The National Weather Service (NWS) has paused language translation services of its products for non-English speakers living in the U.S., including emergency alerts.“Due to a contract lapse, NWS paused the automated language translation services for our products until further notice,” Michael Musher, public affairs specialist and a meteorologist for The National Oceanic and Atmospheric Administration’s (NOAA) National Weather Service told Fast Company in an emailed statement. According to a 2022 report from the U.S. Census Bureau, more than 68 million people in the U.S. speak a language other than English. Of them, around 42 million speak Spanish. As the Trump administration continues to seek major cuts to federal agencies, such as NOAA, the National Weather Service has been stretched thin, meaning translation services had to take a back seat. After recent Department of Government Efficiency (DOGE) cuts, the Associated Press reported that National Weather Service offices have a 20% vacancy rate. The administration hasn’t commented on translation services or whether they plan to ensure they make a comeback amid its cuts to the NWS. Where should non-English speakers get weather alerts? Norma Mendoza-Denton, a professor of anthropology at the University of California, Los Angeles, said, per AP News, alerts are the difference “between life and death” during severe weather. And what’s even more frightening is that there aren’t a ton of other places non-English speakers can get translated weather updates, even when severe weather strikes. While there are many apps that provide weather alerts, few offer built-in translation services.  One that does is AccuWeather. In 2021, AccuWeather partnered with Rubric, a translations company, for global weather updates in over 100 languages and dialects. Online, Accuweather has a 24/7 livestream, and provides “notifications of government issued watches, warnings, advisories, statements and other alerts that are issued by the relevant country’s government weather or civil protection services,” the site reads, noting that users can customize the types of alerts they receive.A representative for AccuWeather told Fast Company, “A localized description of the type of severe warning is displayed in the language the user has selected on their device. For example, a Spanish-language-app user will receive a warning description for flash-flooding in their area in the language they selected on their device.” The representative added, “We have no indication at this time that the changes at NOAA will impact AccuWeather’s services. We remain committed to providing the highest levels of service to the public, the media, and our clients.” The AccuWeather app is available for iPhone and Android in the App Store and on Google Play.In 2022, the Weather Channel launched a free 24/7 Spanish-language weather streaming service, the Weather Channel en Español. But last year, Allen Media Group pulled the plug on the channel. AMG spokesperson told Media Moves at the time, “As part of Allen Media Group’s extensive cost-cutting measures, the Weather Channel en Español will cease operations on December 31, 2024.” While there is currently not a national weather channel offering non-English weather reports in the U.S., residents should also look to their local news outlets, as some may offer news apps with translated weather updates and alerts. 

The Sakeb Carpenter’s Pencil is a carefully considered reinterpretation of the traditional carpenter’s pencil, designed with sustainability and cultural heritage at its foundation. Inspired by local craftsmanship and grounded in environmental responsibility, it reflects a deep respect for materials, making processes, and the people behind them. Visually, the pencil holds onto a raw, dusty appearance that mirrors the environment in which it was conceived. Its texture and finish feel unrefined in an intentional way, aligning with the aesthetics of a workshop rather than a design studio. Interestingly, their matte texture and chunky form give them an almost crayon-like appearance, playful yet practical, inviting users to engage with them in an intuitive and creative way. This authenticity reinforces the pencil’s identity as a working tool rooted in tradition. Designer: Raghad Saqfalhait and Mariam Dahabreh Made from recycled sawdust and bio-based binders, the pencil is a direct response to waste generated in woodworking. The sawdust, collected from Palestinian carpentry workshops, is transformed into a usable material that forms the body of the pencil. This approach not only repurposes discarded material but establishes a cyclical process: once the pencil is used, it can be melted and recast—even using its own shavings. This allows the pencil to be continuously reused, extending its lifespan and reducing waste throughout its lifecycle. The pencil’s form follows its function and context. It is thicker than standard pencils, mirroring the proportions of traditional carpenter’s tools. While this may appear unconventional at first, the thicker body serves a clear purpose. It enables users to make bold marks, shade broad areas, and focus on larger ideas, especially useful during the early phases of ideation or sketching. For carpenters and designers alike, the pencil supports a mindset focused on the overall picture rather than small details. Each Sakeb Carpenter’s Pencil is handcrafted, a deliberate choice that embraces analog production methods. This process ensures that no two pencils are identical. The uniqueness of each piece reflects the individuality of the makers and the variability of natural materials. It also connects users to a slower, more intentional form of production, contrasting with mass-manufactured tools that often lack a personal touch. Even the act of sharpening the pencil draws from carpentry. Instead of using a conventional sharpener, users file the pencil using sandpaper. This method complements the nature of the material and reinforces its connection to woodworking practices. It becomes a subtle ritual that ties the pencil to the tools and techniques used by the carpenters whose waste materials helped create it. Visually, the pencil holds onto a raw, dusty appearance that mirrors the environment in which it was conceived. Its texture and finish feel unrefined in an intentional way, aligning with the aesthetics of a workshop rather than a design studio. This authenticity reinforces the pencil’s identity as a working tool rooted in tradition. The post This Sustainable Sawdust Pencil Transforms Carpentry Waste Into Pure Creativity first appeared on Yanko Design.

