Iran-backed militias in Iraq have claimed they were responsible for an attack on U.S. forces at a strategic base in southeastern Syria. The Islamic Resistance in Iraq, an umbrella group of Iranian-backed militias, said Monday that their forces used two drones to attack the al-Tanf garrison near the Jordanian and Iraqi borders, a sensitive location often used by Iranian-backed militants to transport weapons to Hezbollah. Monday's attack comes after a string of similar attacks on bases housing U.S. military in Iraq and Syria over the past week. In one, the same group attacked two bases in Iraq with drones, causing minor injuries among U.S. forces. The U.S. military has maintained a presence at the al-Tanf garrison since training forces as part of a campaign against the Islamic State group.

The British government should offer tax breaks for businesses developing AI-powered products and services, or applying AI to their existing operations, to "unlock the UK's potential for augmented productivity", according to a new University of Cambridge report. Researchers argue that the UK currently lacks the computing capacity and capital required to build "generative" machine learning models fast enough to compete with US companies such as Google, Microsoft or Open AI. Instead, they call for a UK focus on leveraging these new AI systems for real-world applications – such as developing new diagnostic products and addressing the shortage of software engineers, for example – which could provide a major boost to the British economy. However, the researchers caution that without new legislation to ensure the UK has solid legal and ethical AI regulation, such plans could falter. British industries and the public may struggle to trust emerging AI platforms such as ChatGPT enough to invest time and money into skilling up. The policy report is a collaboration between Cambridge's Minderoo Centre for Technology and Democracy, Bennett Institute for Public Policy, and ai@cam: the University's flagship initiative on artificial intelligence.

Government officials are using artificial intelligence (AI) and complex algorithms to help decide everything from who gets benefits to who should have their marriage licence approved, according to a Guardian investigation. The findings shed light on the haphazard and often uncontrolled way that cutting-edge technology is being used across Whitehall. Civil servants in at least eight Whitehall departments and a handful of police forces are using AI in a range of areas, but especially when it comes to helping them make decisions over welfare, immigration and criminal justice, the investigation shows. An algorithm used by the Department for Work and Pensions (DWP) which an MP believes mistakenly led to dozens of people having their benefits removed. A facial recognition tool used by the Metropolitan police has been found to make more mistakes recognising black faces than white ones under certain settings.

Kate Osamor, the Labour MP for Edmonton, recently received an email from a charity about a constituent of hers who had had her benefits suspended apparently without reason. "For well over a year now she has been trying to contact DWP [the Department for Work and Pensions] and find out more about the reason for the suspension of her UC [Universal Credit], but neither she nor our casework team have got anywhere," the email said. "It remains unclear why DWP has suspended the claim, never mind whether this had any merit … she has been unable to pay rent for 18 months and is consequently facing eviction proceedings." Osamor has been dealing with dozens of such cases in recent years, often involving Bulgarian nationals. She believes they have been victims of a semi-automated system that uses an algorithm to flag up potential benefits fraud before referring those cases to humans to make a final decision on whether to suspend people's claims.

One of California's riskiest volcanoes has for decades been undergoing geological changes and seismic activity, which are sometimes a precursor to an eruption, but -- thankfully -- no supervolcanic eruptions are expected. That's according to Caltech researchers who have been studying the Long Valley Caldera, which includes the Mammoth Lakes area in Mono County. The caldera was classified in 2018 by the U.S. Geological Survey as one of three volcanoes in the state -- along with 15 elsewhere in the U.S. -- considered a "very high threat," the highest-risk category defined by the agency. The two other volcanoes in California with that classification are Mt. Shasta in Siskiyou County and the Lassen Volcanic Center, which includes Lassen Peak in Shasta County.

AI software can land a plane on a carrier deck better than you. Over 5,000 men and women crew each of America's 11 aircraft carriers, but the U.S. Navy's counting on AI to help them fight China. AI will bring carrier planes in for landings, fly unmanned tankers with fuel for combat planes, and even analyze the bug juice in the chow line. Night carrier landings are dangerous feats of combat aviation. Americans think of the "Top Gun" movies starring Tom Cruise as Maverick, an intrepid Navy pilot who can land a 32,000-lb.

The UK government, with its reversals on climate policy and commitment to oil drilling and air pollution, usually seems to be pro-apocalypse. But lately, senior British politicians have been on a save-the-world tour. Prime minister Rishi Sunak, his ministers, and diplomats have been briefing their international counterparts about the existential dangers of runaway artificial superintelligence, which, they warn, could engineer bioweapons, empower autocrats, undermine democracy, and threaten the financial system. "I do not believe we can hold back the tide," deputy prime minister Oliver Dowden told the United Nations in late September. Dowden's doomerism is supposed to drum up support for the UK government's global summit on AI governance, scheduled for November 1 and 2. The event is being billed as the moment that the tide turns on the specter of killer AI, a chance to start building international consensus toward mitigating that risk.

This paper introduces a new IncidentAI dataset for safety prevention. Different from prior corpora that usually contain a single task, our dataset comprises three tasks: named entity recognition, cause-effect extraction, and information retrieval. The dataset is annotated by domain experts who have at least six years of practical experience as high-pressure gas conservation managers. We validate the contribution of the dataset in the scenario of safety prevention. Preliminary results on the three tasks show that NLP techniques are beneficial for analyzing incident reports to prevent future failures. The dataset facilitates future research in NLP and incident management communities. The access to the dataset is also provided (the IncidentAI dataset is available at: https://github.com/Cinnamon/incident-ai-dataset).

Modeling groundwater levels continuously across California's Central Valley (CV) hydrological system is challenging due to low-quality well data which is sparsely and noisily sampled across time and space. A novel machine learning method is proposed for modeling groundwater levels by learning from a 3D lithological texture model of the CV aquifer. The proposed formulation performs multivariate regression by combining Gaussian processes (GP) and deep neural networks (DNN). Proposed hierarchical modeling approach constitutes training the DNN to learn a lithologically informed latent space where non-parametric regression with GP is performed. The methodology is applied for modeling groundwater levels across the CV during 2015 - 2020. We demonstrate the efficacy of GP-DNN regression for modeling non-stationary features in the well data with fast and reliable uncertainty quantification. Our results indicate that the 2017 and 2019 wet years in California were largely ineffective in replenishing the groundwater loss caused during previous drought years.

Recent progress in natural language processing (NLP) owes much to remarkable advances in large language models (LLMs). Nevertheless, LLMs frequently "hallucinate," resulting in non-factual outputs. Our carefully designed human evaluation substantiates the serious hallucination issue, revealing that even GPT-3.5 produces factual outputs less than 25% of the time. This underscores the importance of fact verifiers in order to measure and incentivize progress. Our systematic investigation affirms that LLMs can be repurposed as effective fact verifiers with strong correlations with human judgments, at least in the Wikipedia domain. Surprisingly, FLAN-T5-11B, the least factual generator in our study, performs the best as a fact verifier, even outperforming more capable LLMs like GPT3.5 and ChatGPT. Delving deeper, we analyze the reliance of these LLMs on high-quality evidence, as well as their deficiencies in robustness and generalization ability. Our study presents insights for developing trustworthy generation models.