{
  "text": "歡迎回到On Insight。如果我們要選一個節點作為全球AI產業,從學術夢幻期徹底跨入現實暴擊期的分水裡,我敢打賭,絕不會是哪家大廠又在PPT裡發布了多大規模參數的模型,基準測試又突破了多高的分數。\n真正的变局往往发生在那些\n看起来极度安静\n甚至带着几分诡异的高层人事变动力\n就在最近\n谷歌内部发生了一场\n足以重塑整个硅谷底层格局的震荡\n在谷歌效力了整整27年\n被全球程序员\n尊为代码之神的谷歌首席科学家\nJeff Dean正式宣布离职\n如果这只是一位老将退隐的话\n倒也不至于让硅谷各大投资机构\n和安全部门彻夜未免\n真正认识背后发紧的是\nJav Dien离职的时候\n带走了一个堪称\n硅谷全明星的豪华降维打击阵容\n这其中包括与Jav Dien合作\n20多年的底层系统传奇搭档\n桑杰·格马瓦特\n还有Google DeepMind研究副总裁\nJemline模型最核心的技术操盘手\n奥里奥尔·维尼亚尔斯\n以及AutoML领域先驱\n和Google Brain的联合创始人黎光莲\n这四个人加在一起\n几乎就是半个谷歌AI\n与底层分布式系统的技术灵魂\n他们组建了一家名为Discovery Loop的新公司\n按理说\n面对如此核心的技术骨干离职\n谷歌应该爆发一场血腥的公关战\n对吧\n但诡异的事情来了\n在继续后面的内容之前\n我想先请你帮我一个小忙\n如果你经常观看我的视频\n但还没有订阅我的频道\n欢迎点击订阅\n这样你就不会错过后续更多的科学资讯和深度解析\n那如果你觉得这期视频对你有所帮助\n也请帮忙点赞和分享\n让更多人看到这些内容\n那如果你希望给予我更多的支持\n也欢迎加入我的会员\n每月一杯咖啡的钱\n就是对我持续创作最大的支持\n好的我们继续\n这个诡异的事情就是\n谷歌不仅没有阻拦\n反而第一时间成为了Discovery Loop的初始投资方\n并且双手奉上了Google Cloud的超级算力集群\n双方直接签下了长期云服务绑定协议\n不仅如此\n就在同一天\nGoogle DeepMind内部也完成了一场\n近乎手术刀级别的高层重组\nDeepMind的灵魂人物Demith Hossabies\n正式卸任DeepMind CEO的日常管理\n身任Alphabet的首席科学家\n去操盘更为宏大的AGI理论中举\n而接替他全盘操盘Google DeepMind的新帅\n是原DevMind的CTO\nDQN与WaveNet的发明人\n科雷·卡伍克·乔鲲鲁\n他越过了所有中间管理层\n直接向谷歌CEO桑德尔PX1汇报\n听到这里\n你脑海里一定盘旋着一个极其巨大的悬念\n就是谷歌为什么会心甘情愿的\n让自己的顶尖大脑集体外逃\n甚至还要掏钱掏算力\n帮他们去创业呢\n而那位既懂DQN强化学习\n又懂Vivnet原生语音的工程狂人\n柯雷接棒之后\n到底要把Gemla带向何方呢\n今天我们就有一期节目\n彻底拆解这场看似是高层离职潮\n实则是谷歌AI技术大分流的\n硅谷终极棋局\n要解开这个谜团\n我们首先得搞清楚\nJeff Dean带着这帮神仙阵容杀出去\n到底要去干一件什么样的事情\n在Discovery Loop的官宣里\nJeff Dean直接描绘了公司的愿景\n那就是构建自动化的科学方法闭环\n简单来说\n过去几百年\n人类探索未知的方式\n本质上都是串行实验\n就是科学家提出假设\n亲手写代码\n跑一次实验\n等几天看结果\n分析数据\n然后再手动修改参数\n跑下一次\n在这套流程里\n最瓶颈的变量不是算力\n而是人类脑力的思考速度与手动操作的物理实验\n而Discovery Loop要构建一套完全自主运行的AIA层架构\n做到自动提出科研假设\n自动编写实验代码\n自动部署分布式计算\n自动分析实验指标\n最后自动递归迭代\n你注意看Discovery Loop的首期应用场景选择\n他们没有第一天就跑去见物理实验室\n而是选择吃自己的狗粮\n把机器学习研发本身作为第一个实验战场\n也就是说他们要用AI去自动设计\n自动搜索\n自动验证下一代超越Transformer的模型架构\n很多朋友可能会问\n用AI自动写代码做研究\n之前日本Sakana AI的项目不是也做过吗\n这正是我们接下来要拆解的关键\nSakana AI的项目本质上是一个\n应用层的agent流程封装\n它是拿现成的GPT4O或者Cloud外包一层prompt\n去模拟科学家写代码跑实验写论文\n它的上限完全被底下调用的通用大模型给封死了\n而Discovery Loop完全是另一个维度的存在\nJeff Dean和桑杰·格瓦马特\n是写出MapReduce Bigtable的超级系统架构师\n李光莲是AutoML的鼻祖\n奥里尔尔维尼亚尔斯\n是顶尖的神经网络搜索与多模态专家\n他们要做的是从底层分布式计算\n硬件算力调度\n神经架构搜索到算法闭环的系统级重构\n他们不是在用AI写论文\n他们是在用超级计算机集群\n让AI去做数以万计的硬件级与架构级实验\n直接在算法底层寻找下一个Transformer\n那么回到我们开头的悬念\n就是谷歌为什么不把这个超级团队留在内部\n而是支持他们独立成立一家公共利益公司呢\n因为谷歌算透了一笔账\n像AI自主研发下一代AI这种前沿探索\n风险极高\n烧钱极大\n失败率也极高\n如果放在谷歌Digemine内部\n研究员们每天要面对华尔街的财报审查和季度KPI\n就会被无限牵扯精力去搞产品适配\n通过把JFDN团队外部化\n谷歌既让这一帮顶尖天才\n摆脱了大公司病和商业汇报的枷锁\n又能通过初始投资方和独家GoCloud算力的身份\n牢牢锁定未来潜在的技术回流与云服务收益\n这一招叫做前沿高风险战略的外部结偶\n如果说JFDN的离职是把远期高风险探索拆分了出去\n那么接下来发生的事情\n才是对Jamlight产品生态影响最深远\n也是最暴力的一环\n接掌Google D.My的全局率颖的\n科雷·卡伍克·乔武\n是一个与Demis Hossabies\n风格截然不同的工程技术派\nDemis是国际象棋神童和神经科学家\n他充满着理想主义\n而科雷是把深度强化学习\n真正推向AlphaGo破局\n把WaveNet实施语音\n真正变成极易手机里语音引擎的超级工程操盘手\n在科雷掌舵之后\nGoogle G-Mind释放出了一个极其明确的信号\n就是全面放弃学术包袱\n转向极致工程眼镜与大规模商业化暴击\n在他的主持下\nGamlight的底层架构\n首先,最根本的變化是,徹底打破對預训訓參數崇拜的單位幻想,全面轉向後訓練與推理期計算的深度融合。\n在过去两年 行业盲目堆叠预训练参数\n但预训练越往后 编辑效应递减越严重\n算力成本高到令人发指\n柯雷结合他在DQN强化学习领域的积累\n给Gemla引入了过程奖励模型PRM\n与蒙特卡洛书搜索MCTS\n打个比方 传统的语言模型\n就像在考场上没有任何草稿纸\n靠直觉备答案的考生\n遇到简单的题那是没问题的\n但遇到复杂逻辑推导或者长流程编程\n中间要是错一步后面就全乱了\n而PRM与MCTS的融合\n相当于给GEMLA配了一张无线扩展的虚拟草稿纸\n和一位微观监考官\n在推理的每一步\nPRM都会对它的逻辑合理性进行毫秒级评分\n同时MCTS数搜索会在后台预判\n未来几十步的走向\n如果发现某个分支走向死胡同\n模型会在推理阶段自动撤回并重新搜索最右键\n这意味着在不用无限膨胀预讯链参数的前提下\nGEMLINE处理高难度编程\n复杂数学推导和A乘的规划的成功率\n称指数级提升\n紧接着第二种转变在于\n系统级推理效率的极限拉链\n那就是构建GEMINI Spark端云协同分层架构\n这里稍微提一下我个人对Spark的使用体验\n虽然它还是Beta版本\n但先说结论\n我觉得它超级好用\n我会单独出一期视频\n聊一下Gemla Spark和Hermis Agent的配合使用\n作为Google订阅的长期体验者\n我明显感觉最近Gemla的使用限额提高了\n他们虽然没有发布像Fable 5这类的模型\n但是无论是Gemla 3.6 Flash\n还是Gemla 3.1 Pro都变得更聪明了\n好 我们继续这个第二重转变\n在商业高并发场景中\n用户问一句今天天气怎么样\n如果都要调用一次千亿参数的单体大模型\n显存和电力开销能把大厂烧穿\n克雷推动的Gemlet Spark架构\n把整套AI系统做成了像特种部队一样的分层调度\n前端部署及轻量及低延迟的Gemlet Spark代理模型\n负责高频意图识别\n上下门过滤和轻量任务\n只有当前线判定\n涉及复杂长链条逻辑的时候\n才会无缝调动云端期间大模型\n这种分层架构\n直接把单位Token成本打了下来\n这才是支撑数亿并发的商业硬实力\n而第三种变化\n则是科雷将他的看家本领\nVivnet原生音频与流逝生成\n深度融合到了詹姆賴的全模态底座中\n以往的多模态AI\n往往是文本模型外挂语音识别\n与语音合成模块的拼接车\n延迟高 语调生硬\n而柯雷主导下的詹姆賴\n实现了真正的原生全模态流逝交互\n不仅能实现毫秒级\n实时音视频双向对打\n还能在听懂说话的同时\n自主执行代码\n甚至在手机屏幕上动态生成\nUI界面与交互空间\n这已经不是在聊天了\n这是直接把自然语言变成了实时软件生成的界面\n那底层架构演进的再漂亮\n如果不能转化为商业流水\n在今天也是一张废纸\n科雷·卡伍克·乔鲁这次最让人惊悚的调整\n在于他不仅管科研\n还同时兼任了Gemline产品与开发者生态的最高指挥官\n在过去的大厂体制里\n科研实验室和商品产品部门是完全脱节的\n实验室发了论文\n产品部门要花大半年去沟通打包\n而柯雷一把抓之后\n直接祭出了商业落地的三大杀招\n在终端入口上\n他们开始加速清洗传统的语音助手\n全面用GM来替换掉Android和VLOS上\n那个指挥社闹钟的assistant\n赋予他操作系统级的最高上下文读取权限\n与跨应用API调用能力\n直接把它打造成操作系统的中央控制大脑\n比如在ValOS手表或者车载系统里\n你只需要说帮我安排今晚与张总的商务晚餐\n并准备一份他公司最新财报的背景简报\nGamlight就会自动调取邮件和地图订餐厅\n同时在云端提取财报生成简报推送给你\n与此同时\n他们还大幅压缩了科研成果\n转化到商业API的时间\n在谷歌云Vertex