Author: RadarAI Editorial
Editor: RadarAI Editorial
Last updated: 2026-11-20
Review status: Editorial review pending
Brief
速报
官方
AI动态
开源
Editorial standards and source policy: Editorial standards, Team. Content links to primary sources; see Methodology.
## 🔍 Core Insights
**Anthropic** completed the largest IPO in history with **$81.3 billion** raised, and its first-day market cap surged to **$1.27 trillion**, but controversy over **RSI (recursive self-improvement)** erupted simultaneously [6][9][14]; on the model side, **Gemini 4 Neon** approached **Opus 5.5**'s level at about one-third the cost [4], while **Zoom** topped the streaming transcription leaderboard with a **2.4%** word error rate [0].
## 🚀 Key Developments
- **Anthropic surges 17% on its first day of Nasdaq trading** [9]: Market cap reaches $1.27 trillion, with the IPO raising $81.3 billion—the largest in history.
- **Protests appear at Anthropic headquarters on IPO day** [14]: Former researcher Jacob Coxon calls for a ban on RSI, saying that if the public knew, hundreds of thousands would show up.
- **Cognition evaluates Gemini 4 Neon** [4]: Scores 46.2% on FrontierCode 2.0, less than 1 point behind Opus 5.5, at only about one-third the cost.
- **Zoom Scribe V2.1 Pro Live tops the streaming transcription leaderboard** [0]: AA-WER Streaming word error rate of 2.4%, narrowly beating MAI-Transcribe-2-Streaming's 2.5%.
- **OpenAI tests a brand-new ChatGPT mobile experience** [1]: Prioritized for Pro users; dot gets its own tab, and Codex, Projects, and more can be pinned to the tab bar.
- **ChatGPT Health expands to desktop, dots, Sites, and Android** [10]: Supports connecting health data, can analyze sleep and exercise patterns, and proactively suggests rescheduling workouts.
- **Gemini for Mac launches computer operation capabilities** [16]: Can perform tasks across folders, apps, and web pages on the user's behalf, while screen and cursor control remain with the user.
- **Apple recruits engineers to build its own AI agent** [12]: Leading a multi-step task agent across apps, browsers, command line, and APIs, in collaboration with the Siri team.
## 🔗 Sources
[0] Zoom Scribe V2.1 Pro Live tops the streaming transcription accuracy leaderboard — https://aihot.news/items/mxng1bugzfnky3z0fwpxb7zep
[1] OpenAI tests a brand-new ChatGPT mobile experience — https://aihot.news/items/s0q4y5e0r662omayvlae6yt7n
[4] Cognition evaluates Gemini 4 Neon: 46.2% on FrontierCode 2.0, close to Opus 5.5 at about one-third the cost — https://aihot.news/items/sr374ahztizayy7ozug6u9pii
[6] Just now, Anthropic rang the bell and went public! The largest IPO in history has Silicon Valley up in arms — https://www.bestblogs.dev/article/efa9ba5ca9?utm_source=rss&utm_medium=feed&utm_campaign=resources&entry=rss_article_item
[9] Anthropic surges 17% on its first day of Nasdaq trading, market cap reaches $1.27 trillion, and the $81.3 billion IPO becomes the largest in history — https://aihot.news/items/e7xjysyuhq3nk3kvhci7lwx1c
[10] ChatGPT Health features expand to the desktop app, dots, Sites, and Android — https://aihot.news/items/ng08f00cmwvxon8nlhm3xr5zw
[12
Anthropic completed the largest IPO in history with $81.3 billion raised, and its first-day market cap surged to $1.27 trillion, but controversy over RSI (recursive self-improvement) erupted simultaneously [6][9][14]; on the model side, Gemini 4 Neon approached Opus 5.5's level at about one-third the cost [4], while Zoom topped the streaming transcription leaderboard with a 2.4% word error rate [0].
🚀 Key Developments
- Anthropic surges 17% on its first day of Nasdaq trading [9]: Market cap reaches $1.27 trillion, with the IPO raising $81.3 billion—the largest in history.
- Protests appear at Anthropic headquarters on IPO day [14]: Former researcher Jacob Coxon calls for a ban on RSI, saying that if the public knew, hundreds of thousands would show up.
- Cognition evaluates Gemini 4 Neon [4]: Scores 46.2% on FrontierCode 2.0, less than 1 point behind Opus 5.5, at only about one-third the cost.
- Zoom Scribe V2.1 Pro Live tops the streaming transcription leaderboard [0]: AA-WER Streaming word error rate of 2.4%, narrowly beating MAI-Transcribe-2-Streaming's 2.5%.
- OpenAI tests a brand-new ChatGPT mobile experience [1]: Prioritized for Pro users; dot gets its own tab, and Codex, Projects, and more can be pinned to the tab bar.
- ChatGPT Health expands to desktop, dots, Sites, and Android [10]: Supports connecting health data, can analyze sleep and exercise patterns, and proactively suggests rescheduling workouts.
- Gemini for Mac launches computer operation capabilities [16]: Can perform tasks across folders, apps, and web pages on the user's behalf, while screen and cursor control remain with the user.
- Apple recruits engineers to build its own AI agent [12]: Leading a multi-step task agent across apps, browsers, command line, and APIs, in collaboration with the Siri team.
🔗 Sources
[0] Zoom Scribe V2.1 Pro Live tops the streaming transcription accuracy leaderboard — https://aihot.news/items/mxng1bugzfnky3z0fwpxb7zep
[1] OpenAI tests a brand-new ChatGPT mobile experience — https://aihot.news/items/s0q4y5e0r662omayvlae6yt7n
[4] Cognition evaluates Gemini 4 Neon: 46.2% on FrontierCode 2.0, close to Opus 5.5 at about one-third the cost — https://aihot.news/items/sr374ahztizayy7ozug6u9pii
[6] Just now, Anthropic rang the bell and went public! The largest IPO in history has Silicon Valley up in arms — https://www.bestblogs.dev/article/efa9ba5ca9?utm_source=rss&utm_medium=feed&utm_campaign=resources&entry=rss_article_item
[9] Anthropic surges 17% on its first day of Nasdaq trading, market cap reaches $1.27 trillion, and the $81.3 billion IPO becomes the largest in history — https://aihot.news/items/e7xjysyuhq3nk3kvhci7lwx1c
[10] ChatGPT Health features expand to the desktop app, dots, Sites, and Android — https://aihot.news/items/ng08f00cmwvxon8nlhm3xr5zw
[12
← Back to Updates