arXiv:2606.07561v1 Announce Type: new Abstract: Gaussian processes with stationary kernels on bounded domains exhibit inflated posterior variance near the boundary. Despite being a long-recognized artifact in geostatistics and a source of over-exploration in Bayesian optimization, the causes and effects of boundary-induced acquisition bias are underexplored. We trace the root cause to a simple geometric mechanism: the truncation of the kernel correlation neighborhood at the domain boundary creates an observation-independent distortion that worsens with dimensionality. We show how this distortion manifests across three acquisition classes: variance maximization concentrates selections at the corners, whereas negative integrated posterior variance and expected predictive information gain move selections inward to axis-aligned interior shells. These patterns arise without reference to any objective function, meaning that acquisition behavior can be dominated by kernel geometry rather than the desired task-specific uncertainty. To quantify this, we introduce a function-free selection-profile diagnostic for arbitrary acquisitions, kernels, and bounded-domain geometries.

arXiv:2606.07613v1 Announce Type: new Abstract: Visual evidence has long been treated as a reliable form of legal proof, but advances in artificial intelligence (AI) are undermining that assumption. This article asks how well humans and frontier multimodal large language models (MLLMs) can distinguish authentic evidentiary photographs from AI-generated counterparts in the object-centric scenarios typical of civil disputes. We built Synthetic Legal Evidence Detection (SLED-1400), a dataset of 200 authentic evidence images paired with 1,200 synthetic counterparts produced by six contemporary text-to-image generators across ten evidence categories. The same stimuli and response format were used in a controlled web experiment with 136 lay participants and in a standardized evaluation of four MLLMs (GPT-5.1, Gemini-3-Pro, Gemini-3-Flash, Qwen3-VL-235B). Human accuracy was 64.8% overall, and 48.5% and 51.0% on the two strongest generators (Gemini-3-Pro-Image and Flux-2-Max), indistinguishable from chance. MLLMs never misclassified an authentic image (100% specificity), but missed most synthetic outputs from the harder generators, with average MLLM detection at 5.9% on Gemini-3-Pro-Image outputs. Human and MLLM errors were largely uncorrelated, while the four MLLMs were strongly correlated with each other. Neither group is a reliable standalone authenticator. We argue that visual evidence in legal proceedings should be treated as inherently contestable, and that a workable procedural response must combine trained human review, MLLM screening, and provenance infrastructure such as C2PA Content Credentials.

arXiv:2606.07558v1 Announce Type: new Abstract: Purpose: Digitization projects in the humanities produce vast, heterogeneous archives of historical documents, making manual sorting impractical at scale. This work addresses the need for an automated system to classify scanned page images based on visual content type - text, tables, and graphics - enabling content-specific downstream processing such as Optical Character Recognition (OCR) or structured data extraction. Methods: An image classification system was developed and evaluated on a dataset of over 48,000 annotated historical page images from century-old Czech archaeological archives, refined through four successive annotation stages with domain-expert review. A Random Forest Classifier baseline was established using hand-crafted image features. Subsequently, deep learning architectures were fine-tuned and compared: Convolutional Neural Networks (EfficientNetV2, RegNetY), Vision and Document Image Transformers (ViT, DiT), and multimodal CLIP models. An 11-category label scheme was designed collaboratively with domain experts and evaluated via five-fold cross-validation. Results: The feature-based baseline achieved approximately 75% accuracy. Fine-tuned CNNs and Transformers substantially outperformed it, with RegNetY-16GF achieving 99.16% and ViT-large 99.12% Top-1 accuracy on the held-out test set. CLIP ViT-B/16 reached 99.14% with optimized text descriptions. Conclusion: Image-only models, particularly RegNetY-16GF, deliver near-perfect classification accuracy and produce consistent labels across 649,508 unlabeled archival pages with over 90% inter-model agreement. Fine-tuned CLIP, despite competitive test-set accuracy, showed under 65% agreement with image-only models on unlabeled data, making it less suitable for deployment. The final models, annotated dataset, and software are publicly available under open-source licenses.

