Large Language Models (LLMs) are emerging as powerful tools in healthcare, particularly for complex, domain-specific tasks. This study describes the development and evaluation of the PErioperative AI CHatbot (PEACH), a secure LLM-based system integrated with local perioperative guidelines to support preoperative clinical decision-making. PEACH was embedded with 35 institutional perioperative protocols in the secure Claude 3.5 Sonet LLM framework within Pair Chat (developed by Singapore Government) and tested in a silent deployment with real-world data. Accuracy, safety, and usability were assessed. Deviations and hallucinations were categorized based on potential harm, and user feedback was evaluated using the Technology Acceptance Model (TAM). Updates were made after the initial silent deployment to amend one protocol. In 240 real-world clinical iterations, PEACH achieved a first-generation accuracy of 97.5% (78/80) and an overall accuracy of 96.7% (232/240) across three iterations. The updated PEACH demonstrated improved accuracy of 97.9% (235/240), with a statistically significant difference from the null hypothesis of 95% accuracy (p = 0.018, 95% CI: 0.952-0.991). Minimal hallucinations and deviations were observed (both 1/240 and 2/240, respectively). Clinicians reported that PEACH expedited decisions in 95% of cases, and inter-rater reliability ranged from kappa 0.772-0.893 within PEACH and 0.610-0.784 among attendings. PEACH is an accurate, adaptable tool that enhances consistency and efficiency in perioperative decision-making. Future research should explore its scalability across specialties and its impact on clinical outcomes.

In embodied instruction-following (EIF), the integration of pretrained language models (LMs) as task planners emerges as a significant branch, where tasks are planned at the skill level by prompting LMs with pretrained skills and user instructions. However, grounding these pretrained skills in different domains remains challenging due to their intricate entanglement with the domain-specific knowledge. To address this challenge, we present a semantic skill grounding (SemGro) framework that leverages the hierarchical nature of semantic skills. SemGro recognizes the broad spectrum of these skills, ranging from short-horizon low-semantic skills that are universally applicable across domains to long-horizon rich-semantic skills that are highly specialized and tailored for particular domains. The framework employs an iterative skill decomposition approach, starting from the higher levels of semantic skill hierarchy and then moving downwards, so as to ground each planned skill to an executable level within the target domain. To do so, we use the reasoning capabilities of LMs for composing and decomposing semantic skills, as well as their multi-modal extension for assessing the skill feasibility in the target domain. Our experiments in the VirtualHome benchmark show the efficacy of SemGro in 300 cross-domain EIF scenarios.

In this work, we propose a novel tree-based explanation technique, PEACH (Pretrained-embedding Explanation Across Contextual and Hierarchical Structure), that can explain how text-based documents are classified by using any pretrained contextual embeddings in a tree-based human-interpretable manner. Note that PEACH can adopt any contextual embeddings of the PLMs as a training input for the decision tree. Using the proposed PEACH, we perform a comprehensive analysis of several contextual embeddings on nine different NLP text classification benchmarks. This analysis demonstrates the flexibility of the model by applying several PLM contextual embeddings, its attribute selections, scaling, and clustering methods. Furthermore, we show the utility of explanations by visualising the feature selection and important trend of text classification via human-interpretable word-cloud-based trees, which clearly identify model mistakes and assist in dataset debugging. Besides interpretability, PEACH outperforms or is similar to those from pretrained models.

Marva Bailer, a tech executive, author, and speaker, shares why utilizing AI may be more expensive than physical actors. "Super Mario Bros." star Jack Black is feeling positive about the future with artificial intelligence, even with lingering concerns over the technology in the industry. "It's so new that it's hard to really say what the future holds, but I don't have all doom and gloom," he told The Hollywood Reporter. "I don't feel like, 'Oh no, it's going to be like Terminator where it comes and destroys all the human jobs.' I'm not convinced about that because I can admit, I don't know, and I'm hoping that it's going to be a great new world and that it's going to be a tool that all of us can use to make ourselves better and make the world better."

We propose that symbols are first and foremost external communication tools used between intelligent agents that allow knowledge to be transferred in a more efficient and effective manner than having to experience the world directly. But, they are also used internally within an agent through a form of self-communication to help formulate, describe and justify subsymbolic patterns of neural activity that truly implement thinking. Symbols, and our languages that make use of them, not only allow us to explain our thinking to others and ourselves, but also provide beneficial constraints (inductive bias) on learning about the world. In this paper we present relevant insights from neuroscience and cognitive science, about how the human brain represents symbols and the concepts they refer to, and how today's artificial neural networks can do the same. We then present a novel neuro-symbolic hypothesis and a plausible architecture for intelligent agents that combines subsymbolic representations for symbols and concepts for learning and reasoning. Our hypothesis and associated architecture imply that symbols will remain critical to the future of intelligent systems NOT because they are the fundamental building blocks of thought, but because they are characterizations of subsymbolic processes that constitute thought.

