A federal prosecutor alleged that one defendant boasted that his father "had engaged in similar business for the Chinese Communist Party." US authorities allege four people based in Florida, Alabama, and California conspired to illegally ship supercomputers and hundreds of Nvidia GPUs to China as recently as July. The charges, which were unsealed in federal court on Wednesday, are part of a wider government effort to crack down on the smuggling of advanced AI chips to China. Over the past few years, the US has introduced a series of export control rules designed to prevent Chinese organizations from acquiring computer chips that have become popular for developing AI chatbots . The restrictions aim to slow China in what US officials have described as a race to develop powerful AI systems, including surveillance tools and autonomous weapons .

We explore the use of Artificial Potential Fields (APFs) to solve Multi-Agent Path Finding (MAPF) and Lifelong MAPF (LMAPF) problems. In MAPF, a team of agents must move to their goal locations without collisions, whereas in LMAPF, new goals are generated upon arrival. We propose methods for incorporating APFs in a range of MAPF algorithms, including Prioritized Planning, MAPF-LNS2, and Priority Inheritance with Backtracking (PIBT). Experimental results show that using APF is not beneficial for MAPF but yields up to a 7-fold increase in overall system throughput for LMAPF.

The huge amount of energy required to train artificial intelligence (AI) is becoming a concern. To train the large language model (LLM) powering Chat GPT-3, for example, almost 1,300 megawatt hours of energy was used, according to an estimate by researchers from Google and the University of California, Berkeley, a similar quantity of energy to what is used by 130 American homes in one year. Furthermore, an analysis by OpenAI suggests that the amount of power needed to train AI models has been growing exponentially since 2012, doubling roughly every 3.4 months as the models become bigger and more sophisticated. However, our energy production capacity is not increasing as steeply, and doing so is likely to further contribute to global warming: generating electricity is the single biggest contributor to climate change given that coal, oil, and gas are still widely used to generate electricity, compared to cleaner energy sources. "At this rate, we are running into a brick wall in terms of the ability to scale up machine learning networks," said Menachem Stern, a theoretical physicist at the AMOLF research institute in the Netherlands.

This workshop will grow and consolidate a community of interdisciplinary CSCW researchers focusing on the topic of contestable AI. As an outcome of the workshop, we will synthesize the most pressing opportunities and challenges for contestability along AI value chains in the form of a research roadmap. This roadmap will help shape and inspire imminent work in this field. Considering the length and depth of AI value chains, it will especially spur discussions around the contestability of AI systems along various sites of such chains. The workshop will serve as a platform for dialogue and demonstrations of concrete, successful, and unsuccessful examples of AI systems that (could or should) have been contested, to identify requirements, obstacles, and opportunities for designing and deploying contestable AI in various contexts. This will be held primarily as an in-person workshop, with some hybrid accommodation. The day will consist of individual presentations and group activities to stimulate ideation and inspire broad reflections on the field of contestable AI. Our aim is to facilitate interdisciplinary dialogue by bringing together researchers, practitioners, and stakeholders to foster the design and deployment of contestable AI.

As artificial neural networks grow in complexity, understanding their inner workings becomes increasingly challenging, which is particularly important in healthcare applications. The intrinsic evaluation metrics of autoregressive neural language models (NLMs), perplexity (PPL), can reflect how "surprised" an NLM model is at novel input. PPL has been widely used to understand the behavior of NLMs. Previous findings show that changes in PPL when masking attention layers in pre-trained transformer-based NLMs reflect linguistic anomalies associated with Alzheimer's disease dementia. Building upon this, we explore a novel bidirectional attention head ablation method that exhibits properties attributed to the concepts of cognitive and brain reserve in human brain studies, which postulate that people with more neurons in the brain and more efficient processing are more resilient to neurodegeneration. Our results show that larger GPT-2 models require a disproportionately larger share of attention heads to be masked/ablated to display degradation of similar magnitude to masking in smaller models. These results suggest that the attention mechanism in transformer models may present an analogue to the notions of cognitive and brain reserve and could potentially be used to model certain aspects of the progression of neurodegenerative disorders and aging.