AI和开发者API上\n前沿团队在Achic代码生成\n自动模型评估和推理加速上的突破\n能以周级别的速度直接部署上线\n让科研成果在第一秒\n直接转化为商业账单\n形成技术产品与算力的超级费轮\n更关键的是\n谷歌正在悄无声息地改变自己的盈利模式\n当用户不再翻页点击搜索广告链接\n而是让Gamlet给出最终答案时\n谷歌的解法是\n直接占据agent决策路径的入口\n当Gamlet在每日简报或者个人助理里\n帮你规划行程推荐服务时\n它在生成方案的同时\n精准接入商业API入口\n谷歌由此从一个信息检索广告商\n变成了全局智能决策的开门人\n抽取服务分账和佣金\n拆决员Jeff Dean的离职创业\n与柯磊掌舵下的Deep Mind重组\n我们终于能看清这场硅谷AI大戏的全貌了\n硅谷的AI战争正是告别了依靠概念和论文引用量\n获取高估值的学术竞赛阶段\n全面跨入了拼推理成本\n拼系统稳定性\n拼终端控制力\n以及拼商业造学能力的工程与商业深水去\n对于我们所有在做AI研发架构设计\n或者技术创业的朋友来说\n谷歌的这一系列辩证\n留下了三条极其宝贵的工程启发\n首先不要盲目迷信\n纯域训练参数的单位堆叠\n多去研究推理器计算\n通过过程奖励模型打分与蒙特卡洛数搜索\n让模型在推理阶段学会自我纠错\n这是低成本实现逻辑跨越的唯一解法\n其次要学会拥抱端云分层架构\n平气无论任务大小\n一力调用千亿大模型的粗暴思维\n学习Jamler Spark的设计\n用轻量级模型做好前置意图路由\n才能在高并发任务中守住成本底线\n最后我们需要重新思考软件入口的定义\n未来的软件竞争不再是争夺\n手机屏幕上的APP图标位置\n而是争夺能否被系统及agent\nAPI决策链所调用\n技术不相信幻觉\n商业不相信眼泪\n这里是All Insight\n我们下期见",
  "language": "zh",
  "duration": 929.100063,
  "segments": [
    {
      "start": 0.0,
      "end": 29.1,
      "text": "歡迎回到On Insight。如果我們要選一個節點作為全球AI產業,從學術夢幻期徹底跨入現實暴擊期的分水裡,我敢打賭,絕不會是哪家大廠又在PPT裡發布了多大規模參數的模型,基準測試又突破了多高的分數。",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.13953972816467286,
      "no_speech_prob": 1.1523658763334943e-11,
      "compression_ratio": 1.0871369294605808,
      "source": "whisper"
    },
    {
      "start": 29.1,
      "end": 31.94,
      "text": "真正的变局往往发生在那些",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07450612737328967,
      "no_speech_prob": 1.2842626267994817e-11,
      "compression_ratio": 1.2176656151419558,
      "source": "whisper"
    },
    {
      "start": 31.94,
      "end": 33.68,
      "text": "看起来极度安静",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07450612737328967,
      "no_speech_prob": 1.2842626267994817e-11,
      "compression_ratio": 1.2176656151419558,
      "source": "whisper"
    },
    {
      "start": 33.68,
      "end": 37.24,
      "text": "甚至带着几分诡异的高层人事变动力",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07450612737328967,
      "no_speech_prob": 1.2842626267994817e-11,
      "compression_ratio": 1.2176656151419558,
      "source": "whisper"
    },
    {
      "start": 37.24,
      "end": 38.34,
      "text": "就在最近",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07450612737328967,
      "no_speech_prob": 1.2842626267994817e-11,
      "compression_ratio": 1.2176656151419558,
      "source": "whisper"
    },
    {
      "start": 38.34,
      "end": 39.78,
      "text": "谷歌内部发生了一场",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07450612737328967,
      "no_speech_prob": 1.2842626267994817e-11,
      "compression_ratio": 1.2176656151419558,
      "source": "whisper"
    },
    {
      "start": 39.78,
      "end": 43.4,
      "text": "足以重塑整个硅谷底层格局的震荡",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07450612737328967,
      "no_speech_prob": 1.2842626267994817e-11,
      "compression_ratio": 1.2176656151419558,
      "source": "whisper"
    },
    {
      "start": 43.4,
      "end": 46.04,
      "text": "在谷歌效力了整整27年",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07450612737328967,
      "no_speech_prob": 1.2842626267994817e-11,
      "compression_ratio": 1.2176656151419558,
      "source": "whisper"
    },
    {
      "start": 46.04,
      "end": 47.5,
      "text": "被全球程序员",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07450612737328967,
      "no_speech_prob": 1.2842626267994817e-11,
      "compression_ratio": 1.2176656151419558,
      "source": "whisper"
    },
    {
      "start": 47.5,
      "end": 50.54,
      "text": "尊为代码之神的谷歌首席科学家",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07450612737328967,
      "no_speech_prob": 1.2842626267994817e-11,
      "compression_ratio": 1.2176656151419558,
      "source": "whisper"
    },
    {
      "start": 50.54,
      "end": 52.44,
      "text": "Jeff Dean正式宣布离职",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07450612737328967,
      "no_speech_prob": 1.2842626267994817e-11,
      "compression_ratio": 1.2176656151419558,
      "source": "whisper"
    },
    {
      "start": 52.44,
      "end": 55.26,
      "text": "如果这只是一位老将退隐的话",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07450612737328967,
      "no_speech_prob": 1.2842626267994817e-11,
      "compression_ratio": 1.2176656151419558,
      "source": "whisper"
    },
    {
      "start": 55.26,
      "end": 58.18,
      "text": "倒也不至于让硅谷各大投资机构",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07450612737328967,
      "no_speech_prob": 1.2842626267994817e-11,
      "compression_ratio": 1.2176656151419558,
      "source": "whisper"
    },
    {
      "start": 58.18,
      "end": 59.96,