Mami Wata has an unconventional way to promote sun protection.

Save on rowers and accessories with Hydrow coupons, including an exclusive discount of $125.

The United States is now applying 104% import tariffs against goods imported from China, that will eventually make all Apple hardware dramatically more expensive.Apple CEO Tim Cook [left] with Donald Trump [right] at a Mac Pro factoryFollowing Trump's implementation of a 34% tariff of imports from China into the United States and China's reciprocal tariff response, Trump threatened a 50% extension unless China backed down. On Wednesday, the threat was implemented.White House officials confirmed on Tuesday that a total 104% tariff on China would be implemented starting at 12:01am ET, 12:01pm China Standard Time, on Wednesday. The 104% figure is a combination of an existing 20% tariff, the 34% new tariff, and the 50% penalty for China pushing back with its own reciprocal tariffs on goods imported from the US. Continue Reading on AppleInsider | Discuss on our Forums

Donkey Kong Bananza will be the first mainline 3D Donkey Kong platformer in well over two decades, and based on the gameplay we’ve seen so far, it’s looking more than a little promising. We still don’t know who developed it, however, and Nintendo isn’t ready to share that information either. The leading theory since the game’s announcement last week has been that Donkey Kong Bananza was developed by the team behind Super Mario Odyssey. When asked in a recent interview with IGN if that was the case, however, Nintendo’s vice president of product and player experience Bill Trinen refused to divulge any information. Instead, Trinen says Nintendo intends to reveal that information later. “You’ll have to stay tuned,” Trinen said. “We always like to start by showcasing the game and getting people focused on the gameplay, but we’ll have some information to share about the developer in due time.” Frustratingly enough, this isn’t the first time Nintendo has taken similar steps, having refused to reveal the development teams behind the likes of Super Mario RPG, Mario and Luigi: Brothership, and Princess Peach: Showtime! prior to their respective releases. Based on Trinen’s words, however, it does seem like Donkey Kong’s development team will be revealed at some point pre-launch, so there’s that. Donkey Kong Bananza launches on July 17 for Nintendo Switch 2.