arXiv:2606.07585v1 Announce Type: new Abstract: This thesis addresses group emotion recognition (GER) in-the-wild with a focus on privacy preservation. Unlike traditional emotion recognition methods that rely on individual-level cues such as face, gaze, or voice analysis, this work uses collective audio-video signals to infer emotions at the group level, reducing risks of individual monitoring and surveillance. Two complementary frameworks are proposed. The first is a cross-attention multimodal architecture for audio-video fusion, combined with Frames Attention Pooling (FAP) for temporal aggregation. It is supported by synthetic data augmentation and validated through ablation studies, demonstrating robustness in real-world GER conditions. The second framework, Variational Encoder Multi-Decoder (VE-MD), learns a shared latent space for emotion classification and structural representation prediction, including body and face cues. Two decoding strategies, DETR-based and heatmap-based, are explored to analyze the role of structural representations in group and individual settings. The thesis makes three main contributions: it clarifies the role of multimodality and structural cues in group-level affective computing; introduces two architectures for privacy-preserving multimodal GER; and shows that competitive performance can be achieved without using individual features as input data.

arXiv:2606.07626v1 Announce Type: new Abstract: Perception in dense, unstructured urban traffic remains a major challenge for autonomous driving because of the wide variety of road users, frequent occlusions, irregular motion patterns, and the lack of standardized road layouts. Although recent LiDAR based 3D object detectors have shown strong performance in structured driving scenarios, most are developed and evaluated for limited field of view settings, and their behavior under full surround 360-degree sensing is still not well understood. This paper studies a 360-degree LiDAR perception pipeline for autonomous driving, with particular attention to panoramic sensing, azimuthal sector wise spatial processing, and transformation equivariant feature extraction in complex urban scenes. The paper presents a practical 360-degree perception framework that combines sector wise panoramic processing with rotation equivariant sparse convolutions and evaluates its behavior on a custom Ouster OS0 LiDAR dataset collected across diverse Indian urban traffic conditions. The results show generally stable detection across several object classes, with the strongest performance for cars at 92.02/90.51, buses at 80.53/76.34, and trucks at 78.59/74.16, while lower scores for pedestrians at 67.45/61.02, cyclists at 73.21/69.54, and motorcyclists at 71.20/68.13 reflect the greater difficulty of detecting smaller and more variable road users in dense urban scenes.

arXiv:2606.07550v1 Announce Type: new Abstract: Offline reinforcement learning (RL) offers a promising route for developing plasma controllers from historical tokamak data, since online trial-and-error on real devices is costly and risky. However, progress in this direction remains difficult to measure due to the lack of a standardized offline RL benchmark for realistic multi-actuator, long-horizon plasma control problems in nuclear fusion. We introduce RL4F, an Offline Reinforcement Learning Benchmark for Plasma Control in Nuclear Fusion, providing closed-loop evaluation environments and baseline comparisons across four full-profile tracking tasks: rotation, density, temperature, and pressure. The dynamics function underlying the evaluation environment is built from historical discharge data from DIII-D, a real-world Tokamak. We evaluate a broad set of imitation learning and offline RL baselines under a unified protocol. We find that offline model-based RL methods obtain the best average performance on most objectives, although no single method dominates all tasks, highlighting the importance of dynamics modeling in complex, long-horizon plasma control tasks. To foster further research, we open-source the codebase, datasets, and evaluation framework, providing a benchmark not only for the fusion community but also for algorithm development in offline RL.

arXiv:2606.07553v1 Announce Type: new Abstract: The emergence of machine learning and deep learning has revolutionized the efficiency of diagnostic, therapeutic, and administrative systems in healthcare. However, this rapid adoption has come at the cost of requiring significant computing power and energy consumption, as well as e-waste disposal and carbon emissions. One of the challenges of these models is choosing the right model for classification tasks. To this end, researchers attempt to identify the optimal model using their data through trial and error, which involves energy consumption and waste. The goal of this study is to develop a model-based recommender system for medical image classification. For this purpose, a data set was collected from 3,000 articles in the field of medical image classification. This dataset, publicly available under the name MedicalRec-Bench, contains over 5,000 records of models tested in various tasks, including Skin Cancer Classification, Tumour Classification, Wound Classification, Breast Cancer, and MRI classification. The dataset was evaluated in four different modes, depending on the number of features: MedicalRec I (5 features), MedicalRec II (9 features), MedicalRec III (11 features), and MedicalRec IV (18 features). Collecting all values for the features is challenging due to non-reporting by the authors; hence, the dataset contains significant amounts of missing values. The Medical Recommender System (MedicalRec) is a transformer-based model used for item recommendations in this study. This model achieved remarkable results in the evaluation on the dataset and in the evaluation with 12 base models. This model achieved a maximum HitRate@100 of 75.5%. The dataset and implementations are available through the GitHub link: https://github.com/Ramin1Mousa/MedicalRec