      "text": "和安全部门彻夜未免",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15480281460669734,
      "no_speech_prob": 2.012770505821493e-11,
      "compression_ratio": 1.0978593272171253,
      "source": "whisper"
    },
    {
      "start": 59.96,
      "end": 62.32,
      "text": "真正认识背后发紧的是",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15480281460669734,
      "no_speech_prob": 2.012770505821493e-11,
      "compression_ratio": 1.0978593272171253,
      "source": "whisper"
    },
    {
      "start": 62.32,
      "end": 63.86,
      "text": "Jav Dien离职的时候",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15480281460669734,
      "no_speech_prob": 2.012770505821493e-11,
      "compression_ratio": 1.0978593272171253,
      "source": "whisper"
    },
    {
      "start": 63.86,
      "end": 65.52,
      "text": "带走了一个堪称",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15480281460669734,
      "no_speech_prob": 2.012770505821493e-11,
      "compression_ratio": 1.0978593272171253,
      "source": "whisper"
    },
    {
      "start": 65.52,
      "end": 69.18,
      "text": "硅谷全明星的豪华降维打击阵容",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15480281460669734,
      "no_speech_prob": 2.012770505821493e-11,
      "compression_ratio": 1.0978593272171253,
      "source": "whisper"
    },
    {
      "start": 69.18,
      "end": 71.7,
      "text": "这其中包括与Jav Dien合作",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15480281460669734,
      "no_speech_prob": 2.012770505821493e-11,
      "compression_ratio": 1.0978593272171253,
      "source": "whisper"
    },
    {
      "start": 71.7,
      "end": 74.66,
      "text": "20多年的底层系统传奇搭档",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15480281460669734,
      "no_speech_prob": 2.012770505821493e-11,
      "compression_ratio": 1.0978593272171253,
      "source": "whisper"
    },
    {
      "start": 74.66,
      "end": 76.36,
      "text": "桑杰·格马瓦特",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15480281460669734,
      "no_speech_prob": 2.012770505821493e-11,
      "compression_ratio": 1.0978593272171253,
      "source": "whisper"
    },
    {
      "start": 76.36,
      "end": 79.14,
      "text": "还有Google DeepMind研究副总裁",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15480281460669734,
      "no_speech_prob": 2.012770505821493e-11,
      "compression_ratio": 1.0978593272171253,
      "source": "whisper"
    },
    {
      "start": 79.14,
      "end": 82.14,
      "text": "Jemline模型最核心的技术操盘手",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15480281460669734,
      "no_speech_prob": 2.012770505821493e-11,
      "compression_ratio": 1.0978593272171253,
      "source": "whisper"
    },
    {
      "start": 82.14,
      "end": 83.96,
      "text": "奥里奥尔·维尼亚尔斯",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15480281460669734,
      "no_speech_prob": 2.012770505821493e-11,
      "compression_ratio": 1.0978593272171253,
      "source": "whisper"
    },
    {
      "start": 83.96,
      "end": 86.4,
      "text": "以及AutoML领域先驱",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15480281460669734,
      "no_speech_prob": 2.012770505821493e-11,
      "compression_ratio": 1.0978593272171253,
      "source": "whisper"
    },
    {
      "start": 86.4,
      "end": 88.98,
      "text": "和Google Brain的联合创始人黎光莲",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08651695251464844,
      "no_speech_prob": 1.7370429747365357e-11,
      "compression_ratio": 1.1904761904761905,
      "source": "whisper"
    },
    {
      "start": 88.98,
      "end": 91.06,
      "text": "这四个人加在一起",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08651695251464844,
      "no_speech_prob": 1.7370429747365357e-11,
      "compression_ratio": 1.1904761904761905,
      "source": "whisper"
    },
    {
      "start": 91.06,
      "end": 93.2,
      "text": "几乎就是半个谷歌AI",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08651695251464844,
      "no_speech_prob": 1.7370429747365357e-11,
      "compression_ratio": 1.1904761904761905,
      "source": "whisper"
    },
    {
      "start": 93.2,
      "end": 96.04,
      "text": "与底层分布式系统的技术灵魂",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08651695251464844,
      "no_speech_prob": 1.7370429747365357e-11,
      "compression_ratio": 1.1904761904761905,
      "source": "whisper"
    },
    {
      "start": 96.04,
      "end": 99.6,
      "text": "他们组建了一家名为Discovery Loop的新公司",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08651695251464844,