Multilingual pre-training significantly improves many multilingual NLP tasks, including machine translation. Most existing methods are based on some variants of masked language modeling and text-denoising objectives on monolingual data. Multilingual pre-training on monolingual data ignores the availability of parallel data in many language pairs. Also, some other works integrate the available human-generated parallel translation data in their pre-training. This kind of parallel data is definitely helpful, but it is limited even in high-resource language pairs. This paper introduces a novel semi-supervised method, SPDG, that generates high-quality pseudo-parallel data for multilingual pre-training. First, a denoising model is pre-trained on monolingual data to reorder, add, remove, and substitute words, enhancing the pre-training documents' quality. Then, we generate different pseudo-translations for each pre-training document using dictionaries for word-by-word translation and applying the pre-trained denoising model. The resulting pseudo-parallel data is then used to pre-train our multilingual sequence-to-sequence model, PEACH. Our experiments show that PEACH outperforms existing approaches used in training mT5 and mBART on various translation tasks, including supervised, zero- and few-shot scenarios. Moreover, PEACH's ability to transfer knowledge between similar languages makes it particularly useful for low-resource languages. Our results demonstrate that with high-quality dictionaries for generating accurate pseudo-parallel, PEACH can be valuable for low-resource languages.

A newly released trailer for Nintendo's upcoming movie, "The Super Mario Bros. Movie," heavily featured the voice of Princess Peach, voiced by Anya Taylor-Joy. Long a secondary character in the main franchise, the trailer revealed just how involved Peach would be in Nintendo's new movie, showcasing scenes where she dons her iconic Mario Kart biker outfit and even wields a halberd, all but ensuring her place as more than just a princess in another castle.

Peach, the global market leader in video advertising workflow and delivery has announced support of Netflix's new ad-supported service Basic with Ads. To coincide with the launch of the service, Peach has launched new destinations enabling clients to deliver ads to Netflix across multiple territories including UK, Australia, Germany, France, Italy, Spain, Mexico, Brazil with more to follow. Peach provides a connected advertising workflow, enabling clients to get their ads delivered to Netflix straight from the edit suite, while ensuring the highest possible quality, formatting and accuracy. Doug Conely, Chief Product and Technology Officer at Peach, said: "This is a pivotal moment for TV advertising. As leaders in global creative ad delivery for over 25 years, we've seen ad spend in Connected TV grow rapidly in the UK* and the rest of the world, and we expect to see further acceleration of growth driven by ad-supported tiers such as Netflix. AI and ML News: An Investment Into Artificial Intelligence as Daktela Buys Coworkers.ai "Netflix's Basic with Ads will bring our clients new audiences in a premium environment, creating opportunities for more addressable and premium content.

The Pentagon has made big plans to adopt artificial intelligence across the department, but two very large hurdles stand in the way of that goal, its new chief data officer said Wednesday: structuring data and recruiting the talent to manage it. "You can't feed the algorithms if you don't have data. Solid, clean data in large volumes, well-tagged and well organized," Michael Conlin said at the ACT-IAC Artificial Intelligence and Intelligent. "People will tell you that the machine learning algorithms, AI technologies can clean the data for you. The Department of Defense has no shortage of data to pull from, but for it to be of any use to the AI capabilities, the department has to make sure that data is recorded in consistent, machine-readable formats for accuracy and to ensure it doesn't present the algorithms with unintended bias, Conlin said. "The more data you have to train your algorithms, the more accurate the algorithms are and the faster you get your results," he said. As an example, he detailed the department's efforts to improve the flight readiness of aircraft by tracking the lifecycle of parts that are replaced frequently versus those that can be sustained for longer -- dubbed "lemons" and "peaches," respectively. Conlin said the department tracked the serial numbers for the parts from aircraft maintenance records and determined with 99.9 percent accuracy which parts were lemons and which were peaches after nine maintenance stops on each part. The problem stems from the data itself, however. Because department officials used the serial numbers to identify the lemons versus peaches, Conlin said, only 25 percent of the data was useful. "Some [records] had a blank or a'To be completed later,' or'I don't know,' or something that wasn't the serial number," he said. "So you couldn't connect the maintenance records together in order to be able to identify to nine consistent maintenance activities." Considering that Silicon Valley is focused more on delivering AI solutions tailored for specific use cases rather than enterprisewide applications, as well as the growing importance of edge computing, the quality of the structured data becomes that much more important. Equally important is the Pentagon's need for data scientists to help oversee the AI systems, Conlin said. But the challenge is the current federal workforce structure isn't designed for the job. We don't have the career path for data professionals, let alone data scientists," he said.

Having a robotic butler hand you a steaming cup of coffee and the newspaper in the morning is something science fiction has made us yearn for and modern robotics has brought into the realm of possibility. Yet roboticists are still having trouble teaching machines how to complete tasks that even children are capable of. That's why two researchers at the University of California, Berkeley have begun teaching a robot as if it were a five-year-old in the hopes of turning them into the taskmaster robots of the silver screen. "We're teaching robots how to pull off sorting and organizational tasks by simply watching a human do them once," Tianhe Yu co-author of the study tells Inverse. "Today's robots are able to perform a few specific tasks well, but they still don't come close to what a human is capable of. We hope that by teaching robots through demonstration we can enable them to carry out more general tasks."