arXiv:2606.07620v1 Announce Type: new Abstract: With the growth of Vision Transformers in safety-critical domains like autonomous systems and medical imaging, ensuring their reliability against soft errors is paramount. While ViTs offer state-of-the-art accuracy, their massive parameter counts render exhaustive fault injection campaigns infeasible. To bridge this gap, a statistical fault injection framework is presented, leveraging finite-population sampling theory to provide formal reliability guarantees. It is demonstrated that failure rates are bounded within a 1% margin at 99\% confidence using only a few thousand samples, regardless of model scale. This methodology achieves up to a 10,700 times reduction in experimental cost compared to exhaustive approaches, while preserving the ability to localize vulnerabilities across architectural components. Through extensive evaluation of different architectures like ViT-Tiny and ViT-Small, a highly non-uniform reliability landscape is uncovered. It is shown that while only 3% of FP32 bit-flips result in failure, the vast majority of these events lead to catastrophic accuracy collapse. Specific vulnerabilities are localized to normalization layers and critical exponent bits within the IEEE-754 format, providing a mathematical foundation and actionable insights for the design of hardened, edge-deployed ViT architectures.

arXiv:2606.07595v1 Announce Type: new Abstract: Vision-language agents increasingly consume screenshots, documents, and user interfaces before writing to memory, sending messages, or invoking external tools. We study a concrete failure mode in this setting: action-boundary propagation, where sensitive or unsafe visible text is copied from an image into downstream tool arguments. We present VisualLeakBench, a diversified 500-image benchmark spanning UI, chat, document, form, and dashboard scenes, and evaluate a stratified 100-image agent subset with four production VLM systems under two workflows: note capture and external handoff. At baseline, target strings are propagated into tool arguments in 78.8% of PII cases and 85.5% of rendered unsafe-text cases. Under a defensive system prompt, rendered unsafe-text propagation remains high at 52.6%, while PII tool propagation falls to 2.0%, largely by suppressing tool use rather than preserving utility. Rates are tool-surface dependent: search-like tools suppress PII propagation, but rendered unsafe text still crosses tool boundaries. We measure visual-to-tool propagation rather than downstream instruction execution. We additionally provide a labeled-target oracle upper-bound diagnostic that localizes most failures at the tool boundary while leaving response-side leakage as residual risk.

arXiv:2606.07593v1 Announce Type: new Abstract: The widespread use of image classification models in high-risk, real-world situations necessitates making these models robust to slight disturbances or perturbations, such as blurring or sharpening, in the input images. While vision transformers (ViTs) play an integral role in many modern-day multi-modal models like Vision-Language-Models (VLMs) and Vision-Language-Action (VLA) models, they have received a lack of attention in the setting of robustness. In this work, we analyze the effects of adversarial fine-tuning, a popular method for improving model robustness to image perturbations, on a ViT's performance on perturbed and regular images through a mechanistic lens. We adversarially train a ViT on low-frequency and high-frequency image corruptions, and attempt to explain changes in downstream model performance through an examination of the model's attention mechanisms, internal representations, and knowledge evolution. Overall, our results suggest that, while fine-tuning on inputs with common corruptions improves model performance and certainty on new instances of corrupted data, these improvements do not transfer to other classes of corruptions not seen in the training. Additionally, despite observing changes in visual attention and knowledge evolution across layers, we found that adversarial training did not lead to fundamental changes in the sparse representations learned by ViTs.

arXiv:2606.07590v1 Announce Type: new Abstract: Pathology foundation models are pretrained on large streams of WSI-derived patches, while supervision during data construction is often slide-level, sparse, or heterogeneous. This mismatch makes it difficult to understand and control which biological patterns enter the pretraining data. We propose SlideCheck, a lightweight pretraining data guidance tool built on frozen pathology foundation model patch features. Rather than serving as a standalone patch diagnostic model, SlideCheck provides explicit abnormality and malignancy scores for organizing, filtering, and auditing pathology pretraining data. SlideCheck uses a dual-head MLP to separately model broad abnormal morphology and malignant evidence. A regularized feature-space scorer provides a supervised anchor for patch-level evidence estimation, while score-attention agreement combines patch scores with WSI-level MIL attention to mine high-confidence pseudo labels. The same scores are then used to construct broad-positive ViT pretraining subsets, where a patch is selected if either abnormality or malignancy evidence exceeds a threshold. Experiments show that SlideCheck-defined data distributions influence the downstream behavior of self-supervised ViT pretraining, indicating that biological composition is an important controllable factor in pathology foundation model development. Curated subsets can approach full-data performance, suggesting that explicitly scored patch pools may support more efficient and auditable pretraining data construction. These findings position SlideCheck as a data guidance and auditing layer for transforming large, undifferentiated patch pools into controllable and reusable pretraining datasets.