      "no_speech_prob": 1.7370429747365357e-11,
      "compression_ratio": 1.1904761904761905,
      "source": "whisper"
    },
    {
      "start": 99.6,
      "end": 100.68,
      "text": "按理说",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08651695251464844,
      "no_speech_prob": 1.7370429747365357e-11,
      "compression_ratio": 1.1904761904761905,
      "source": "whisper"
    },
    {
      "start": 100.68,
      "end": 103.62,
      "text": "面对如此核心的技术骨干离职",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08651695251464844,
      "no_speech_prob": 1.7370429747365357e-11,
      "compression_ratio": 1.1904761904761905,
      "source": "whisper"
    },
    {
      "start": 103.62,
      "end": 106.7,
      "text": "谷歌应该爆发一场血腥的公关战",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08651695251464844,
      "no_speech_prob": 1.7370429747365357e-11,
      "compression_ratio": 1.1904761904761905,
      "source": "whisper"
    },
    {
      "start": 106.7,
      "end": 107.3,
      "text": "对吧",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08651695251464844,
      "no_speech_prob": 1.7370429747365357e-11,
      "compression_ratio": 1.1904761904761905,
      "source": "whisper"
    },
    {
      "start": 107.3,
      "end": 109.0,
      "text": "但诡异的事情来了",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08651695251464844,
      "no_speech_prob": 1.7370429747365357e-11,
      "compression_ratio": 1.1904761904761905,
      "source": "whisper"
    },
    {
      "start": 109.0,
      "end": 111.18,
      "text": "在继续后面的内容之前",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08651695251464844,
      "no_speech_prob": 1.7370429747365357e-11,
      "compression_ratio": 1.1904761904761905,
      "source": "whisper"
    },
    {
      "start": 111.18,
      "end": 113.26,
      "text": "我想先请你帮我一个小忙",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08651695251464844,
      "no_speech_prob": 1.7370429747365357e-11,
      "compression_ratio": 1.1904761904761905,
      "source": "whisper"
    },
    {
      "start": 113.26,
      "end": 115.1,
      "text": "如果你经常观看我的视频",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.08651695251464844,
      "no_speech_prob": 1.7370429747365357e-11,
      "compression_ratio": 1.1904761904761905,
      "source": "whisper"
    },
    {
      "start": 115.1,
      "end": 116.48,
      "text": "但还没有订阅我的频道",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.0802341586067563,
      "no_speech_prob": 1.5635298611371695e-11,
      "compression_ratio": 1.386138613861386,
      "source": "whisper"
    },
    {
      "start": 116.48,
      "end": 118.04,
      "text": "欢迎点击订阅",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.0802341586067563,
      "no_speech_prob": 1.5635298611371695e-11,
      "compression_ratio": 1.386138613861386,
      "source": "whisper"
    },
    {
      "start": 118.04,
      "end": 122.46,
      "text": "这样你就不会错过后续更多的科学资讯和深度解析",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.0802341586067563,
      "no_speech_prob": 1.5635298611371695e-11,
      "compression_ratio": 1.386138613861386,
      "source": "whisper"
    },
    {
      "start": 122.46,
      "end": 125.8,
      "text": "那如果你觉得这期视频对你有所帮助",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.0802341586067563,
      "no_speech_prob": 1.5635298611371695e-11,
      "compression_ratio": 1.386138613861386,
      "source": "whisper"
    },
    {
      "start": 125.8,
      "end": 127.58,
      "text": "也请帮忙点赞和分享",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.0802341586067563,
      "no_speech_prob": 1.5635298611371695e-11,
      "compression_ratio": 1.386138613861386,
      "source": "whisper"
    },
    {
      "start": 127.58,
      "end": 129.36,
      "text": "让更多人看到这些内容",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.0802341586067563,
      "no_speech_prob": 1.5635298611371695e-11,
      "compression_ratio": 1.386138613861386,
      "source": "whisper"
    },
    {
      "start": 129.36,
      "end": 132.28,
      "text": "那如果你希望给予我更多的支持",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.0802341586067563,
      "no_speech_prob": 1.5635298611371695e-11,
      "compression_ratio": 1.386138613861386,
      "source": "whisper"
    },
    {
      "start": 132.28,
      "end": 133.88,
      "text": "也欢迎加入我的会员",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.0802341586067563,
      "no_speech_prob": 1.5635298611371695e-11,
      "compression_ratio": 1.386138613861386,
