build

According to CoinDesk, Ethereum developers have fixed a vulnerability in the Glamsterdam upgrade that could allow operators to withdraw ETH intended for new validators from the staking service. The fix changed the handling rules for validator deposits and was merged on Monday, after developers had rejected the modification a week earlier.Researcher avsetsin discovered that if an operator first registers as a builder with a specific public key, and the staking service subsequently deposits ETH using the same key to activate the validator, the draft rules would credit both deposits to the builder. Avsetsin stated that this is a conditional code discovery and not a confirmed real-time exploit; existing checks can prevent the attack, and there are currently no reports of losses. The new rules treat deposits as sorted by withdrawal receipts, ensuring that validator deposits still belong to the validators.Glamsterdam has been running on the Sepolia testnet, but the launch on the Hoodi testnet has been delayed, with a new date to be determined before the mainnet launch.

According to Zhitong Finance, Intel has agreed to acquire the Israeli quantum computing startup Classiq Technologies, with the transaction price undisclosed and expected to be completed in the first half of 2027. Classiq's clients include BMW and AT&T, and its technology helps clients write code that can be tested and run on any type of quantum hardware.This acquisition will enhance Intel's research and development capabilities in quantum hardware and make its products easier to deploy on quantum computers. Intel CEO Pat Gelsinger stated that partnering with Classiq will transform Intel's quantum computing division into a full-stack quantum computing enterprise. Classiq has raised over $200 million in funding and has approximately 130 employees, and after the acquisition, it will continue to help clients develop code that can run on various types of quantum hardware, including competitors' hardware. Intel has been committed for over a decade to transforming its silicon-based manufacturing advantages into a position in quantum computing, focusing on spin qubit technology.

According to Bloomberg, the UK government has announced an investment of £100 million (approximately $134 million) to build an artificial intelligence biosafety laboratory, supported by OpenAI, Anthropic, and Google DeepMind. According to a statement from the UK government, the laboratory will utilize artificial intelligence to detect and monitor biological threats.In addition to the aforementioned funding, OpenAI and Anthropic will also provide up to £200 million in usage credits for artificial intelligence models. The laboratory will also collaborate with Google DeepMind and life sciences company Basecamp Research, with the UK government allocating an additional £65 million for developing responses to pathogens.The announcement also includes an additional £120 million to the Artificial Intelligence Safety Institute, which is responsible for regularly assessing cutting-edge artificial intelligence models; and an additional £23 million to the under-construction Artificial Intelligence Economic Research Institute for analyzing the impact of artificial intelligence on the economy.

On October 24, Solana co-founder Anatoly Yakovenko (Toly) posted an idea on X, planning to utilize Solana validators to build an oracle that allocates voting power based on staking weight, providing price assessments for tokenized assets and using such assets as collateral.According to this concept, validators will participate in price judgments based on the amount of SOL they have staked. The profits generated by the related mechanism could increase revenue for the Solana network through taxation. However, this idea is still in the conceptual stage and is not a formal governance proposal, with no clear implementation timeline or governance voting arrangements at present. If realized in the future, this mechanism could introduce a native alternative to Solana's existing oracle network, but it also involves supply changes brought about by the increase in SOL issuance, as well as potential conflicts of interest and manipulation risks arising from validators participating in collateral pricing.

The official account of Alibaba Cloud stated that the large model company Step Star has globally released its flagship open-source model Step 5 Preview, positioned as a primary working model for real-world intelligent agent tasks, excelling in AI programming, software engineering, professional knowledge work, and financial scenarios. The model has a total of 600 billion parameters and 27 billion active parameters, utilizing sparse MoE, with a maximum context window of 1 million, capable of processing text, images, and videos. The article mentions that it recently ranked second globally in both weekly and daily call volume on OpenRouter. On the day of the release, Alibaba Cloud announced the completion of DAY0 adaptation, optimized for joint inference based on chips, super node instances, and models, with inference computing power provided by the Lingjun super node.Step Star has developed its own intelligent agent reinforcement learning training platform, with the sandbox platform managed by Alibaba Cloud. The solution is based on Alibaba Cloud's Agent Sandbox and Kubernetes container service, aimed at elastic computing power in training mode, claiming to create tens of thousands of sandboxes in a single minute, with a maximum concurrency of up to one million, compatible with E2B API, and providing native Kubernetes interfaces. Through multi-level image caching, intelligent pre-scheduling, and control link optimization, it achieves millisecond-level application and minute-level creation of 30,000 sandboxes, with the sandbox lifecycle uniformly managed by Alibaba Cloud's backend.In terms of storage, each Pod starts with 15GB of free temporary storage, automatically increasing by 30GB each time when usage exceeds 70%, up to a maximum of 300GB. For security, MicroVM-level isolation is adopted by default, covering computing, networking, storage, and permissions, with zero-trust access to cloud products and large model interfaces. The results provided in the article include: a 90% reduction in image startup time, model evaluation cycles compressed from weekly to daily, and an average savings of about 40% on EBS storage costs; relying on ACK One Fleet, it can support the stable operation of hundreds of thousands of sandboxes simultaneously.