arXiv:2606.07569v1 Announce Type: new Abstract: Accurate carbon emission monitoring is critical for climate policy and emerging regulatory mechanisms such as the EU Carbon Border Adjustment Mechanism, yet city-level high-frequency monitoring data remain extremely scarce, severely limiting data-hungry deep learning models. Time series generation is a natural remedy, but existing GAN and diffusion-based generators often provide limited explicit supervision for the domain structure of carbon emission data: they may match marginal distributional statistics while insufficiently preserving cross-variable correlations between CO$_2$ and co-emitted pollutants and meteorological factors, and tend to collapse the first-difference statistics of atmospheric measurements, producing sequences that are smooth on average but lack the realistic step-wise variability of the underlying signals. We propose TriHead-GAN, a Transformer-based adversarial framework whose triple-head discriminator jointly supervises three complementary aspects of the joint distribution: distributional authenticity via a Wasserstein critic, cross-variable dependency via leakage-free regression of the target variable, and step-wise temporal smoothness via adjacent-difference prediction. The generator combines global self-attention with local temporal convolution, per-step noise injection, and an anti-smoothing loss that matches first-difference statistics. Experiments on the self-collected Changsha Carbon dataset, two public carbon datasets (China, US), and the ETTh1 benchmark show that TriHead-GAN achieves favorable performance over mainstream baselines on the vast majority of settings, and that the resulting synthetic windows improve downstream forecasting accuracy in low-resource carbon monitoring scenarios.

Andrew Orr / AppleInsider: Apple overhauls the App Store with subscription options for groups, businesses, and schools, retention messaging to reduce churn, bundled IAP reviews, and more  —  Apple is overhauling the App Store with new subscription tools, personalized recommendations, and marketing features …

Mark Gurman / Bloomberg: The first developer beta of iOS 27 offers the clearest signs yet of Apple's upcoming foldable iPhone, with references to folding hardware and flexible displays  —  Apple Inc.'s iOS 27 and related software updates offer the clearest public signs yet of the company's upcoming foldable iPhone …

Jose Antonio Lanz / Decrypt: Xiaomi claims MiMo-V2.5-Pro-UltraSpeed tops 1K tokens/second, a first at the 1T-parameter scale, using a standard 8-GPU commodity node; API trial starts June 9  —  Most people know Xiaomi as the Chinese phone brand.  The one that makes cheap electric scooters and air purifiers.

Can Apple's new AI glow up put to bed accusations that it's losing an all-important industry race?

OpenAI: Sam Altman and Jakub Pachocki claim OpenAI is “entering the third phase”, aiming to automate AI research, boost the economy, and give everyone a personal AGI  —  Every few generations, a new technology changes everything.  —  Imagine electricity reaching a rural American town in the 1920s.

