CZ: The development of Real World Assets (RWA) will ultimately benefit Bitcoin, attracting U.S. stock market funds to enter the crypto market.

According to Ethereum's GitHub discussion page, the EIP-8450 proposal has been recently updated, aiming to introduce an anti-quantum public key registration mechanism for Ethereum validators. The mechanism adopts hash-based cryptography schemes, including leanXMSS keys and VRF roots, to enhance the resilience of the validator key system against future quantum computing threats. The update clarifies that anti-quantum key registration and replacement will be authorized via BLS signature keys, with a replacement delay of at least one epoch and anti-replay mechanisms set in place. The proposal remains in the Draft stage and has not been deployed yet.

BNB Chain stated in an official post that the Jenner hard fork upgrade has been deployed on the BSC testnet. The upgrade includes three features: Payment Lane (which reserves 5% of block gas by default for eligible payments), millisecond-precision block timestamps for smart contracts, and CAS20, a new compliance asset standard. Both Payment Lane and CAS20 will not be fully enabled until approved via on-chain governance, with more details to be announced later.

According to Onchain Lens monitoring, Hyperliquid has permanently burned 49.87 million HYPE tokens, valued at approximately $3.89 billion based on current metrics, accounting for 4.99% of HYPE’s maximum total supply of 1 billion tokens. Over the past 24 hours, Hyperliquid burned 19,700 HYPE tokens worth around $1.59 million; in the same period, its estimated fee revenue reached about $1.69 million, with cumulative revenue totaling roughly $1.38 billion.

Nonprofit Ethereum R&D lab Ethlabs announced in a statement that Ethereum co-founder Mihai Alisie has joined as an anchor funder, while Payward and Gattaca have come on board as contributors. The lab’s initial anchor funders include BitMine, Sharplink, and fellow Ethereum co-founder Joseph Lubin.

According to official news, OKX’s Flash Earn has been officially upgraded. In addition to the existing "Stake to Earn" and "Trade to Earn" models, a new "Hold to Earn" mode has been added. The first phase of the event will run from 08:00 on October 29, 2026 to 07:59 on November 28, 2026 (UTC+8), with a total prize pool of 400,000 USDT. Eligible users can participate by registering, hold 10 designated tokenized assets including xNVDA, xTSLA, xGOOGL and xQQQ, and join reward distribution based on their holding share—no need to subscribe or stake assets into the event prize pool. It is noted that a user’s daily total value of eligible holdings must be no less than 100 USDT, with a maximum daily reward of 650 USDT per user. Rewards are calculated starting from the next UTC day after successful registration, based on the minimum eligible holding value from the daily random snapshot. Users can register and participate via the "Flash Earn" - "Hold to Earn" entry.

Breaking news from Dongcha Beating AI: DeepSeek’s new model has seen a huge capability boost, likely thanks to Astra. The Information, citing multiple sources with direct knowledge, reports that DeepSeek’s V4.1-Flash-1017, updated on October 17, and Moonshot AI’s newly released Kimi K3.1 Pro both made extensive use of OpenAI’s GPT-6 Astra outputs for distillation during training. The new DeepSeek model performed particularly well on browser operation tasks: in 180 real web task tests from Browser Use, its score rose from 70.9 in the old version to 75.0, close to Astra’s 77.9. However, its per-question model inference cost is only $0.22, compared to Astra’s $10.77—nearly a 49x difference. Previously, Claude was the main target for distillation among Chinese AI firms. But Anthropic has tightened its safeguards over the past two months, making it far harder to extract the model’s reasoning processes, so domestic labs have turned their attention to Astra. DeepSeek founder Liang Wenfeng was initially reluctant to rely on distillation for V4’s development, but after seeing domestic competitors like Zhipu AI’s GLM make progress via distillation, he ultimately caved to competitive pressure.