Title: 154: From Qwen's Personnel Changes to 'AI Legends'—A Conv...
URL Source: https://www.bestblogs.dev/podcast/0fb34ca
Published Time: 2026-03-11 22:30:00
Markdown Content: Skip to main content Toggle navigation menu Toggle navigation menuArticlesPodcastsVideosTweetsSourcesNewsletters
⌘K
Change language Switch ThemeSign In
Narrow Mode
Sign in to bookmark podcasts and track your listening history. Sign in now
154: 从千问变动到「AI 英雄传」，与 DINQ 高岱恒聊传奇 AI 研究员们 =========================================
One Sentence Summary
This episode explores the drastic shifts in the top-tier AI talent market, likening researchers to Renaissance masters while analyzing the changes within Alibaba's Qwen team and the future of AI labor models.
Podcast Info
From:晚点聊 LateTalk
Published At:Yesterday
AI Score:
86
Categories
Business & Tech
Chinese
#### Tags
Artificial Intelligence AI Talent Alibaba Qwen Open Source Community Reinforcement Learning
Your browser does not support the audio player.
晚点聊 LateTalk
Now Playing
154: From Qwen's Personnel Changes to 'AI Legends'—A Conversation with DINQ's Sam Gao about Legendary AI Researchers --------------------------------------------------------------------------------------------------------------------
00:00 00:00
1x
Show Notes Content Overview Transcript Q&A Recap 「美第奇和达芬奇，大公司和 AI 研究员。」
今天的节目分两次录制。嘉宾是高岱恒，Sam，他在 AI 技术圈和开源社区非常活跃，曾是双非土木老哥，后来以因为对 AI 开源项目的贡献进入阿里达摩院，现在自己创业，做了一个找到 AI 人才的 Agent 产品 DINQ。
而本期我们聊的主要话题，却是他的副业和 “消遣”：挖掘 AI 人物故事。Sam 对此非常着迷，在达摩院时，他就很喜欢搜集 AI 大牛的人生故事，那些非典型的、充满反差的经历尤其吸引他。
本期的第一部分，我们聊了上周刚发生的阿里千问的人事变化。Sam 分享了他作为开源模型受益者和前达摩院成员的观察。这整个过程的更具体情况，大家也可以看《晚点 LatePost》从上周到本周发布的 3 篇报道，以及我们的上一期播客。
本期第二部分，我们聊了 Sam 津津乐道的 AI 人物英雄传，在 Sam 眼里，这些不问出处、不拘一格的研究者就像文艺复兴时期的大师。他们有各自的性格、癖好、执念和技术浪漫，这对大型商业组织来说是陌生的、需要适应和理解的，我们也由此聊到了未来的研发型人才市场会如何变化。 本期嘉宾：
高岱恒 Sam DINQ 创始人 本期主播：
程曼祺 晚点 LatePost 科技报道负责人 时间线： 千问变动后续 02:42 千问候选人被搜索量翻了 3 倍 09:34 Model Scope 魔搭社区价值几何？ 14:30 美第奇和达芬奇，大公司和 AI 研究员 AI 英雄传（第一次聊，春节前） 22:38 DINQ 的诞生，AI 研究员需要更好地展示、更好地匹配 28:30 他们就像文艺复兴时期的大师 42:26 从土木老哥到达摩院，“当你有了作品，没人在问你学校” 49:52 Adam 发明者、被拒的 GPT-1 一作、旅居越南的 RAG 发明者………创新来自边缘地带 01:10:23 他们的追求、畏惧、不屑和弱点 01:19:38 AI 下的人才市场：从人和公司的匹配，到人和具体任务的匹配 01:43:48 连点成线 相关链接：
134 期:Meta AI 人才动荡，上亿美元为何留不住人？| 与 Pokee AI 朱哲清盘点 AI 组织 本期剪辑：Nick 小红书@曼祺_火柴Q即刻@曼祺_火柴Q
!Image 6 ☆《晚点聊 LateTalk》建立「 播客听友群」啦！☆
欢迎关注科技、商业大公司动态和创业创新的小伙伴进群交流，第一时间收听新节目。
这里有更多互动，更多话题讨论。欢迎贡献选题 & 推荐嘉宾。
请先添加「晚点」小助手的微信号，备注：“晚点聊”，我们邀请您入群。
!Image 7 关注公众号《晚点 LatePost》和《晚点对话》，阅读更多商业、科技文章：
Related Podcasts ---------------- 字节跳动技术团队 ### Deconstructing Agentic Coding from First Principles: From Theory to Practice This article deeply analyzes Agentic Coding from first principles, covering the essence of LLM, attention mechanisms, reinforcement learning, and Agent implementation principles, while proposing practical best practices such as short conversations, streamlined context, efficient configuration files, compounding engineering, and optimizing developer experience to enhance AI-assisted programming efficiency.MiniMax 稀宇科技 ### MiniMax M2.5 Released: $1/Hour, the King of Real-World Work MiniMax releases the M2.5 model, featuring SOTA performance for programming and office scenarios, achieving extreme cost-effectiveness ($1/hour) through a native Agent RL framework.宝玉的分享 ### Year-end AI Reflections by Salvatore Sanfilippo, the Creator of Redis Salvatore Sanfilippo, creator of Redis, offers eight profound insights into AI's state and future at the close of 2025, drawing on his dual role as a tech guru and an everyday AI user. He highlights the underestimated power of LLMs, the breakthroughs driven by reinforcement learning, and AI's ultimate challenge: preventing extinction.晚点聊 LateTalk ### 148: Interview with Tashi Founder Yilun Chen: Three Dawns and the First Hurdle of Embodied AI Yilun Chen, founder of Tashi, shares his experience transitioning from autonomous driving to general-purpose robotics, emphasizing the critical roles of data acquisition, model design, and application scenarios in the development of Embodied AI.腾讯科技 ### After Reading 200 Papers: Analyzing DeepMind, Meta, DeepSeek - What AGI Narratives Are Chinese and American Giants Describing? | 2025 AI Annual Review This article reviews 200 AI papers from 2025, deeply analyzing breakthroughs and bottlenecks in AGI across four key areas: reasoning, memory, spatial intelligence, and meta-learning, while looking ahead to 2026 development trends.晚点聊 LateTalk ### 146: Gemini 3's Resurgence: The Inside Story, What LLMs Do Agents Need, RL Startup Opportunities, Discussing Bay Area Trends with Former Google Entrepreneur & Silicon Valley Investor This podcast episode discusses the latest advancements in Silicon Valley's AI landscape, the competitive landscape of large models, the development of the Agent ecosystem, reinforcement learning entrepreneurship opportunities, and the ecosystem competition between Google TPUs and Nvidia GPUs.晚点聊 LateTalk ### 150: Year-End AI Review: From Models to Applications, Technology to Business Wars, Grasping the Thread of Meaning in the Torrent A comprehensive in-depth review of the 2025 global AI industry landscape, covering underlying model evolution, Agent implementation, Big Tech organizational battles, and embodied AI trends.晚点聊 LateTalk ### 151: Interview with MuleRun Founder Chen Yusen: Claude Code Brings a New Paradigm of Agent Creation, Future Software is Disposable An interview with MuleRun founder Chen Yusen, deeply exploring the Claude Code-driven paradigm of Agent creation and the future of "disposable software." HomeArticlesPodcastsVideosTweets
154: From Qwen's Personnel Changes to 'AI Legends'—A Conv... ===============