News Square
Global Crypto News · Real-time Aggregation
📚 Deep Analysis
More →Discord is making public chats searchable on Google: How to opt out
Discord conversations on large public servers will soon reach audiences far beyond their communities, raising privacy concerns. Discord is making a policy change that will make conversations from its largest public servers discoverable in search engines, including Google, raising privacy concerns among users whose conversations were previously confined to the platform. The change applies to Discoverable Servers — communities with at least 1,000 members listed in Discord’s public directory that do not require an invitation to join. From Nov. 16, every Discoverable Server will have a web page that may include summaries of public channel discussions. From mid-December, individual messages posted in eligible channels may appear in web search, Discord<a href="https://discord.com/blog/making-content-from-discoverable-servers-easier-to-discover" rel="nofollow noopener" target="_blank">> <span>said</span></a> Tuesday. Messages posted before the rollout will not be included, while usernames and avatars will be excluded from published messages. Example of an individual message being shown online. Source: <a href="https://discord.com/blog/making-content-from-discoverable-servers-easier-to-discover" rel="nofollow noopener" aria-label="Source: Discord" target="_blank">Discord</a> Discord said the shift of online communities from message boards to its servers had made useful answers harder to find. “This has led to valuable information being locked in Discord conversations, inaccessible to the wider internet. We want to fix this,” the company said. The move has <a href="https://www.reddit.com/r/discordapp/comments/1xgxb68/comment/19rr367kxqlk2/?utm_source=share&utm_medium=web3x&utm_name=web3xcss&utm_term=1&utm_content=share_button" rel="nofollow noopener" target="_blank">><span>drawn</span></a> criticism from users concerned about their conversations reaching an audience outside of their intended community, even if the servers are public in the first place. <em><strong>Related: </strong></em><a href="https://cointelegraph.com/news/starknet-introduces-wallet-privacy-feature-for-defi-activity" target="_blank" rel="nofollow"><span><em><strong>Starknet introduces wallet privacy feature for DeFi activity</strong></em></span></a> Others welcomed the change, saying it could make useful answers easier to find. One user in Discord’s announcement thread <a href="https://www.reddit.com/r/discordapp/comments/1xgxb68/official_megathread_discoverable_servers_are/" rel="nofollow noopener" target="_blank">><span>called</span></a> it “very much necessary,” while another said moving communities from forums to Discord had made their discussions harder to find through search. The change could also affect cryptocurrency communities that use Discord for technical discussions, support and project announcements. Cointelegraph found nearly 800 servers tagged “cryptocurrency,” around 1,400 tagged “trading” and more than 700 tagged “finance” in Discord’s public directory. Disc
AI begins to take on real engineering design: GPT-6 Astra achieves a 37% pass rate, while five models score zero.
CoinWorld news: An RWE-Bench test released by AI startup Xplore shows that GPT-6 Astra ranks first among 11 models with a single-run average pass rate of 37%, followed by Claude Opus 5.5 at 29% and GPT-6.1 SOL at 25%. The test covers 20 mechanical, electrical, and aerospace tasks, including engine crankshaft design, circuit board routing, and turbine blade cooling simulation. Each model uses its own accompanying agent toolchain to directly operate professional engineering software. All designs are automatically verified by engineering software and physics solvers, and failure to meet any single hard requirement counts as a failure. Xplore hopes to turn these engineering tasks into reinforcement learning training environments to improve AI's engineering capabilities.
U.S. spot Bitcoin ETFs recorded $286.3 million in inflows this week.
Bitcoin Magazine reports that U.S. spot Bitcoin ETFs have attracted $286.3 million in inflows so far this week, according to Bijie data.
Bitget partners with Reality and Keyring to release a research report, planning to launch an on-chain stock basis trading strategy.
