Welcome to our February 2022 monthly digest, where you can catch up with any AIhub stories you may have missed, get the low-down on recent events, and much more. This month, we cover our latest New voices in AI interview, hear from a NeurIPS award winner, and get stuck into AAAI 2022. You may have seen the launch of our new series last month. In the latest episode, Isabel Cachola talks about how she got into AI and her work on interpretability of NLP models. In this interview, Bernard Koch tells us about research that won him, and co-authors Emily Denton, Alex Hanna and Jacob Foster, a best paper prize at NeurIPS 2021.

Organized crime is very active throughout both the private and public sectors. From banks to major retailers, federal governments to small businesses, fraudsters aren't discerning when choosing which type of organization to target. What's more, crime rings are actively recruiting and deploying groups of thieves across the U.S., making it increasingly difficult for companies to catch bad actors before they strike. Overwhelmed by unmanageable amounts of data, companies have left gaps in their IT systems that criminals are eager to exploit. It's well past time for businesses and governments to take steps to thwart nefarious actors by using technology that outpaces them, keeps customer data safe, and protects organizations across the board.

If you are a PhD student or ML/AI/data researcher impacted by current Russian aggression, let me know, we can arrange a paid research internship (up to one year) in Prague for you.

America is getting older faster. According to the U.S. Census Bureau, the number of people aged 65 or older in the United States will grow to 95 million by the year 2060 and will account for nearly one-quarter of the population. Artificial intelligence (AI) and technology solutions have a significant potential to transform quality of life and improve health care outcomes for older Americans, including those with Alzheimer's Disease and Related Dementias (AD/ADRD). To meet this challenge, the AI/Tech Aging (a2) Collective is announcing the a2 Pilot Awards, a national competition that will earmark $40 million over the next 5 years for promising pilot projects that leverage AI and other transformative technology to support healthy aging and persons living with AD/ADRD. The a2 Collective represents the National Institute on Aging's (NIA) Artificial Intelligence and Technology Collaboratories for Aging Research (AITC) program, which is dedicated to helping Americans live longer, better through the application of AI and emerging technologies.

Ukrainian military official releases video showing what appears to be Russian missile system being destroyed by a drone. A Ukrainian official released footage on Sunday, appearing to show a drone taking out a Russian missile system. The Russian invasion of Ukraine continued for a fifth day on Monday, Feb. 28. Ukrainian forces have offered stiff resistance as Russia launched attacks from Crimea in the south, an attack on the second-largest city of Kharkiv, and an attack on the Ukrainian capital city of Kyiv. The United States and its allies have responded with stiff sanctions, including cutting off some Russian banks from the international payment system Swift.

MadJax is a tool for generating and evaluating differentiable matrix elements of high energy scattering processes. As such, it is a step towards a differentiable programming paradigm in high energy physics that facilitates the incorporation of high energy physics domain knowledge, encoded in simulation software, into gradient based learning and optimization pipelines. MadJax comprises two components: (a) a plugin to the general purpose matrix element generator MadGraph that integrates matrix element and phase space sampling code with the JAX differentiable programming framework, and (b) a standalone wrapping API for accessing the matrix element code and its gradients, which are computed with automatic differentiation. The MadJax implementation and example applications of simulation based inference and normalizing flow based matrix element modeling, with capabilities enabled uniquely with differentiable matrix elements, are presented.

We propose a framework for online meta-optimization of parameters that govern optimization, called Amortized Proximal Optimization (APO). We first interpret various existing neural network optimizers as approximate stochastic proximal point methods which trade off the current-batch loss with proximity terms in both function space and weight space. The idea behind APO is to amortize the minimization of the proximal point objective by meta-learning the parameters of an update rule. We show how APO can be used to adapt a learning rate or a structured preconditioning matrix. Under appropriate assumptions, APO can recover existing optimizers such as natural gradient descent and KFAC. It enjoys low computational overhead and avoids expensive and numerically sensitive operations required by some second-order optimizers, such as matrix inverses. We empirically test APO for online adaptation of learning rates and structured preconditioning matrices for regression, image reconstruction, image classification, and natural language translation tasks. Empirically, the learning rate schedules found by APO generally outperform optimal fixed learning rates and are competitive with manually tuned decay schedules. Using APO to adapt a structured preconditioning matrix generally results in optimization performance competitive with second-order methods. Moreover, the absence of matrix inversion provides numerical stability, making it effective for low precision training.

Artificial intelligence is revolutionary, but it's not without its controversies. Some believe it can take us down a dangerous path, potentially arming governments with dangerous Orwellian surveillance and mass control capabilities. We have to remember that any technology is only as'good' or'bad' as the people who use it. Consider the EU's hailed'blueprint for AI regulation' and China's proposed crackdown on AI development; these instances seek to regulate AI as if it were already an autonomous, conscious technology. The U.S. must think wisely before following in their footsteps and consider addressing the actions of the user behind the AI.

I love constructing knowledge graphs from various sources. I've wanted to create a government knowledge graph for some time now but was struggling to find any data that is easily accessible and doesn't require me to spend weeks developing a data pipeline. At first, I thought I would have to use OCR and NLP techniques to extract valuable information from public records, but luckily I stumbled upon UK Gazette. The UK Gazette is a website that holds the United Kingdom's official public record information. All the content on the website and via its APIs is available under the Open Government License v3.0.

February may be the shortest and coldest month of the year. But for many, it is a time to give special recognition to often overlooked aspects of world history (Black History Month) and recognize what may be the single greatest threat to health in the world. For many, February is also known as Heart Health Month, and 2022 will be the 58th consecutive year it is recognized. Cardiovascular disease is a global problem that claims the lives of more people a year than cancer, strokes, or other prevalent diseases. Luckily, advanced research is leading to effective solutions for improving cardiovascular health.