Anthropic的Claude在付费消费者中人气飙升Anthropic’s Claude popularity with paying consumers is skyrocketing

1. Hiroki, founder of Epismo, has launched the Epismo CLI, a tool designed to make human-AI workflows reusable and shareable. The CLI, available via npm, recorded over 380 downloads shortly after release. The tool addresses the challenge of reproducing complex, multi-step AI-assisted processes that span chat histories, tool configurations, and iterative prompts—common in tools like Claude Code. Currently, while code and models are easily reusable (via platforms like GitHub and Hugging Face), end-to-end workflows involving human-AI collaboration are not. Epismo aims to solve this by enabling users to save, share, and execute full workflows as structured, repeatable processes. The platform supports both human users and AI agents, allowing them to import and run shared workflows. A key innovation is the use of markdown-based chains to represent workflows, which agents can parse and execute step-by-step. The CLI is built for terminal-native agents, supports login via email OTP without a browser, and integrates with tools like Claude Code and Codex. An open-source skills repository is also available under MIT license.

Key Takeaways:
Epismo CLI enables reusable human-AI workflows via markdown-based task chains
Over 380 downloads post-launch shows early developer interest
Supports agent-native execution with terminal-based authentication and integration
Open-source skills repo promotes shared best practices across users and agents

Source: Original Article

1. Innovative Fishing Technologies Show Promise in Reducing Marine Bycatch
   Marine bycatch—the unintended capture of non-target species such as turtles, dolphins, sharks, and seabirds—remains a major ecological and economic challenge in global fisheries. Each year, millions of marine animals die due to entanglement in fishing gear, which can cause injury, suffocation, or fatal trauma even if animals are released. Beyond environmental harm, bycatch damages fishing equipment and increases operational costs, affecting fishermen’s livelihoods and industry reputations. While various mitigation strategies have been developed over decades, widespread adoption has been inconsistent due to practical, economic, or logistical barriers. Recent advancements, however, demonstrate measurable success. For example, nets equipped with LED lights have been shown to reduce sea turtle bycatch by up to 70% in some studies, as the lights deter turtles without significantly affecting target catch. Similarly, low-cost modifications—such as attaching plastic bottles to gillnets to increase visibility—have proven effective in reducing bycatch of certain species while remaining accessible to small-scale fishers. Researchers emphasize that unlike many conservation issues, bycatch reduction aligns the interests of fishermen, conservationists, consumers, and regulators. This broad consensus offers a strong foundation for scaling proven solutions. Despite progress, implementation gaps persist, particularly in regions with limited resources or regulatory oversight. Continued innovation, stakeholder collaboration, and policy support are essential to expand the use of effective bycatch reduction technologies.
   Key Takeaways:
   LED-lit nets significantly reduce sea turtle bycatch in tested fisheries
   Low-cost gear modifications offer practical bycatch solutions for small-scale fishers
   Stakeholder alignment on bycatch reduction enables broader conservation cooperation
   Implementation remains limited despite proven technological solutions
   Source: Original Article

1. A study published in New Scientist reveals that AI data centers can increase local air temperatures by up to 9.1°C in surrounding areas due to waste heat emissions. Researchers analyzed satellite data and climate models to assess the thermal impact of large-scale data infrastructure, particularly in regions with high concentrations of AI-driven computing facilities. The findings indicate that the heat output from these centers can significantly alter microclimates, potentially affecting local ecosystems, agriculture, and urban planning. While data centers have long been known to consume substantial energy, this research highlights a previously underappreciated environmental consequence: localized warming. The study calls for improved urban planning and more efficient cooling technologies to mitigate thermal pollution. Detailed information on specific locations or mitigation strategies is limited, but the research underscores the growing environmental footprint of digital infrastructure as AI adoption accelerates globally.

Key Takeaways:
AI data centers can raise local temperatures by up to 9.1°C
Waste heat from computing infrastructure alters surrounding microclimates
Environmental impact of digital growth extends beyond energy consumption
Improved cooling and planning needed to reduce thermal pollution

Source: Original Article

1. Researchers captured the first detailed drone footage of a sperm whale birth off the coast of Dominica on July 8, 2023. The event, lasting 34 minutes, was recorded in high resolution and documented the full delivery process, including postpartum behavior. The mother, a well-studied female sperm whale named Rounder (whale #5714), was assisted by other adult females in her pod, who positioned themselves around her during labor. At 11:46 a.m., the birth was confirmed by visible blood plumes and the appearance of the newborn calf. Immediately after delivery, pod members engaged in coordinated behavior, taking turns lifting and supporting the calf, indicating complex social cooperation. This footage provides unprecedented insight into sperm whale reproductive behavior and maternal care, which had previously been poorly documented due to the challenges of observing deep-diving marine mammals in their natural habitat. The study was conducted by a team from Project CETI, a research initiative focused on understanding sperm whale communication and social structures. The use of drones allowed non-invasive observation, minimizing disturbance to the animals. This advancement in data collection could inform future marine conservation strategies and deepen scientific understanding of cetacean social dynamics. Detailed information on the calf’s long-term survival or health outcomes is limited.
   Key Takeaways:
   First high-resolution drone footage of sperm whale birth recorded
   Pod members actively assisted mother and newborn after delivery
   Observation advances understanding of sperm whale social behavior
   Non-invasive drone use enables deeper marine mammal research
   Source: Original Article

1. Hugging Face has introduced a solution for users whose OpenClaw, Pi, or Open Code agents have been disrupted, enabling migration to open-source models via two primary methods: using Hugging Face Inference Providers or running models locally on user hardware. The hosted option via Inference Providers offers quick access to high-performance open models without requiring local computational resources. Users must generate a Hugging Face API token and configure their OpenClaw environment using the onboard command. The platform supports thousands of models, with GLM-5 recommended due to strong performance on Terminal Bench evaluations. Configuration is done through a JSON file specifying the model repo ID, such as huggingface/zai-org/GLM-5:fastest. Hugging Face PRO subscribers receive $2 monthly credits applicable to Inference Providers usage. Alternatively, users can run models locally using Llama.cpp, a lightweight, open-source inference library compatible with macOS, Linux, and Windows. This approach ensures full privacy, eliminates API costs, and avoids rate limits. A local server can be launched with models like Qwen3.5-35B-A3B-GGUF, which performs well on systems with 32GB RAM, though hardware compatibility varies by model. Detailed information limited on specific performance benchmarks and broader ecosystem integration.

Key Takeaways:
OpenClaw agents can migrate to open models via Hugging Face or local setup
Hugging Face Inference Providers offer fast access with API token authentication
Local deployment ensures privacy and zero API costs using Llama.cpp
GLM-5 and Qwen3.5 models recommended for terminal and local use

Source: Original Article