Corny dialogue has been part of video games almost since they have existed. From 1989's Zero Wing spawning the decades old "All your base are belong to us" internet meme, to the clunky translations of the pre-remake Resident Evil games ("the master of unlocking"), to Skyrim's infamous adventurer who once took an arrow to the knee and never shuts up about it, non-playable character (NPC) dialogue has rarely been exactly Shakespearean, and the frequent repetition doesn't help. But could AI tools change that, enabling a world full of characters that respond believably when you talk to them? In collaboration with Google, a team of researchers from Stanford have built a game demo called Smallville that integrates the AI writing tool ChatGPT. Instead of just walking into walls and setting themselves on fire like the classic Sims characters we all knew and loved, the game's 25 characters can instead comfortably discuss topics such as local politics and composing music, pulling from ChatGPT's enormous database.

We map the surface of Pluto using an unsupervised machine learning technique using the near-infrared observations of the LEISA/Ralph instrument onboard NASA's New Horizons spacecraft. The principal component reduced Gaussian mixture model was implemented to investigate the geographic distribution of the surface units across the dwarf planet. We also present the likelihood of each surface unit at the image pixel level. Average I/F spectra of each unit were analyzed -- in terms of the position and strengths of absorption bands of abundant volatiles such as N${}_{2}$, CH${}_{4}$, and CO and nonvolatile H${}_{2}$O -- to connect the unit to surface composition, geology, and geographic location. The distribution of surface units shows a latitudinal pattern with distinct surface compositions of volatiles -- consistent with the existing literature. However, previous mapping efforts were based primarily on compositional analysis using spectral indices (indicators) or implementation of complex radiative transfer models, which need (prior) expert knowledge, label data, or optical constants of representative endmembers. We prove that an application of unsupervised learning in this instance renders a satisfactory result in mapping the spatial distribution of ice compositions without any prior information or label data. Thus, such an application is specifically advantageous for a planetary surface mapping when label data are poorly constrained or completely unknown, because an understanding of surface material distribution is vital for volatile transport modeling at the planetary scale. We emphasize that the unsupervised learning used in this study has wide applicability and can be expanded to other planetary bodies of the Solar System for mapping surface material distribution.

Photo credit: Matheius Stern via Peta Pixel Midjourney AI is a fantastic tool for anyone wanting to test their imagination, and Mathieu Stern takes it to the next level with a series of vintage cameras that never existed, but look like they should. Generating each of these images took less than a minute after a prompt was provided, although finding the right text to use took him hours to get just right. Once the cameras were generated, Stern then polished them off in Photoshop to make it look as if they were all shot in a professional studio. The only thing left to do was create a convincing backstory as to who might have built these cameras, and that required the use of ChatGPT that can not only answer questions, but also write entire essays at the click of a button. Mucha's designs featured brass frames adorned with intricate mother-of-pearl inlays, and the lenses were made from the finest European glass.

From the NEJM Catalyst event AI and Machine Learning for Health Care Delivery, sponsored by Advisory Board, March 24, 2022. In the special artificial intelligence theme issue of NEJM Catalyst Innovations in Care Delivery, "AI Insurance: How Liability Insurance Can Drive the Responsible Adoption of Artificial Intelligence in Health Care" explores how AI liability insurance can mitigate predictable risks and uncertainties to health care AI adoption. The big challenge for health care delivery is overcoming institutional mismatch, according to Stern. "The technologies that have the greatest potential to transform health care delivery --this includes, but is not limited, to AI -- would be unrecognizable to the 20th-century architects of our regulatory and health care delivery institutions," says Stern. "And this problem is getting worse. The pace of innovation that we see today coupled with our rapidly transforming analytical and technological capabilities is increasingly mismatched to our existing institutions."

Michelle Wu, cofounder and CEO and KK (Qiang Kou 寇强) tech cofounder at Nyquist Data, an AI powered ... [ ] cloud-based platform providing business, clinical, and regulatory intelligence and analytics for medical devices and pharmaceuticals companies How long does it take to get FDA approval for a heart-failure drug? It sounds like a simple question, but without the help of an artificial intelligence (AI) powered MedTech cloud-based platform, it could take months and millions of dollars to find out. The market size for AI in healthcare is projected to reach $187.95 billion by 2030, according to Precedence Research. When Michelle Wu was first asked this question, global clinical and regulatory healthcare information was publicly available, but it was scattered around the world in different databases and languages. Worse yet, keywords were misspelled or there were handwritten notes included in the databases, making what should be searchable unsearchable.