The National Highway Traffic Safety Administration (NHTSA) has concluded an investigation into Tesla's Autopilot driver assistance system after reviewing hundreds of crashes, including 13 fatal incidents that led to 14 deaths. The organization has ruled that these accidents were due to driver misuse of the system. However, the NHTSA also found that "Tesla's weak driver engagement system was not appropriate for Autopilot's permissive operating capabilities." Riders using Autopilot or the company's Full Self-Driving technology "were not sufficiently engaged," because Tesla "did not adequately ensure that drivers maintained their attention on the driving task." The organization investigated nearly 1,000 crashes from January of 2018 until August of 2023, accounting for 29 total deaths.

Amazon's Fallout TV series is pretty good, yeah? Not only is it some darn great television in its own right, this high-budget, high-profile show might just be the most faithful adaptation of a video game ever put to screens big or small. It's so good that the Fallout video games, the most recent of which is almost seven years old, have been shooting back up the charts. But if you're new to the crumbling, irradiated world of Fallout, you might feel a little lost when the credits roll on the last episode. What's this New Vegas place hinted at in the post-credits scene? Why did the pre-war flashbacks look like Marty McFly's 1955, but have nuclear-powered robots? How did people invent Iron Man-style power armor if they can't make a computer smaller than a bread box?

Sam Altman, OpenAI's CEO, Microsoft chief Satya Nadella, Alphabet CEO Sundar Pichai are joining the government's Artificial Intelligence Safety and Security Board, according to The Wall Street Journal. They're also joined by Nvidia's Jensen Huang, Northrop Grumman's Kathy Warden and Delta's Ed Bastian, along with other leaders in the tech and AI industry. The AI board will be working with and advising the Department of Homeland Security on how it can safely deploy AI within the country's critical infrastructure. They're also tasked with conjuring recommendations for power grid operators, transportation service providers and manufacturing plants on how they can can protect their systems against potential threats that could be brought about by advances in the technology. The Biden administration ordered the creation of an AI safety board last year as part of a sweeping executive order that focuses on regulating AI development.

Prime Minister Fumio Kishida will visit France, Brazil and Paraguay from Wednesday through May 6, the government said Friday. In Paris on Thursday, Kishida plans to give a keynote speech at a ministerial council meeting of the OECD and meet with French President Emmanuel Macron. The speech will reflect Kishida's intention to lead discussions to resolve socio-economic challenges for the international community, Chief Cabinet Secretary Yoshimasa Hayashi said at a news conference. Kishida is also set to deliver speeches at OECD events themed on generative artificial intelligence and on cooperation with Southeast Asia. In Brasilia on May 3, Kishida will meet with President Luiz Inacio Lula da Silva, this year's chair of the Group of 20 major economies, and hold a joint news conference.

Artificial intelligence is enabling India's politicians to be everywhere at once in the world's largest election by cloning their voices and digital likenesses. Even dead public figures, such as politician and actress Jayaram Jayalalithaa, are getting digitally resurrected to canvass support in what is shaping up to be the biggest test yet of democratic elections in the age of AI-generated deepfakes. How will AIs like ChatGPT affect elections this year?

Addressing complex meteorological processes at a fine spatial resolution requires substantial computational resources. To accelerate meteorological simulations, researchers have utilized neural networks to downscale meteorological variables from low-resolution simulations. Despite notable advancements, contemporary cutting-edge downscaling algorithms tailored to specific variables. Addressing meteorological variables in isolation overlooks their interconnectedness, leading to an incomplete understanding of atmospheric dynamics. Additionally, the laborious processes of data collection, annotation, and computational resources required for individual variable downscaling are significant hurdles. Given the limited versatility of existing models across different meteorological variables and their failure to account for inter-variable relationships, this paper proposes a unified downscaling approach leveraging meta-learning. This framework aims to facilitate the downscaling of diverse meteorological variables derived from various numerical models and spatiotemporal scales. Trained at variables consisted of temperature, wind, surface pressure and total precipitation from ERA5 and GFS, the proposed method can be extended to downscale convective precipitation, potential energy, height, humidity and ozone from CFS, S2S and CMIP6 at different spatiotemporal scales, which demonstrating its capability to capture the interconnections among diverse variables. Our approach represents the initial effort to create a generalized downscaling model. Experimental evidence demonstrates that the proposed model outperforms existing top downscaling methods in both quantitative and qualitative assessments.