Bitget, in collaboration with Reality and Keyring, has released a research report exploring the synergy mechanisms among tokenized stocks, perpetual contracts, and decentralized lending, and evaluating the potential returns and risks of related strategies through simulation-based calculations. The report notes that the application of tokenized stocks is expanding from traditional stock trading to more scenarios such as collateralized lending, asset management, and systematic investment strategies. Based on Reality's rtoken, the study combines stock perpetual contract hedging with on-chain lending mechanisms to explore stock basis return strategies. Historical simulation data shows that from December 15, 2025, to October 19, 2026, the strategy index achieved annualized net returns of approximately 7% and 4% in-sample and out-of-sample, respectively, with a gross funding rate premium of approximately 11%–14%. The report also points out that financing costs, trade execution fees, market liquidity, and risk management are key factors affecting the actual returns of the strategy.
Guotai Haitong and other parties jointly establish a 20 billion yuan Shanghai Linghang Science and Technology Innovation Fund.
According to CoinWorld news, Guotai Haitong (601211.SH) announced that its wholly-owned subsidiary Guotai Haitong Kaiyuan plans to jointly contribute capital with International Group Asset Management Company and other parties to establish the Shanghai Linghang Science and Technology Innovation Fund, with a total subscribed capital of 20 billion yuan, of which Guotai Haitong Kaiyuan will subscribe 4 billion yuan and serve as the fund manager, and International Group Asset Management Company will subscribe 2 billion yuan. The partnership will invest only in underlying funds approved by the partners' meeting, with investment directions including next-generation information technology, high-end equipment, new materials, biomedicine, artificial intelligence, the digital economy, and other key industries that align with national strategic priorities.
Financial Associated Press, October 28 (Xinhua) - Netanyahu's camp won 54 seats in the Israeli parliament, while the opposition won 49 seats; 77% of the votes have been counted.
According to a report by Bijie, Caixin reported on October 28 that, according to Israeli sources, Netanyahu's camp has won 54 seats in the Israeli parliament, while the opposition has 49 seats; 77% of the votes have been counted.
DBS Bank survey: 88% of large enterprises have used or piloted tokenized currencies
According to CoinWorld news, DBS Bank commissioned EY to conduct a survey of 422 senior finance executives at large enterprises with annual revenue exceeding US$500 million across 11 markets. The results showed that 88% of respondents have already used or piloted tokenized currencies. Among them, stablecoins accounted for 63%, tokenized deposits for 53%, and tokenized money market funds for 52%. 71% of respondents believe that real-time cross-border payments have improved or are expected to improve within two years, and about 64% hope that banks will become the unified gateway into the tokenized currency ecosystem.
BTC perpetual contract open interest totals 15.47 billion USD.
CoinWorld data: Under the statistical methodology of selected mainstream platforms, the total notional open interest of BTC USDT-margined perpetual contracts is $15.47 billion. This is down 0.82% compared with the same-methodology snapshot 24 hours earlier and down 1.96% compared with the same-methodology snapshot 7 days earlier.
Bitcoin's market capitalization share stands at 59.47%.
CoinWorld data: Under CoinWorld's unified market methodology, Bitcoin's market capitalization accounts for 59.47% of the total cryptocurrency market capitalization.
Liquid AI发布LFM3，1.7B小模型跑网页Agent胜过9B
Liquid AI has released its third-generation foundation model, LFM3, unveiling six small models with approximately 2 billion parameters at once, covering text, vision, speech, and decision-making tasks. Among them, the 1.7B text model supports a 256K context, and its 4-bit quantized weights are only about 1GB. All six models support fine-tuning on a single GPU, allowing enterprises to train them specifically with their own data and then deploy them to phones, computers, or private servers. Compared with the previous generation LFM2.5-2.6B, the new models have about 35% fewer parameters, yet their performance on some tasks has improved substantially. In official tests, LFM3-1.7B scored 38.42 on the WebArena-verified browser operation task, compared with only 3.02 for the previous generation, and it surpassed Qwen3.5-9B (28.94), which has about 5 times its parameter count. On the AutomationBench enterprise workflow test, its task success rate reached 41.29%, while Qwen3.5-9B scored 15.10%. However, LFM3-1.7B's performance declined somewhat on complex programming and long-text tasks.