      "source": "whisper"
    },
    {
      "start": 133.88,
      "end": 135.74,
      "text": "每月一杯咖啡的钱",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.0802341586067563,
      "no_speech_prob": 1.5635298611371695e-11,
      "compression_ratio": 1.386138613861386,
      "source": "whisper"
    },
    {
      "start": 135.74,
      "end": 138.44,
      "text": "就是对我持续创作最大的支持",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.0802341586067563,
      "no_speech_prob": 1.5635298611371695e-11,
      "compression_ratio": 1.386138613861386,
      "source": "whisper"
    },
    {
      "start": 138.44,
      "end": 139.9,
      "text": "好的我们继续",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.0802341586067563,
      "no_speech_prob": 1.5635298611371695e-11,
      "compression_ratio": 1.386138613861386,
      "source": "whisper"
    },
    {
      "start": 139.9,
      "end": 141.84,
      "text": "这个诡异的事情就是",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.0802341586067563,
      "no_speech_prob": 1.5635298611371695e-11,
      "compression_ratio": 1.386138613861386,
      "source": "whisper"
    },
    {
      "start": 141.84,
      "end": 143.76,
      "text": "谷歌不仅没有阻拦",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.0802341586067563,
      "no_speech_prob": 1.5635298611371695e-11,
      "compression_ratio": 1.386138613861386,
      "source": "whisper"
    },
    {
      "start": 143.76,
      "end": 147.98,
      "text": "反而第一时间成为了Discovery Loop的初始投资方",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11725490821335843,
      "no_speech_prob": 1.6349982132068952e-11,
      "compression_ratio": 1.1476923076923078,
      "source": "whisper"
    },
    {
      "start": 147.98,
      "end": 152.24,
      "text": "并且双手奉上了Google Cloud的超级算力集群",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11725490821335843,
      "no_speech_prob": 1.6349982132068952e-11,
      "compression_ratio": 1.1476923076923078,
      "source": "whisper"
    },
    {
      "start": 152.24,
      "end": 156.54,
      "text": "双方直接签下了长期云服务绑定协议",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11725490821335843,
      "no_speech_prob": 1.6349982132068952e-11,
      "compression_ratio": 1.1476923076923078,
      "source": "whisper"
    },
    {
      "start": 156.54,
      "end": 157.12,
      "text": "不仅如此",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11725490821335843,
      "no_speech_prob": 1.6349982132068952e-11,
      "compression_ratio": 1.1476923076923078,
      "source": "whisper"
    },
    {
      "start": 157.12,
      "end": 158.34,
      "text": "就在同一天",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11725490821335843,
      "no_speech_prob": 1.6349982132068952e-11,
      "compression_ratio": 1.1476923076923078,
      "source": "whisper"
    },
    {
      "start": 158.34,
      "end": 160.68,
      "text": "Google DeepMind内部也完成了一场",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11725490821335843,
      "no_speech_prob": 1.6349982132068952e-11,
      "compression_ratio": 1.1476923076923078,
      "source": "whisper"
    },
    {
      "start": 160.68,
      "end": 163.42,
      "text": "近乎手术刀级别的高层重组",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11725490821335843,
      "no_speech_prob": 1.6349982132068952e-11,
      "compression_ratio": 1.1476923076923078,
      "source": "whisper"
    },
    {
      "start": 163.42,
      "end": 165.96,
      "text": "DeepMind的灵魂人物Demith Hossabies",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11725490821335843,
      "no_speech_prob": 1.6349982132068952e-11,
      "compression_ratio": 1.1476923076923078,
      "source": "whisper"
    },
    {
      "start": 165.96,
      "end": 168.94,
      "text": "正式卸任DeepMind CEO的日常管理",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11725490821335843,
      "no_speech_prob": 1.6349982132068952e-11,
      "compression_ratio": 1.1476923076923078,
      "source": "whisper"
    },
    {
      "start": 168.94,
      "end": 171.46,
      "text": "身任Alphabet的首席科学家",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.11725490821335843,
      "no_speech_prob": 1.6349982132068952e-11,
      "compression_ratio": 1.1476923076923078,
      "source": "whisper"
    },
    {
      "start": 171.46,
      "end": 174.5,
      "text": "去操盘更为宏大的AGI理论中举",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1439477284749349,
      "no_speech_prob": 1.6168512709247018e-11,
      "compression_ratio": 1.1412103746397695,
