Банк Amina получил лицензию на оказание криптоуслуг в Гонконге

cryptonews.ruОпубліковано о 2025-11-16Востаннє оновлено о 2025-11-18

Швейцарский криптобанк Amina сообщил, что получил одобрение Комиссии по ценным бумагам и фьючерсам Гонконга (SFC) на оказание юрлицам услуг по хранению и торговле криптовалютами.

Amina будет обслуживать местные компании через свою дочернюю компанию Amina HK, зарегистрированную в этом особом регионе Китая. Гонконгская лицензия позволит банку предлагать услуги с 13 криптовалютами, включая биткоин, эфир, стейблкоины USDC и USDT.

Глава подразделения Amina в Гонконге и Азиатско-Тихоокеанском регионе Майкл Бенц (Michael Benz) сказал, что банк стремится расширить деятельность по управлению частными фондами, структурированными продуктами, криптодеривативами и токенизированными активами. Руководство компании сообщило, что Amina стала первой международной банковской группой, получившей лицензию гонконгского регулятора на оказание услуг, связанных с криптовалютами.

Власти Гонконга стали все больше ужесточать требования к деятельности криптокомпаний. В сентябре Денежно-кредитное управление Гонконга (HKMA) предупредило, что ограничит выдачу лицензий эмитентам стейблкоинам — разрешение на работу получат лишь те компании, которые готовы соблюдать строгие нормативные требования.

«Гонконг зарекомендовал себя как самый развитый рынок по надзору за криптовалютами. Получив лицензию, Amina сможет удовлетворить растущий спрос со стороны профессиональных инвесторов, желающих получить регулируемый доступ к криптовалютам», — заявил Майкл Бенц.

По данным Amina, в первой половине года объем торгов на гонконгских криптобиржах вырос на 233% в сравнении с аналогичным периодом 2024-го.

Недавно дочерняя компания банка, Amina Austria, получила от австрийского регулятора лицензию поставщика услуг по управлению криптоактивами (CASP). Банк сможет оказывать европейским клиентам услуги стейкинга и хранения криптоактивов, а также обменивать их на фиатные деньги.

Пов'язані матеріали

Embodied Intelligence 'Gaokao' is Insanely Hard, Humans Score 100, Best Model Only 12.8

Embodied AI Faces a Daunting "Everest": New Benchmark Reveals Huge Gap Between Models and Humans A comprehensive new benchmark for robotic manipulation, RoboDojo, has been released, painting a stark picture of the current state of embodied AI. It serves as a unified evaluation platform covering both simulation and real-world robot tasks. The benchmark assesses five core capabilities: Generalization (adapting to new scenes/objects), Memory, Precision manipulation, Long-Horizon multi-step tasks, and Open semantic understanding. It includes 42 simulation tasks and 18 standardized real-world tasks across three dual-arm robot platforms. The results are sobering. In simulation, the best-performing generalist robot policy achieved an average success rate of only 8.80%. Performance in the real world was slightly higher but still low, with the top model succeeding 12.8% of the time on average. In stark contrast, human experts scored 76.03% in simulation and 100% in real-world tests. The benchmark highlights significant, uneven gaps in current models' abilities. While some excel in specific areas like visual recognition or simple actions, they struggle with reliability, especially in long-horizon tasks where errors accumulate and in open-ended semantic instructions. The low scores, particularly in real-world deployment with physical uncertainties like camera noise and contact dynamics, underscore that today's models are far from being robust, general-purpose operational robots. RoboDojo is more than just a ranking; it's an infrastructure designed for fair, reproducible comparison. Its companion system, XPolicyLab, standardizes the interface for different models to be evaluated. Maintained by an academic consortium without commercial ties, it aims to provide a community-wide "altitude meter" to track genuine progress toward reliable and generalizable robot manipulation.

marsbit2 хв тому

Embodied Intelligence 'Gaokao' is Insanely Hard, Humans Score 100, Best Model Only 12.8

marsbit2 хв тому

Zuckerberg's 'Mango' Image Generation Model Trails Only GPT Image 2, It Learned to Revise Prompts on Its Own

Meta's MSL has launched Muse Image, an advanced image generation model nicknamed "Mango," which ranks second globally in text-to-image benchmarks, closely trailing OpenAI's GPT Image 2. Its key innovation is agent-like behavior: it searches for factual information, writes code for charts, and, most notably, has developed self-correction abilities through reinforcement learning, allowing it to revise its own outputs without explicit programming. This shift emphasizes reasoning over immediate generation. Integrated with Meta's ecosystem, Mango connects with the Muse Spark language model for complex tasks and features a unique "@" function that can incorporate public Instagram photos into generated images—raising privacy concerns as it's enabled by default. The model is directly accessible in Meta AI, Instagram, and WhatsApp, leveraging Meta's vast user base for distribution rather than competing solely on image quality. Accompanying Mango is the preview of Muse Video, a video generation model with integrated audio, currently ranked third in its category. All Mango-generated images include an invisible, persistent watermark (Content Seal) for AI identification, alongside a public detection tool. While Mango advances "thinking" image models, its use of social data poses new ethical questions about consent and digital boundaries.

marsbit1 год тому

Zuckerberg's 'Mango' Image Generation Model Trails Only GPT Image 2, It Learned to Revise Prompts on Its Own

marsbit1 год тому

Weng Li's New Blog Proposes 'Self-Evolution Should Start from Harness', DeepSeek's Cui Tianyi Endorses with Repost

Lilian Weng, former OpenAI security VP and co-founder of Thinking Machines Lab, has published a new blog post titled "Harness Engineering for Self-Improvement," proposing a pragmatic path for AI self-evolution. She argues that Recursive Self-Improvement (RSI) may practically begin at the "Harness" layer—the external runtime system governing how models use tools, manage context, and execute tasks—rather than directly from the model rewriting its own weights. The blog outlines a progression from optimizing prompts (Context Engineering) to designing workflows, and ultimately to Self-Improving Harness systems. These systems can identify their own weaknesses, propose targeted, verifiable modifications to the harness code, and validate improvements. Works like Self-Harness and Darwin Gödel Machine (DGM) demonstrate significant performance gains on benchmarks like SWE-bench through such automated harness evolution, rivaling handcrafted agents. DeepSeek researcher Tianyi Cui endorsed the view, noting harness-based self-evolution is as promising as model-based approaches. Weng emphasizes this is complementary to model training, with both reinforcing each other. However, key challenges remain: weak evaluators for subjective tasks, reward hacking, diversity collapse, managing long-term system health versus short-term success, and defining the human oversight role. The consensus is growing: the harness is a critical variable, as the same model can exhibit vastly different capabilities within different harness systems.

marsbit1 год тому

Weng Li's New Blog Proposes 'Self-Evolution Should Start from Harness', DeepSeek's Cui Tianyi Endorses with Repost

marsbit1 год тому

Торгівля

Спот
活动图片