According to official news from Bitget, the platform announced an upgrade to its RWA strategy, further expanding its tokenized stock business from asset trading access to institutional-level liquidity infrastructure construction. This upgrade introduces the Super Match Engine, breaking through the traditional single model of direct connection to brokers, adopting an internal matching priority mechanism, and providing market makers with lower latency trade execution, clearer liquidity distribution, and more efficient capital utilization through exclusive quoting channels and layered depth market data. At the same time, Super Match supports differentiated order processing and market services to meet the diverse trading and operational needs of institutional traders.Previously, Bitget launched the rToken US stock transfer function on October 13, supporting over 3,000 US stocks and ETFs. Users can transfer eligible stock assets from traditional brokerage accounts to Bitget, and once the stocks are credited, they will be automatically converted into US stock spot rTokens based on the corresponding number of shares. On this basis, relying on the Super Match framework, the efficiency of US stock trading will be continuously enhanced.Bitget CEO Gracy Chen stated that the next phase of tokenized asset development is crucial for bridging the asset and liquidity channels between traditional finance and the digital asset market. Bitget will continue to improve cross-market trading infrastructure, promote the application of rToken in scenarios such as spot trading, lending, margin, and asset management, enhance asset liquidity and capital efficiency, and deepen the strategic layout of the panoramic exchange (UEX).

Tesla (TSLA.O) executives stated that the company is building its own maps for Full Self-Driving (FSD).

Meta Platforms (META.O) CEO Mark Zuckerberg stated that he plans to build a "large-scale Muse network" to achieve agent collaboration.

Elon Musk stated that SpaceX will build a ground communication network.

According to on-chain analyst Ai Yi's monitoring, a certain whale that built a position in ETH at the top of this bull market started accumulating around $4600 in September 2025. Subsequently, during the price decline, it continued to add to its position, extracting a total of 15,810 ETH, worth $63.85 million, with an average price of $4,101.85.From October 2025 to October 2026, this whale deposited a total of 15,361.1 ETH to exchanges, worth $49.55 million, during which it made deposits at price levels such as $4,478 and $1,910. The most recent deposit occurred 5 hours ago, with a final average deposit price of $3,226.22. If the aforementioned ETH were sold, it would incur a loss of $13.45 million. Currently, there are still 572 ETH remaining on-chain.

Sui co-founder and Chief Product Officer Adeniyi Abiodun wrote in a review: Several years ago, when the team began to build an L1, the venture capital firms and institutions they contacted suggested that they fork EVM to avoid the difficulties of self-development and to quickly launch another version of a similar product before continuing to advance.Abiodun stated that the team rejected this suggestion because they had a clear judgment on the direction of industry development and believed that replicating a flawed design was not advisable. Therefore, Sui was built from scratch, taking on the corresponding risks and time costs. He mentioned that the current trend aligns with the team's initial judgment, and among the batch of application scenarios people hope to build today, Sui is the only viable option, making the next few years very interesting.

X-Agent released the latest technical article "From Agent Payments to an Onchain Personal Assistant," revealing the important product direction for the next phase: moving from Agent Payments to an Onchain Personal Assistant that can understand user intent, obtain authorization, execute on-chain operations, and continuously track results.Unlike simply allowing the AI Agent to "have a wallet," X-Agent aims to solve a more critical issue: how to enable the Agent to safely complete an on-chain task on behalf of the user. For example, when a renewal bill arrives, the AI assistant can understand the bill, verify the payee and amount, generate a payment plan, and complete the transaction after user authorization; even if there are timeouts or interruptions during the process, it should resume the original task and check the payment status, rather than blindly repeating the payment. After the payment is completed, the Agent also needs to continue confirming whether the service purchased by the user has truly been delivered.To achieve this goal, X-Agent plans to further connect the three major modules: Agent ID, Agentic Email, and Agentic Wallet, which will respectively handle identity and responsibility attribution, task and communication entry, and controlled on-chain fund execution. Through mechanisms such as payment pre-calculation, one-time authorization, budget control, recovery, and reconciliation, clear security boundaries will be established for the Agent's autonomous execution. Its core logic can be summarized as: user intent → clear authorization → on-chain execution → result verification.X-Agent believes that a true on-chain AI assistant is not just "initiating a transaction on behalf of the user," but is capable of continuously completing tasks and being accountable for the final results under controlled permissions. This also means that X-Agent is exploring a persistent on-chain AI assistant aimed at individual users, moving beyond the infrastructure of Agent payments.

U.S. President Trump stated that GlobalFoundries will invest $21 billion in Malta, New York, to build two large semiconductor manufacturing facilities, with funding provided by Japan based on a previously reached trade agreement.

Confido Health has completed a total of $22.5 million in Series A financing, including a $12.5 million extension round led by z47_vc. The funds will be used to build an AI entity that connects patients, healthcare providers, insurance companies, and pharmacies, automating processes such as appointments, insurance, referrals, and prescriptions.The data cited indicates that the annual healthcare labor expenditure in the United States exceeds $800 billion, and healthcare IT expenditure exceeds $200 billion. By 2030, the shortage of healthcare personnel could reach 3 million.

GlobalFoundries plans to build two chip factories with up to $21 billion in funding support from the Japanese government. The U.S. government will acquire about 4% of the company's shares.The new factories will produce foundational chips, widely used in fields such as automotive, industrial equipment, and communication devices. This move aims to increase the production of foundational chips in the U.S. and reduce dependence on East Asian suppliers.