The forecast accuracy of deep-learning-based weather prediction models is improving rapidly, leading many to speak of a "second revolution in weather forecasting". With numerous methods being developed, and limited physical guarantees offered by deep-learning models, there is a critical need for comprehensive evaluation of these emerging techniques. While this need has been partly fulfilled by benchmark datasets, they provide little information on rare and impactful extreme events, or on compound impact metrics, for which model accuracy might degrade due to misrepresented dependencies between variables. To address these issues, we compare deep-learning weather prediction models (GraphCast, PanguWeather, FourCastNet) and ECMWF's high-resolution forecast (HRES) system in three case studies: the 2021 Pacific Northwest heatwave, the 2023 South Asian humid heatwave, and the North American winter storm in 2021. We find evidence that machine learning (ML) weather prediction models can locally achieve similar accuracy to HRES on record-shattering events such as the 2021 Pacific Northwest heatwave and even forecast the compound 2021 North American winter storm substantially better. However, extrapolating to extreme conditions may impact machine learning models more severely than HRES, as evidenced by the comparable or superior spatially- and temporally-aggregated forecast accuracy of HRES for the two heatwaves studied. The ML forecasts also lack variables required to assess the health risks of events such as the 2023 South Asian humid heatwave. Generally, case-study-driven, impact-centric evaluation can complement existing research, increase public trust, and aid in developing reliable ML weather prediction models.

The rapid evolution of Natural Language Processing (NLP) has favored major languages such as English, leaving a significant gap for many others due to limited resources. This is especially evident in the context of data annotation, a task whose importance cannot be underestimated, but which is time-consuming and costly. Thus, any dataset for resource-poor languages is precious, in particular when it is task-specific. Here, we explore the feasibility of repurposing existing datasets for a new NLP task: we repurposed the Belebele dataset (Bandarkar et al., 2023), which was designed for multiple-choice question answering (MCQA), to enable extractive QA (EQA) in the style of machine reading comprehension. We present annotation guidelines and a parallel EQA dataset for English and Modern Standard Arabic (MSA). We also present QA evaluation results for several monolingual and cross-lingual QA pairs including English, MSA, and five Arabic dialects. Our aim is to enable others to adapt our approach for the 120+ other language variants in Belebele, many of which are deemed under-resourced. We also conduct a thorough analysis and share our insights from the process, which we hope will contribute to a deeper understanding of the challenges and the opportunities associated with task reformulation in NLP research.

Clinical trial outcome prediction seeks to estimate the likelihood that a clinical trial will successfully reach its intended endpoint. This process predominantly involves the development of machine learning models that utilize a variety of data sources such as descriptions of the clinical trials, characteristics of the drug molecules, and specific disease conditions being targeted. Accurate predictions of trial outcomes are crucial for optimizing trial planning and prioritizing investments in a drug portfolio. While previous research has largely concentrated on small-molecule drugs, there is a growing need to focus on biologics-a rapidly expanding category of therapeutic agents that often lack the well-defined molecular properties associated with traditional drugs. Additionally, applying conventional methods like graph neural networks to biologics data proves challenging due to their complex nature. To address these challenges, we introduce the Language Interaction Network (LINT), a novel approach that predicts trial outcomes using only the free-text descriptions of the trials. We have rigorously tested the effectiveness of LINT across three phases of clinical trials, where it achieved ROC-AUC scores of 0.770, 0.740, and 0.748 for phases I, II, and III, respectively, specifically concerning trials involving biologic interventions.

Recent advancements in machine learning and deep learning have brought algorithmic fairness into sharp focus, illuminating concerns over discriminatory decision making that negatively impacts certain individuals or groups. These concerns have manifested in legal, ethical, and societal challenges, including the erosion of trust in intelligent systems. In response, this survey delves into the existing literature on algorithmic fairness, specifically highlighting its multifaceted social consequences. We introduce a novel taxonomy based on 'tolerance', a term we define as the degree to which variations in fairness outcomes are acceptable, providing a structured approach to understanding the subtleties of fairness within algorithmic decisions. Our systematic review covers diverse industries, revealing critical insights into the balance between algorithmic decision making and social equity. By synthesizing these insights, we outline a series of emerging challenges and propose strategic directions for future research and policy making, with the goal of advancing the field towards more equitable algorithmic systems.