AI agents quickly become core components in handling complex human interactions, particularly in business environments where conversations span multiple turns and involve task execution, information extraction, and adherence to specific procedural rules. Unlike traditional chatbots that handle single-turn questions, these agents must hold context over several dialogue exchanges while integrating external data and tool usage. These challenges demand systems capable of navigating user goals incrementally, engaging in feedback loops, and invoking structured functions like API calls based on the conversation state. These capabilities heavily depend on the availability of training datasets that reflect such tasks’ natural complexity and sequence. As these AI agents are expected to operate under domain-specific constraints and execute task-relevant functions in finance, retail, and customer support, the demand for nuanced and verified training data grows significantly. The central bottleneck in scaling agent capability has been the lack of high-quality, multi-turn datasets that reflect realistic user interactions. Collecting such data manually is slow and costly and requires domain knowledge to construct tasks that represent actual use cases. Also, even leading language models tend to underperform in conversations that require tracking prior context, using tools precisely, or dynamically adjusting their strategy. Without structured training datasets that reflect these challenges, models are prone to errors in execution and struggle with maintaining goal alignment across turns. These limitations become more pronounced in scenarios that involve tool usage, such as executing function calls, retrieving external data, or fulfilling service requests with multiple stages of information exchange. Various frameworks have attempted to bridge this gap through synthetic data generation or task-specific tuning. Some efforts like APIGen and knowledge distillation methods have helped generate single-turn task data or simplified templates. Tool-usage models have been enhanced using frameworks that provide fixed sets of functions but often lack the flexibility to adapt to dynamic tool environments. Other attempts, such as MAG-V, MATRIX, and BUTTON, use multi-agent systems to simulate training interactions but suffer from inadequate quality controls or rely on fixed instruction structures. Many of these tools either fail to capture long-term dependency or rely on brittle rule-based systems that lack generalizability. Even popular evaluation benchmarks like MultiChallenge and ToolDial struggle to emulate the intricacies of realistic conversations, often due to overly simplified interaction formats. A research team from Salesforce AI Research introduced APIGen-MT, a novel two-phase data generation pipeline designed to create high-quality, multi-turn interaction data between agents and simulated human users. The approach focuses on realism, structure, and verification by constructing validated task blueprints and then simulating detailed agent-human conversations in executable environments. Unlike earlier approaches, this method employs a layered validation mechanism using both automated checkers and committees of large language models to assess task coherence, accuracy, and feasibility. The researchers train a family of models under the xLAM-2-fc-r series, ranging from 1 billion to 70 billion parameters, using this synthetic data to outperform major benchmarks in multi-turn agent evaluation significantly. The architecture behind APIGen-MT is split into two main operational phases. In Phase 1, a task configuration is created using an LLM-driven generator that proposes user intent instructions, a sequence of groundtruth actions, and the expected outputs. These proposals are then validated for format correctness, executability, and semantic coherence using a combination of rule-based checkers and a multi-agent LLM review committee. If a proposal fails at any stage, a feedback mechanism will reflect on the errors and propose improvements. Successful tasks move to Phase 2, where a simulation engine generates realistic dialogues between a simulated human user and a test agent. The agent responds to user inputs by calling APIs, interpreting outputs, and evolving the conversation across turns. Only those dialogue trajectories that match the expected groundtruth are included in the final training dataset, ensuring functional accuracy and natural dialogue flow. Models trained on APIGen-MT data, specifically the xLAM-2-fc-r models, demonstrate superior performance across two industry-standard evaluation benchmarks: τ-bench and BFCL v3. For example, on the BFCL v3 benchmark in the Retail domain, the xLAM-2-70b-fc-r model achieved a score of 78.2, surpassing Claude 3.5 (56.5) and GPT-4o (72.1). Similarly, the airline domain scored 67.1 compared to GPT-4o’s 62.8. In more complex environments involving iterative interactions, the xLAM-2-8b-fc-r model outperformed larger traditional models, illustrating the impact of higher-quality training data. These results confirm that detailed and verified training interactions are more valuable than sheer model size when structured carefully through feedback loops and task validation. Also, the consistency of these models across multiple trials shows enhanced robustness, a critical factor for deployment in enterprise environments. The APIGen-MT framework is impactful not only because of its performance but also because of its scalability and open-source contribution. By releasing both the synthetic datasets and the xLAM-2-fc-r models to the public, the researchers aim to democratize access to high-quality agent training data. This modular, verifiable, and interaction-grounded approach opens avenues for future advancements in AI agents. It enables researchers to extend the framework across different domains, functions, and tools, making it adaptable to specific industrial requirements without sacrificing dialogue realism or execution integrity. Some Key Takeaways from the Research: APIGen-MT creates multi-turn interaction datasets using a two-phase task blueprint generation followed by simulated conversation.   The system integrates validation via format checks, execution tests, and LLM review committees.   Feedback loops allow the improvement of failed tasks, creating a learning mechanism within the pipeline.   Models trained with this data outperform GPT-4o and Claude 3.5 across τ-bench and BFCL v3 benchmarks.   The xLAM-2-70b-fc-r scored 78.2 on Retail and 67.1 on Airline under BFCL v3, higher than all baselines.   Smaller models like xLAM-2-8b-fc-r also beat larger alternatives in long-turn interactions, indicating better efficiency.   The open-source release of both data and models ensures wider accessibility for research and industrial use.   The framework enhances realism and technical reliability in agent training, setting a new standard for synthetic interaction data. Check out the Paper and Model. All credit for this research goes to the researchers of this project. Also, feel free to follow us on Twitter and don’t forget to join our 85k+ ML SubReddit. Asif RazzaqWebsite |  + postsBioAsif Razzaq is the CEO of Marktechpost Media Inc.. As a visionary entrepreneur and engineer, Asif is committed to harnessing the potential of Artificial Intelligence for social good. His most recent endeavor is the launch of an Artificial Intelligence Media Platform, Marktechpost, which stands out for its in-depth coverage of machine learning and deep learning news that is both technically sound and easily understandable by a wide audience. The platform boasts of over 2 million monthly views, illustrating its popularity among audiences.Asif Razzaqhttps://www.marktechpost.com/author/6flvq/A Code Implementation to Use Ollama through Google Colab and Building a Local RAG Pipeline on Using DeepSeek-R1 1.5B through Ollama, LangChain, FAISS, and ChromaDB for Q&AAsif Razzaqhttps://www.marktechpost.com/author/6flvq/A Step-by-Step Coding Guide to Building a Gemini-Powered AI Startup Pitch Generator Using LiteLLM Framework, Gradio, and FPDF in Google Colab with PDF Export SupportAsif Razzaqhttps://www.marktechpost.com/author/6flvq/Transformer Meets Diffusion: How the Transfusion Architecture Empowers GPT-4o’s CreativityAsif Razzaqhttps://www.marktechpost.com/author/6flvq/Meta AI Just Released Llama 4 Scout and Llama 4 Maverick: The First Set of Llama 4 Models