      "source": "whisper"
    },
    {
      "start": 174.5,
      "end": 177.88,
      "text": "而接替他全盘操盘Google DeepMind的新帅",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1439477284749349,
      "no_speech_prob": 1.6168512709247018e-11,
      "compression_ratio": 1.1412103746397695,
      "source": "whisper"
    },
    {
      "start": 177.88,
      "end": 180.12,
      "text": "是原DevMind的CTO",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1439477284749349,
      "no_speech_prob": 1.6168512709247018e-11,
      "compression_ratio": 1.1412103746397695,
      "source": "whisper"
    },
    {
      "start": 180.12,
      "end": 182.52,
      "text": "DQN与WaveNet的发明人",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1439477284749349,
      "no_speech_prob": 1.6168512709247018e-11,
      "compression_ratio": 1.1412103746397695,
      "source": "whisper"
    },
    {
      "start": 182.52,
      "end": 184.02,
      "text": "科雷·卡伍克·乔鲲鲁",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1439477284749349,
      "no_speech_prob": 1.6168512709247018e-11,
      "compression_ratio": 1.1412103746397695,
      "source": "whisper"
    },
    {
      "start": 184.02,
      "end": 186.56,
      "text": "他越过了所有中间管理层",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1439477284749349,
      "no_speech_prob": 1.6168512709247018e-11,
      "compression_ratio": 1.1412103746397695,
      "source": "whisper"
    },
    {
      "start": 186.56,
      "end": 189.44,
      "text": "直接向谷歌CEO桑德尔PX1汇报",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1439477284749349,
      "no_speech_prob": 1.6168512709247018e-11,
      "compression_ratio": 1.1412103746397695,
      "source": "whisper"
    },
    {
      "start": 189.44,
      "end": 190.66,
      "text": "听到这里",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1439477284749349,
      "no_speech_prob": 1.6168512709247018e-11,
      "compression_ratio": 1.1412103746397695,
      "source": "whisper"
    },
    {
      "start": 190.66,
      "end": 194.42,
      "text": "你脑海里一定盘旋着一个极其巨大的悬念",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1439477284749349,
      "no_speech_prob": 1.6168512709247018e-11,
      "compression_ratio": 1.1412103746397695,
      "source": "whisper"
    },
    {
      "start": 194.42,
      "end": 196.94,
      "text": "就是谷歌为什么会心甘情愿的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1439477284749349,
      "no_speech_prob": 1.6168512709247018e-11,
      "compression_ratio": 1.1412103746397695,
      "source": "whisper"
    },
    {
      "start": 196.94,
      "end": 199.22,
      "text": "让自己的顶尖大脑集体外逃",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1439477284749349,
      "no_speech_prob": 1.6168512709247018e-11,
      "compression_ratio": 1.1412103746397695,
      "source": "whisper"
    },
    {
      "start": 199.22,
      "end": 201.36,
      "text": "甚至还要掏钱掏算力",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1439477284749349,
      "no_speech_prob": 1.6168512709247018e-11,
      "compression_ratio": 1.1412103746397695,
      "source": "whisper"
    },
    {
      "start": 201.36,
      "end": 202.74,
      "text": "帮他们去创业呢",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12129249572753906,
      "no_speech_prob": 1.5878588374707014e-11,
      "compression_ratio": 1.1246006389776357,
      "source": "whisper"
    },
    {
      "start": 202.74,
      "end": 205.72,
      "text": "而那位既懂DQN强化学习",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12129249572753906,
      "no_speech_prob": 1.5878588374707014e-11,
      "compression_ratio": 1.1246006389776357,
      "source": "whisper"
    },
    {
      "start": 205.72,
      "end": 208.88,
      "text": "又懂Vivnet原生语音的工程狂人",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12129249572753906,
      "no_speech_prob": 1.5878588374707014e-11,
      "compression_ratio": 1.1246006389776357,
      "source": "whisper"
    },
    {
      "start": 208.88,
      "end": 210.26,
      "text": "柯雷接棒之后",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12129249572753906,
      "no_speech_prob": 1.5878588374707014e-11,
      "compression_ratio": 1.1246006389776357,
      "source": "whisper"
    },
    {
      "start": 210.26,
      "end": 212.84,
      "text": "到底要把Gemla带向何方呢",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12129249572753906,
      "no_speech_prob": 1.5878588374707014e-11,
      "compression_ratio": 1.1246006389776357,
      "source": "whisper"
