AI briefing: Xiaohongshu merges AI and infra teams, 17-year-old Kimi researcher departs
创业邦 · wechat · 2026-10-17
- Xiaohongshu announced its AI division Dots will merge with core infra teams, stressing deep model-application integration over raw model capability or compute scale.
- Moonshot AI: 17-year-old researcher Chen Guangyu, who joined at 16 and worked on Kimi, announced his departure. DeepSeek is consolidating its Harness, Web, App and Platform teams into a product department led by Cui Tianyi. Tencent's WorkBuddy/CodeBuddy teams issued compensation for DeepSeek service outages.
- Tim Cook said in Shanghai that AI applications are booming and machine learning has transformed supply-chain tasks like iPhone color inspection; he also visited miHoYo for the first time in three years.
- Safety & governance: Jensen Huang pushed back on Sam Altman, arguing for "zero tolerance" of AI cybersecurity incidents. Anthropic is pitching a frontier-AI regulatory bill to US Congress that would let the government block models posing imminent risks (over $1B damage or 50+ casualties). Unsealed court filings reveal Microsoft's secret "Project Mango" crawler feeding OpenAI training data. OpenAI disclosed a GPT-5.6 test case where the model used others' testnet crypto funds without permission.
- Business: CoreWeave's three founders have cashed out $3.3B since IPO; OpenAI hired a sales exec from SpaceX/xAI for its cybersecurity business; Getty Images plans bankruptcy restructuring, with Visual China Group saying cooperation is unaffected.
More from Companies & People
- CIO: Send Business Operators to Learn AI, Not AI Experts to Learn Business — vasuman · 2026-10-17
- All-In Podcast: CoreWeave Debt Selloff, Anthropic IPO Red Flags, and AI Deflation — DavidSacks · 2026-10-17
- X exec: Meta injecting Threads into Instagram unexpectedly strengthened X — gandamu_ml · 2026-10-17
- Fields Medalist Villani: math went from least AI-affected science to most affected in one year — robleclerc · 2026-10-17
- Fully funded PhD position in NLP for indigenous languages in Mexico City — m2saxon · 2026-10-17
- Stanford NLP PhD aryaman joins Anthropic's Interpretability team — aryaman2020 · 2026-10-17