Earlier this year, Dell rebooted the legendary Alienware Area-51 lineup of prebuilt gaming PCs, but until recently you could only configure it with one graphics card option - the RTX 5080. That's no longer the case. Starting now, you can choose to equip your Alienware Area-51 with an Intel Core Ultra 9 285K CPU and - finally - an Nvidia GeForce RTX 5090 GPU starting at $5,499.99. Better yet, you won't have to wait months for your system to arrive; Dell estimates an early April ship date.Alienware Area-51 RTX 5090 Prebuilt Gaming PCs Are AvailableNew ReleaseAlienware Area-51 Intel Core Ultra 9 285K RTX 5090 Gaming PC (32GB/2TB)$5,499.99 at AlienwareThe $5,500 config might be the "base" configuration for an Alienware Area-51 gaming PC equipped with a 5090 GPU, but at that price, I would still expect stacked specs, and fortunately, this configuration delivers. It's equipped with an Intel Core Ultra 9 285K CPU, 32GB of DDR5-6400MHz RAM, and a 2TB NVMe SSD. The Core Ultra 9 285K is Intel's latest flagship CPU and offers stellar workstation and gaming performance. It's not quite the performance uplift we wanted from the i9-14900K, but right now there's nothing better from the Intel camp. The processor is cooled by a massive 360mm all-in-one liquid cooler, and the system is generously powered by a 1,500W 80Plus Platinum-rated power supply.New for 2025: The Alienware Area-51 ChassisDell unveiled the new Alienware Area-51 gaming PC at CES 2025. The chassis looks similar to the 2024 R16 system with aesthetic and cooling redesigns and updated components. The I/O panel is positioned at the top of the case instead of the front, and the tempered glass window now spans the entire side panel instead of just a smaller cutout. As a result, the side panel vents are gone, and instead air intakes are located at the bottom as well as the front of the case. Alienware is now pushing a positive airflow design (more intake than exhaust airflow), which means a less dusty interior. The internal components have been refreshed with a new motherboard, faster RAM, and more powerful power supply to accommodate the new generation of CPUs and GPUs.The RTX 5090 Is the Most Powerful Graphics Card EverThe Nvidia GeForce RTX 5090 has emerged as the most powerful consumer GPU on the market. Although Nvidia has prioritized software updates, AI features, and DLSS 4 technology to improve gameplay performance, the 5090 still boasts an impressive 25%-30% uplift over the RTX 4090 in terms of pure hardware-based raster performance. The 5090 also has more (32GB vs 24GB) and faster (GDDR7 vs GDDR6) VRAM compared to the 4090. This GPU is extremely difficult to find at retail price and is currently selling for $3,500-$4000 on eBay.Nvidia GeForce RTX 5090 FE Review by Jackie Thomas"The Nvidia GeForce RTX 5090 has officially taken the performance crown from the RTX 4090, but with less force than previous generations. When it comes to traditional non-AI gaming performance, the RTX 5090 provides one of the smallest generational uplifts in recent memory. However, in games that support it, DLSS 4 really does deliver huge performance gains – you just have to make your peace with the fact that 75% of the frames are generated with AI."Check out more of the best Dell and Alienware gaming deals of 2025.Some Other 5090 Prebuilt Alternatives:Amazon listed a couple of Skytech-branded RTX 5090 gaming PCs for an even more attractive price point of $4,799.99 shipped. Note that these don't ship out immediately. They're equipped with capable AMD Ryzen 7 7800X3D processors, which perform nearly as well in terms of gaming although they do fall behind when it comes to workstation performance.Skytech Prism 4 AMD Ryzen 7 7800X3D RTX 5090 Gaming PC with 32GB RAM, 2TB SSD$4,799.99 at AmazonSkytech Legacy AMD Ryzen 7 7800X3D RTX 5090 Gaming PC (32GB/2TB)$4,799.99 at AmazonWhy Should You Trust IGN's Deals Team?IGN's deals team has a combined 30+ years of experience finding the best discounts in gaming, tech, and just about every other category. We don't try to trick our readers into buying things they don't need at prices that aren't worth buying something at. Our ultimate goal is to surface the best possible deals from brands we trust and our editorial team has personal experience with. You can check out our deals standards here for more information on our process, or keep up with the latest deals we find on IGN's Deals account on Twitter.Eric Song is the IGN commerce manager in charge of finding the best gaming and tech deals every day. When Eric isn't hunting for deals for other people at work, he's hunting for deals for himself during his free time.

