; Artificial Intelligence enables faster and better decisions if employed properly. The U.S. military and intelligence community have been using this technology for decades now and it's becoming increasingly more prevalent.