    },
    {
      "start": 212.84,
      "end": 215.3,
      "text": "今天我们就有一期节目",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12129249572753906,
      "no_speech_prob": 1.5878588374707014e-11,
      "compression_ratio": 1.1246006389776357,
      "source": "whisper"
    },
    {
      "start": 215.3,
      "end": 218.46,
      "text": "彻底拆解这场看似是高层离职潮",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12129249572753906,
      "no_speech_prob": 1.5878588374707014e-11,
      "compression_ratio": 1.1246006389776357,
      "source": "whisper"
    },
    {
      "start": 218.46,
      "end": 221.72,
      "text": "实则是谷歌AI技术大分流的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12129249572753906,
      "no_speech_prob": 1.5878588374707014e-11,
      "compression_ratio": 1.1246006389776357,
      "source": "whisper"
    },
    {
      "start": 221.72,
      "end": 223.14,
      "text": "硅谷终极棋局",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12129249572753906,
      "no_speech_prob": 1.5878588374707014e-11,
      "compression_ratio": 1.1246006389776357,
      "source": "whisper"
    },
    {
      "start": 223.14,
      "end": 224.82,
      "text": "要解开这个谜团",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12129249572753906,
      "no_speech_prob": 1.5878588374707014e-11,
      "compression_ratio": 1.1246006389776357,
      "source": "whisper"
    },
    {
      "start": 224.82,
      "end": 226.26,
      "text": "我们首先得搞清楚",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12129249572753906,
      "no_speech_prob": 1.5878588374707014e-11,
      "compression_ratio": 1.1246006389776357,
      "source": "whisper"
    },
    {
      "start": 226.26,
      "end": 229.36,
      "text": "Jeff Dean带着这帮神仙阵容杀出去",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.12129249572753906,
      "no_speech_prob": 1.5878588374707014e-11,
      "compression_ratio": 1.1246006389776357,
      "source": "whisper"
    },
    {
      "start": 229.36,
      "end": 231.72,
      "text": "到底要去干一件什么样的事情",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1065929942048354,
      "no_speech_prob": 1.8470898421885096e-11,
      "compression_ratio": 1.170807453416149,
      "source": "whisper"
    },
    {
      "start": 231.72,
      "end": 234.22,
      "text": "在Discovery Loop的官宣里",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1065929942048354,
      "no_speech_prob": 1.8470898421885096e-11,
      "compression_ratio": 1.170807453416149,
      "source": "whisper"
    },
    {
      "start": 234.22,
      "end": 236.96,
      "text": "Jeff Dean直接描绘了公司的愿景",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1065929942048354,
      "no_speech_prob": 1.8470898421885096e-11,
      "compression_ratio": 1.170807453416149,
      "source": "whisper"
    },
    {
      "start": 236.96,
      "end": 240.6,
      "text": "那就是构建自动化的科学方法闭环",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1065929942048354,
      "no_speech_prob": 1.8470898421885096e-11,
      "compression_ratio": 1.170807453416149,
      "source": "whisper"
    },
    {
      "start": 240.6,
      "end": 241.9,
      "text": "简单来说",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1065929942048354,
      "no_speech_prob": 1.8470898421885096e-11,
      "compression_ratio": 1.170807453416149,
      "source": "whisper"
    },
    {
      "start": 241.9,
      "end": 243.38,
      "text": "过去几百年",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1065929942048354,
      "no_speech_prob": 1.8470898421885096e-11,
      "compression_ratio": 1.170807453416149,
      "source": "whisper"
    },
    {
      "start": 243.38,
      "end": 245.5,
      "text": "人类探索未知的方式",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1065929942048354,
      "no_speech_prob": 1.8470898421885096e-11,
      "compression_ratio": 1.170807453416149,
      "source": "whisper"
    },
    {
      "start": 245.5,
      "end": 247.3,
      "text": "本质上都是串行实验",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1065929942048354,
      "no_speech_prob": 1.8470898421885096e-11,
      "compression_ratio": 1.170807453416149,
      "source": "whisper"
    },
    {
      "start": 247.3,
      "end": 249.6,
      "text": "就是科学家提出假设",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1065929942048354,
      "no_speech_prob": 1.8470898421885096e-11,
      "compression_ratio": 1.170807453416149,
      "source": "whisper"
    },
    {
      "start": 249.6,
      "end": 250.88,
