X
PulseAugur coverage of X — every cluster mentioning X across labs, papers, and developer communities, ranked by signal.
- developed by X Corp. 100%
- developed by Jack Dorsey 100%
- founded by Jack Dorsey 100%
- authored by Jack Dorsey 100%
- authored by Evan Williams 100%
- founded by Evan Williams 100%
- subsidiary of Elon Musk 100%
- developed by Evan Williams 100%
- subsidiary of X Corp. 100%
- used by Reddit 90%
- used by Claude Code 90%
- founded Grok 90%
- 2026-05-22 research_milestone SpaceX's S-1 filing revealed user numbers for X and Grok.
- 2026-05-19 product_launch X restricted the Grok fact-check feature to premium subscribers. 来源
- 2026-05-18 product_launch X introduced new daily posting limits for unverified accounts. 来源
- 2026-05-15 regulatory Multiple governments and the European Commission launched investigations into X's Grok AI and its rollout. 来源
- 2026-05-15 regulatory X commits to Ofcom to reduce hate and terror content on its platform in the UK. 来源
- 2026-05-13 product_launch X launched a new private History tab on its iOS app. 来源
- 2026-05-11 regulatory French cybercrime authorities are investigating X and its associated entities for alleged algorithmic manipulation, political interference, and the spread of harmful content.
- 2026-05-08 regulatory French prosecutors seek charges against Elon Musk and X for child sexual abuse images, deepfakes, disinformation, and complicity in denying crimes against humanity via its AI system.
- 2026-04-30 regulatory French prosecutors have opened a criminal investigation into Elon Musk and X.
- 2025-01-24 product_launch X launched a rebuilt, AI-powered advertising platform.
20 天有情绪数据
-
Elon Musk backs Warren Buffett's 5-minute plan for national debt
Elon Musk has publicly endorsed Warren Buffett's proposed solution to address the U.S. national debt. Buffett's 2011 plan suggests that if the deficit exceeds 3% of GDP, all current members of Congress should be ineligi…
-
网站所有者审计主页以兼容Google AI Overview
一位网站所有者审计了自己的主页，以了解Google的AI Overview为何没有引用其内容。审计发现缺乏结构化数据，包括缺少组织模式（organization schema）和错误的元标签（meta tags），这很可能阻止了Google的AI正确索引和引用该网站。所有者通过添加相关的组织模式和更正元信息来实施修复，旨在提高在AI生成搜索结果中的可见性。
-
AlphaGo 周年纪念、离线 AI 进展和 Claude Mythos 性能
Demis Hassabis 回顾了 AlphaGo 发布十周年，强调了它对围棋比赛的影响以及强化学习在人工智能领域的更广泛意义。另外，一条推文称赞了完全离线本地 AI 模型令人印象深刻的推理和长期记忆能力。另一篇帖子指出，Claude Mythos 在人类可在三小时内完成的软件任务上表现出近乎完美的性能，在一小时内的任务失败率接近于零。
-
Ex-CEO Shapiro flags immediate availability as hiring red flag
Gary Shapiro, former CEO of the Consumer Technology Association, employs a unique interview question to gauge candidate loyalty, viewing a two-week availability as a significant red flag. He prefers candidates who requi…
-
通过代码而非提示强制执行 AI 内容质量
一个名为“反滑坡门”的新系统正被用于强制执行质量并防止在 LLM 生成的内容中出现过度宣传性语言。该系统通过使用基于代码的过滤器（例如正则表达式）来在内容发布前拒绝特定的营销术语和结构模式。这种方法比仅依赖提示指令更有效，因为它确保了在 X、Bluesky 和 Dev.to 等各种平台上严格遵守语气和长度的指南。
-
Former UN Human Rights Director Craig Mokhiber comments on Trump, Epstein
Craig Mokhiber, the former UN Human Rights Director, has made statements regarding Donald Trump, Israel, and the Epstein case. These statements were shared via social media platforms, including Mastodon and X (formerly …
-
SocialCrawl 发布 AI 驱动的社交媒体搜索；AI '垃圾分叉' 挑战开源
SocialCrawl 是一项新服务，它聚合了来自 Reddit、X 和 TikTok 等 27 个社交和广告平台的实时数据。它利用 AI 进行情感分析、摘要和观点分布，并能够搜索标准搜索引擎无法访问的私有内容。该服务提供 API 以集成到 AI 代理和数据管道中，通过展示原始内容和 AI 生成的简报来强调透明度。Builder.io 正在应对开源软件中 AI 生成的“垃圾分叉”的兴起，这些分叉增加了维护者的拉取请求和审查负担。为了对…