abra R&D id seeking an experienced C++ WEB applications engineer to join our control systems software team. In this role, you will be responsible for designing, developing, and maintaining both new and existing software components for our mission-critical control and monitoring systems. Responsibilities:Analyze requirements and architect robust, high-performance C++ WEB applicationsImplement complex algorithms and data structures to solve challenging problemsCollaborate with cross-functional teams to deliver end-to-end solutionsMaintain and enhance existing codebase using modern C++ practicesWork in a Linux environment utilizing tools like Git, TFS, and EclipseBachelor's degree in Computer Science or a related technical fieldMinimum 1 year of professional C++ WEB development experience Proven track record of developing server-side applicationsStrong problem-solving skills and attention to detailExcellent teamwork and interpersonal communication abilitiesProficiency in both Hebrew and EnglishApply NowLet's start your dream job Apply now

Submitted by WA Contents Maud Caubet Architectes built an elevated house creating a contrast in the forest in Landes France Architecture News - Apr 09, 2025 - 05:16   html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd" Maud Caubet Architectes, an architecture firm based in Paris and Brussels, has designed a house elevated on stilts, creating a striking contrast in the forest of Landes, France.Maud Caubet has recently completed a new project for a private client on a 1,510-square-metre plot in Lit-et-Mixe in Landes, France, nestled in the forest just steps from the ocean.This 178-square-metre House Of Landes house, designed as an extension of a project finished in 2018 - with 130 square meters built then and an additional 48 square meters added in 2024, embodies an evolving approach to architecture.Six years after her initial intuition, Maud Caubet fine-tunes her idea of a home that is in sync with its surroundings. The volumes, construction techniques, materials, and finishes interact with the existing landscape to create a cohesive ensemble where contemporary architecture harmoniously integrates with the natural surroundings. A raw concrete staircase leads to a panoramic terrace with an ocean view, accentuating the H-shaped layout.A house in the treesSince the beginning in 2018, Maud Caubet aimed to blend the house into its surroundings by using local resources and respecting the site’s existing features. The house, raised above the ground on concrete stilts, aligns with the verticality of the pine trees while introducing an unexpected contrast to the forest’s uniformity. The façades, covered in Landes maritime pine with a dark brown saturation, resonate with the texture of tree bark, strengthening this continuity.A local carpenter implemented the off-site prefabrication of timber-frame walls and wood fiber insulation, enabling the assembly of both the house and its extension with minimal disturbance to the site. Natural ground preservation keeps the under-house area available for storage or shaded comfort in summer, with further enhancement from a lap pool discreetly positioned beneath the second volume. Wooden louvers that can be adjusted are used for the green roof and joineries, filtering sunlight and ensuring privacy.An evolving habitatThe house consists of two interconnected volumes and is designed for potential future expansions—adding a third or fourth module could further emphasize the idea of a home made up of interlinked spaces. The living room, kitchen, garage, and two bedrooms are located in the west volume, which features a south-facing terrace that is sheltered by an extended roof overhang. The smaller volume houses the master suite, which includes a bedroom, dressing room, office, and bathroom, and opens onto a terrace facing west with stunning views of the forest. These two self-sufficient “cabins” are linked by a sheltered walkway.Inside, the layout, composition and form of the rooms are shaped by usage and perspectives; the atmospheres are created in harmony with the natural surroundings beyond: water green, dark green, ocean blue, bark brown. The colors extend into the central patio, creating a unique atmosphere in every room through the interplay of interior and exterior elements, enhanced by the wooden cladding. And the aroma of the sea persists in the air throughout the house.Roof floor planSectionEast façadeWest façadeFounded in 2006, the agency brings together around twenty employees with strong skills and personalities. Adept at an instinctive approach that allows individuality to express itself, Maud Caubet Architectes brings small and large scales to the same level in the design of its projects, ensuring to weave links between architecture, design and the landscape in which they are implanted.Project factsProject name: House Of LandesArchitects: Maud Caubet ArchitectesLocation: Lit-et-Mixe, Landes, FranceCompletion date: 2024Project manager: Nataly TelloSite area: 1,510m2Total floor area: 178m2Main house: 130m2 /
 Extension: 48m2Partner: Amaury LaparraStructural Engineering: IDCEnvironmental approach: Wooden structure and layout carried out by Atelier AgencementAll images © Amaury Laparra. All drawings © Maud Caubet Architectes.> via Maud Caubet Architectes