Apple unveiled Siri AI and new features at 2026 WWDC. Josh Edelson / AFP via Getty Images Apple unveiled Siri AI and new features at the 2026 WWDC. Apple faced criticism and a lawsuit after delaying features promised for 2024. Multiple analysts said Apple could gain an edge over AI companies if it delivers on its promises. All eyes are on Apple as the company spent Monday trying to answer the question that's hung over the company since 2024: Can it finally deliver on the future it promised? At its 2026 Worldwide Developers Conference on Monday, the iPhone maker unveiled Siri AI, a long-awaited overhaul of its voice assistant, powered by a new generation of Apple Intelligence models developed in partnership with Google. It also announced a slate of new AI features across its ecosystem. The stakes are high for Apple. The company spent the past two years facing criticism and a class action lawsuit after many Apple Intelligence capabilities it previewed in 2024 were delayed or failed to arrive as initially promised. Here is what the smartest voices in business and tech are saying about Apple's flurry of announcements. Gene Munster, Apple analyst and managing partner at Deepwater Asset Management Gene Munster said Apple's moves would be successful if it could deliver everything it showed in the demo. Brian Ach/Getty Images for LocationWorld 2016 Gene Munster wrote in a post on X that Monday's moves would be a success for Apple if it could deliver everything it showed in the demo. "Keep in mind, it's still a demo, they overpromised with demos two years ago," Munster wrote. "That said, if they deliver what they showed today, it will drive hardware sales." "It goes beyond one-shot prompts and can reference previous conversations. The examples felt 10x better than using ChatGPT for personal tasks," Munster added. Munster wrote in a subsequent tweet that Apple is "aggressive" in its new child safety tools, which he called a "smart move because these little features tighten the grip that keeps parents in the Apple ecosystem." Christina Warren, developer relations executive at GitHub Warren wrote on X that Apple's AI strategy reflects the growing reality in the industry that smartphones aren't powerful enough to run the most advanced AI models entirely on-device. "As I expected, Apple is going to punt the 'on-device' story for Apple Intelligence and push towards the 'private cloud compute' story for the models that you'll actually want to use," Warren said on Monday. "I'm glad on-device isn't going away, but it's clear a hybrid approach is absolutely necessary," Warren added. "We will probably reach the point in the future where you could do good-enough on-device processing for certain tasks, but that day isn't today." Ben Bajarin, CEO and principal analyst at Creative Strategies Bajarin said that Apple seems to be turning Siri into a central hub for a one-stop AI experience rather than a collection of separate tools. "As we talk about enterprise workflow control plane, the Apple opportunity here is Siri is turning into the control plane for consumer AI," Bajarin wrote on X on Monday. "Personal context, screen awareness, app actions, writing, search, and visual intelligence route through one assistant layer, which is a different product experience all together," Bajarin added. Max Weinbach, consumer technology analyst at Creative Strategies Weinbach wrote on Monday that the Siri upgrade is potentially bad news for AI companies looking to build consumer products. "If the new Siri and AI features are good, and Apple can provide them for free with every new Apple device, everyone should be terrified of what happens to the AI companies," Weinbach said on X. "Siri AI is basically what most consumers use ChatGPT and Gemini for," Weinbach added in a follow-up post. "RIP consumer ambitions for AI companies." Alex Heath, tech journalist Alex Heath said in a series of posts on X that Apple is aiming for the same thing as the big AI players. Big Event Media/Getty Images for HumanX Conference Heath said in a series of tweets that Apple is aiming for exactly what the big AI players are: a "super personalized, all-knowing/present assistant." "What Apple is showing — Siri understanding whatever is on an iPhone's screen and providing contextual responses — is a huge unlock if it works as shown," Heath wrote on X on Monday. "Also shows how Apple continues to leverage platform control to do things folks at the app layer can't," Heath added. Heath was previously a senior reporter for Business Insider. Ernest Wong, head of research at Baskin Wealth Management Apple's lengthy introduction of its new child safety feature is "a perfect example" of "strategy credit," meaning an uncomplicated decision that improves the company's image relative to other competitors, Wong said. "It's an easy decision for Apple that simultaneously brings kids/teens into their ecosystem and handicaps competitors (social media apps), and looks good while doing it," Wong wrote on X. Chris Pirillo, founder of LockerGnome Chris Pirillo said that Apple's AI could enable the creation of automations similar to vibe coding. Kimberly White/Getty Images for Disney Apple's AI could make creating automations as simple as describing the result you want, much like vibe coding, Pirillo said. "This is basically vibe coding, folks. At a macro level," Pirillo wrote on X. Joanna Stern, chief tech analyst at NBC Joanna Stern said that she hopes Apple's child safety features have major under-the-hood improvements. Joy Malone/Getty Images Stern said in a series of posts on X that she used to be an "Apple Shortcuts hater," but the AI updates might significantly change her perception. "I'm glad Apple is finally giving parental controls some attention. But parents don't just need more controls. We need the controls to actually work," Stern added in subsequent posts regarding Apple's child safety updates. "I hope there's a major under-the-hood improvement here focused on syncing, reliability and consistency across devices." Dan Ives, tech analyst at Wedbush Securities Ives said in a note on Monday that Apple was finally delivering on what it promised two years ago with a "robust AI strategy and the announcement of Siri AI." "Overall, this was an impressive event that did not disappoint as Cook and Apple finally unveiled an AI strategy that will unleash the true monetization opportunity for AI in the Cupertino consumer ecosystem after a few years of promise … now it's finally here," he wrote. Read the original article on Business Insider