-
Developer builds ORAG platform for organizational RAG and AI agent data access
Anmol Sharma has developed ORAG, a platform designed to make internal organizational data accessible and usable for AI agents. The system addresses the challenge of providing AI with relevant, trustworthy, and permissio…
-
Neverness To Everness 游戏用生成式AI艺术替换生成式AI艺术，引发粉丝强烈不满
游戏《Neverness To Everness》面临指控，称其用另一种生成式AI艺术替换了最初的生成式AI艺术，尽管开发商 Hotta Games 曾承诺会重新制作被标记的素材。粉丝注意到，游戏中一个最初因其AI生成的图像酷似《天气之子》而受到批评的广告牌，似乎已被更新为一张同样具有生成式AI特征的新图像，特别是上面有一个苏打罐。此举导致玩家感到沮丧，一些人质疑开发商的透明度以及他们移除AI内容的承诺，尤其是在其他AI素材已被占位符替换之后。
-
ShinyHunters hackers disrupt Canvas, impacting 9,000 schools during final exams
A cyberattack by the group ShinyHunters disrupted Canvas, a widely used educational platform, impacting nearly 9,000 schools globally. The outage occurred during a critical period for students preparing for final exams,…
-
FFmpeg 开发者在 Lex Fridman 播客上详述与谷歌的代码纠纷
Dwarkesh Patel，一位 FFmpeg 开发者，也是该项目 X 账号的发言人，讨论了近期与谷歌发生的一场争议。该纠纷围绕谷歌涉嫌滥用 FFmpeg 的开源代码展开，Patel 解释了这场冲突的技术和伦理层面。他强调了开源贡献的重要性，以及当大型企业不遵守许可协议时开发者所面临的挑战。
-
Twitch CEO announces new penalties for streamers caught viewbotting
Twitch is implementing new penalties for streamers who engage in viewbotting, a practice that artificially inflates viewership numbers. CEO Dan Clancy announced that channels identified as persistently viewbotting will …
-
Google Chrome 用户现在可以禁用 Gemini Nano AI 模型
Google 已将其 Gemini Nano AI 模型集成到 Chrome 桌面浏览器中，此举引起了用户对数据隐私和资源使用的关注和担忧。虽然该模型支持无需将数据发送到云端的设备端功能（如诈骗检测和开发者 API），但其存在和大小（约 4 GB）并未广为人知。此后，Google 在 Chrome 设置中引入了一个“设备端 AI”切换开关，允许用户禁用该模型，从而停止下载或更新。
-
White House calls Mark Hamill 'sick' over AI image of Trump in grave
The White House has condemned actor Mark Hamill for posting an AI-generated image depicting Donald Trump in a grave. Hamill, known for his role as Luke Skywalker, shared the image with the caption "If Only" and "Donald …
-
Elon Musk mocks French prosecutors investigating X over deepfakes and child abuse images
Parisian prosecutors have initiated a judicial investigation into Elon Musk and his social media platform X, formerly Twitter. The probe focuses on alleged algorithmic abuse and fraudulent data usage, with potential cha…
-
GameStop CEO's eBay account suspended after selling items to fund acquisition bid
GameStop CEO Ryan Cohen had his eBay account suspended after attempting to sell items to fund his $56 billion bid to acquire the marketplace. Cohen listed various personal and company-related items, including old store …
-
欧盟议员同意放宽人工智能法案限制，优先考虑创新
欧盟国家已达成临时协议，将放宽欧盟《人工智能法案》中的某些限制。主要变化包括将高风险人工智能使用法规的执行推迟到2027年12月，并豁免工业人工智能应用不受严格的法律监管。此外，除非用户辅助人工智能工具的故障会带来健康或安全风险，否则它们不会面临高风险义务，公司将有三个月的宽限期来为人工智能生成的内容添加水印。
-
Spotify 推出人类艺术家认证徽章以打击 AI
据报道，Spotify 正在开发一个人类艺术家认证系统，旨在将其与人工智能生成的内容区分开来。这一举措与推特过去用于认证真实账号的系统有相似之处。目标是提高音乐流媒体平台内的信任度和透明度。
-
CNN-BiLSTM outperforms AutoML for Indonesian Twitter hate speech detection
This paper compares PyCaret AutoML and a CNN-BiLSTM model for detecting hate speech on Indonesian Twitter. The CNN-BiLSTM model achieved superior performance, with an accuracy of 83.8% and an F1-score of 81.2%, outperfo…
-
BuyWhere MCP misses Product Hunt launch but ships content and gains organic traction
BuyWhere MCP encountered significant obstacles during its planned Product Hunt launch on May 6, primarily due to credential blockers preventing the setup of necessary human accounts across various platforms. Despite the…