A crypto whale has opened a 15x short position on ETH, with the position value reaching 18.05 million USD.

As the market rebounds, Bonk Guy(@theunipcs)'s portfolio has grown by $1M+ in the past 24 hours. He remains No. 1 on both #FOMO's 24H and All-Time leaderboards.

Hong Kong Exchanges and Clearing (HKEX) has published a consultation summary document on shortening the settlement cycle of Hong Kong’s spot stock market, which shows widespread market support for implementing a T+1 settlement cycle in Hong Kong’s securities market. Based on feedback from various stakeholders, HKEX plans to complete the relevant transition in the second half of 2028, provided the market is ready and regulatory approval is obtained. The consultation received a total of 109 non-duplicate responses from market stakeholders including brokers, custodians, investment managers, industry associations, technology vendors, and investors. Respondents supported Hong Kong’s adoption of a T+1 settlement cycle, agreeing that this move will help enhance market efficiency, reduce settlement risks, improve capital efficiency, and consolidate Hong Kong’s competitiveness as an international financial center.

Beating AI News: Microsoft announces completion of pre-training for its new self-developed model MAI-Base-2. The model has a total of 2.8 trillion parameters, adopts a Mixture-of-Experts (MoE) architecture, and activates only 80 billion parameters per forward pass. Between August and September this year, Microsoft used 16,000 NVIDIA GB200 GPUs for 38 days of training, processing approximately 50 trillion tokens. The new model also introduces native visual pre-training for the first time, with visual data accounting for 18% of its training data. Internal tests released by Microsoft show that MAI-Base-2 outperforms models including DeepSeek-V4-Pro-Base, Kimi K3, and Qwen3.8 on text prediction benchmarks for code, question answering, mathematics, and science and engineering disciplines. Microsoft also claims that training a model of equivalent performance to DeepSeek-V4-Pro-Base would require only 1/4 to 1/12 of the computing power using its approach. However, Microsoft has invested additional computing power to build a larger, more capable model; MAI-Base-2’s actual training computing power consumption is still higher than that of DeepSeek. Microsoft AI head Mustafa Suleyman stated at an internal meeting on October 15 that MAI-Base-2’s pre-training base is already stronger than the OpenAI model supporting GPT-5.6 Sol, while requiring only one-fourth the training computing power. Microsoft improved MoE expert scheduling, Muon optimizer, and data filtering methods to reduce GPU memory waste and enhance the stability of large-scale training. The company also emphasized that all training data is processed in-house, and the model’s capabilities were not obtained through distillation of other firms’ models. MAI-Base-2 is currently in the post-training phase. Microsoft plans to first deploy it for its own products, then license the model weights to enterprise partners via Foundry.

Solana Foundation Chair Lily Liu posted that the large-scale light display on the entire exterior wall of Citigroup Building in Shanghai’s Lujiazui area on the evening of October 14 was meant to showcase Solana’s deep affection for the Chinese community and express gratitude for the warm welcome extended by the community. Due to promotion efforts, registrations for the "Solana Shanghai Event" quadrupled overnight, far exceeding the maximum number declared to local authorities. Officials were unable to prevent the excessive influx of attendees. To comply with local regulations, the event will be postponed, pending a time when it can better accommodate the community’s scale and reach. We extend our apologies to all friends who traveled near and far to join the Shanghai Solana event.

Lightning Labs has proposed a Bitcoin scaling solution called LayerTree, a non-custodial architecture that allows up to around 1 million users to share a single UTXO and connect to the Lightning Network. Balances are recorded in an off-chain Merkle-sum tree, with the shared output committing to the tree’s root, enabling users to exit unilaterally via a single transaction. The scheme depends on as-yet-unactivated Bitcoin consensus layer changes—such as covenant-related opcodes—and cannot be directly deployed on the mainnet at this time. The project team states that LayerTree is a design and reference implementation, and the opcodes it uses remain proposals rather than active consensus rules.

Beating AI Express News: Vercel launched Daemons at its Ship 2026 Developer Conference, enabling developers to directly deploy long-running background processes. These processes can continuously pull tasks from message queues or databases, process one after another, and automatically restart if they crash. Previously, Vercel primarily executed code on a request basis, so developers needing to run long-running tasks often had to deploy separate background services. Daemons is now in public beta. Another update is the official launch of Drives, which provides persistent file storage for AI Agents. Previously, files inside temporary sandboxes would disappear once the sandbox was shut down. Now, Agents can save code, data, and work products in independent storage for reuse when a new sandbox is launched next time. Developers can also duplicate workspaces, allowing multiple Agents to start from the same set of files, process different tasks in parallel, and avoid overwriting each other’s modifications. Vercel also added MCP Events to Connect, enabling Agents to subscribe to events from external systems. For example, when a new GitHub Issue is created or a document receives a new comment, compatible MCP services can proactively send notifications, triggering Agents to process them without repeated polling for new tasks. This feature is also in public beta and requires external services to support event subscriptions. These updates expand Vercel’s business scope. Developers who previously used the platform mainly for website deployment can now host long-running background tasks, Agent workspaces, and event integrations all on Vercel, reducing the need to build and maintain separate infrastructure.