Clea Barry moved from the US to Chiang Mai with her husband and their three kids. Arada Photography. Clea Barry, 44, left the US for Thailand with her husband and their three kids in 2022. To make the move work, her husband accepted a teaching job at an international school in Chiang Mai. She says she and her husband have been able to enjoy their kids' childhood years more. This as-told-to essay is based on a conversation with Clea Barry, 44, an interior designer who moved from Kansas City to Chiang Mai, Thailand, with her husband and three kids in 2022. It has been edited for length and clarity. My husband and I met 20 years ago, and our shared love of travel brought us together. Over the years, we lived in several cities across the US with our three kids, including San Diego and Denver. Before moving abroad, we were living in Kansas City, where my husband taught middle school and I worked as an interior designer. I traveled quite a bit for work. Life with little kids already felt hectic, and then the pandemic happened. After about a year and a half of online school, we felt ready for a change. To make the move happen, her husband. accepted a job to teach at an international school in Thailand. Arada Photography. We had friends working abroad as international teachers, and seeing their experiences inspired us. We thought it would be a good opportunity for my husband to teach abroad for a couple of years while our kids attended the same school. It felt like a chance for all of us to experience life in a new country together. A new chapter abroad While we had considered moving abroad before, and had even received a few offers over the years, the timing never felt quite right. However, when a teaching opportunity in Chiang Mai, a city in northern Thailand, came up in 2022, we jumped at it. Have you relocated abroad — especially to Asia — for career opportunities, retirement, or lower living costs? We want to hear from you. Fill out this short form to share your story. By then, I had already scaled back my work after our third child, and I was freelancing remotely, which made it easier for us to move. It also helped that my husband and I had visited Chiang Mai on our honeymoon in 2009, so we knew we loved the Thai culture, food, and lifestyle. The family lives in a three-bedroom house about 15 minutes outside of Chiang Mai city center. Clea Barry. When we talked to our kids — who were 9, 7, and 2 at that point — about the move, we framed it as a family adventure. We also showed them YouTube videos about fun things they could try in Thailand, such as rolled ice cream, a popular dessert available at the night markets. That really excited them. In July 2022, we packed up our bags and officially moved to Chiang Mai. The original plan was to live for two years, but now it's been four. We're living in a three-bedroom home in Hang Dong, a district about 15 minutes outside Chiang Mai's city center. Rent is about $980 each month. Clea Barry. The home came furnished, including plants. We have a pool and a garden, and it's an eight-minute drive to the international school. There's a coffee shop across the street, and we're also right by walking trails. Rent is 32,000 Thai baht, or about $975, each month. Our kids have adjusted well. Since people move in and out of Chiang Mai throughout the year, it's common for a new student to arrive in August, January, or any other time. Instead of thinking it's strange, the kids get excited. They're very open to meeting new people because they know everyone comes from different parts of the world. I started volunteering at my kids' school over the holiday season, and that experience has been great because I get to see my kids at school, have lunch with them, and head home together. On days when I'm not volunteering at school, I drop off my kids at 8 o'clock in the morning, head home to do my design work, then pick them up again in the afternoon. She says the lifestyle in Chiang Mai is very family-friendly, with plenty of activities for kids to participate in. Clea Barry. Another thing I appreciate is that the school has club activities, like swimming and soccer, for the kids. Back in the US, I often had to shuttle them from place to place for activities. A family-friendly lifestyle The slower pace of life has been great for our family. On the weekends, we enjoy hiking together and going out to eat. One thing we love about Chiang Mai is its play cafés — restaurants with built-in play areas or activities for kids. So much of the dining here is outdoors, which means the kids get to run around. It almost feels like stepping back into the '90s. It feels safe here, and we feel comfortable letting the kids have a bit more independence. My daughter, now a tween, likes to go to the mall and meet her friends. She says living in Chiang Mai has enabled her to enjoy her kids' childhoods more. Clea Barry. It's also been very easy for us to find our community, especially through the school and through social media groups. There are a lot of events happening every weekend in Chang Mai, you almost have to pick and choose what you want to do. I still work remotely, but I don't have to work as many hours as I used to. In the US, I only had a few weeks of vacation each year and didn't have summers off, which made family travel much harder. Since moving to Chiang Mai, we've been able to travel to 10 different countries together, including Laos and Vietnam. What has been nice is that we've had a lot more time together. I'm able to enjoy my children's childhoods more, and that's something I'm really grateful for. Read the original article on Business Insider

The waiver "serves the public interest by promoting a second large satellite broadband constellation."