Tempo has completed integration with Safe{Wallet} to launch a programmable multi-signature account, supporting financial institutions and payment service providers in managing transaction fees with stablecoins, reducing transfer costs to below $0.001, and enhancing blockchain payment efficiency.

Volcano Engine has launched a short-drama production Agent, “Volcano Drama Creation,” using ByteDance’s Seedance 2.0 model to enable end-to-end automation across the entire process, including script parsing and shot generation. It is expected to reduce the production cycle by 80%. Core functions include automatically generating professional shot breakdowns, managing character and scene consistency, and performing precise shot parsing, greatly improving production efficiency, and also supporting multi-person collaboration and asset management.

On March 27, Bifrost launched SLPx Skill, the first AI-native multi-chain liquidity staking interface, supporting four EVM chains including Ethereum. Users can interact with the AI system through a smart Agent to achieve ETH staking and redemption, ensuring private key security and providing open-source compatibility.

Trust Wallet is accelerating its expansion into the AI field by launching the Trust Wallet Agent Kit, which allows authorized AI agents to perform on-chain operations on behalf of users, transforming crypto wallets from "storage tools" into "smart execution gateways." This tool supports a multi-chain ecosystem, enhancing the automated decision-making capabilities of Web3 applications, enabling users to set strategies for automatic execution by the system. In the future, an agent marketplace will be launched to further promote the platformization of the wallet ecosystem.

Gate News reports that on March 27, Renren Skill announced the launch of its public beta, aimed at creating an AI skill store specifically designed for Chinese users. The platform is fully compatible with the Chinese environment, providing users with a complete one-stop experience for browsing, installing, learning, using, sharing, and creating. According to the introduction, Renren Skill has established a three-tiered community system, covering multi-level needs from beginner users to hardcore developers. The platform allows users to easily build their own Skills and plans to open tipping and monetization mechanisms in the future to improve community ecosystem development.

Vietnamese police have cracked a massive cryptocurrency fraud case, arresting businessman Eric Vuong and six accomplices, with the amount involved reaching several billion dollars. The gang sold counterfeit coins to users through the fake platform ONUS, and police have summoned over 140 people for investigation.

The open-source vector database Chroma has released Context-1, a 20 billion parameter search model for multi-turn retrieval. This model utilizes self-editing context technology, trained on multiple tasks through reinforcement learning and a curriculum mechanism, demonstrating outstanding performance, particularly in the fields of web, finance, and law, and also showcasing cross-domain search capabilities in the email sector.

Gate News bot message, according to on-chain data, a newly created wallet address "0x5b1" withdrew 2,000 ETH valued at $8.77 million from multiple centralized exchanges. Meanwhile, whale address "0x49d" withdrew 800 XAUT ($3.55M), 210 ETH ($656.7K), and 70 PXAG ($312K). The whale currently holds 5,

Former technical lead of Alibaba's Tongyi Qianwen, Lin Junyang, explained in a lengthy article that the AI industry will shift from reasoning-based thinking to agent-based thinking, emphasizing the importance of environment construction. He believes that agent-based thinking will become mainstream, but it faces the risk of reward hacking. Future competitive advantages will depend on better environment design and multi-agent collaboration capabilities.

# Article Source: Axis Axis Robotics is restructuring data diversity and scaled production methods for embodied AI through a simulation-first strategy. In 2025, multiple technological pathways in the robotics industry are rapidly converging: the commoditization upgrade of embodied hardware supply chains has made it possible for expensive prototypes to achieve scaled deployment for the first time; Vision-Language-Action (VLA) models bring robots the "brain" for semantic understanding, reasoning, and planning; and the multi-layered data pyramid composed of video priors to simulation synthesis continues to supply steady fuel for the continuous evolution of embodied AI. However, the industry still faces one critical bottleneck: data. Compared to large language models and autonomous driving, embodied AI still faces a massive data