      "text": "亲手写代码",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1065929942048354,
      "no_speech_prob": 1.8470898421885096e-11,
      "compression_ratio": 1.170807453416149,
      "source": "whisper"
    },
    {
      "start": 250.88,
      "end": 251.76,
      "text": "跑一次实验",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1065929942048354,
      "no_speech_prob": 1.8470898421885096e-11,
      "compression_ratio": 1.170807453416149,
      "source": "whisper"
    },
    {
      "start": 251.76,
      "end": 253.16,
      "text": "等几天看结果",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1065929942048354,
      "no_speech_prob": 1.8470898421885096e-11,
      "compression_ratio": 1.170807453416149,
      "source": "whisper"
    },
    {
      "start": 253.16,
      "end": 253.9,
      "text": "分析数据",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1065929942048354,
      "no_speech_prob": 1.8470898421885096e-11,
      "compression_ratio": 1.170807453416149,
      "source": "whisper"
    },
    {
      "start": 253.9,
      "end": 255.88,
      "text": "然后再手动修改参数",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1065929942048354,
      "no_speech_prob": 1.8470898421885096e-11,
      "compression_ratio": 1.170807453416149,
      "source": "whisper"
    },
    {
      "start": 255.88,
      "end": 256.6,
      "text": "跑下一次",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1065929942048354,
      "no_speech_prob": 1.8470898421885096e-11,
      "compression_ratio": 1.170807453416149,
      "source": "whisper"
    },
    {
      "start": 256.6,
      "end": 258.3,
      "text": "在这套流程里",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1065929942048354,
      "no_speech_prob": 1.8470898421885096e-11,
      "compression_ratio": 1.170807453416149,
      "source": "whisper"
    },
    {
      "start": 258.3,
      "end": 260.72,
      "text": "最瓶颈的变量不是算力",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07288794571094298,
      "no_speech_prob": 1.4329096936771535e-11,
      "compression_ratio": 1.255663430420712,
      "source": "whisper"
    },
    {
      "start": 260.72,
      "end": 265.6,
      "text": "而是人类脑力的思考速度与手动操作的物理实验",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07288794571094298,
      "no_speech_prob": 1.4329096936771535e-11,
      "compression_ratio": 1.255663430420712,
      "source": "whisper"
    },
    {
      "start": 265.6,
      "end": 271.28,
      "text": "而Discovery Loop要构建一套完全自主运行的AIA层架构",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07288794571094298,
      "no_speech_prob": 1.4329096936771535e-11,
      "compression_ratio": 1.255663430420712,
      "source": "whisper"
    },
    {
      "start": 271.28,
      "end": 273.72,
      "text": "做到自动提出科研假设",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07288794571094298,
      "no_speech_prob": 1.4329096936771535e-11,
      "compression_ratio": 1.255663430420712,
      "source": "whisper"
    },
    {
      "start": 273.72,
      "end": 275.62,
      "text": "自动编写实验代码",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07288794571094298,
      "no_speech_prob": 1.4329096936771535e-11,
      "compression_ratio": 1.255663430420712,
      "source": "whisper"
    },
    {
      "start": 275.62,
      "end": 277.4,
      "text": "自动部署分布式计算",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07288794571094298,
      "no_speech_prob": 1.4329096936771535e-11,
      "compression_ratio": 1.255663430420712,
      "source": "whisper"
    },
    {
      "start": 277.4,
      "end": 279.22,
      "text": "自动分析实验指标",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07288794571094298,
      "no_speech_prob": 1.4329096936771535e-11,
      "compression_ratio": 1.255663430420712,
      "source": "whisper"
    },
    {
      "start": 279.22,
      "end": 281.16,
      "text": "最后自动递归迭代",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07288794571094298,
      "no_speech_prob": 1.4329096936771535e-11,
      "compression_ratio": 1.255663430420712,
      "source": "whisper"
    },
    {
      "start": 281.16,
      "end": 285.16,
      "text": "你注意看Discovery Loop的首期应用场景选择",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07288794571094298,
      "no_speech_prob": 1.4329096936771535e-11,
      "compression_ratio": 1.255663430420712,
      "source": "whisper"
    },
    {
      "start": 285.16,
      "end": 288.02,
      "text": "他们没有第一天就跑去见物理实验室",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07288794571094298,
      "no_speech_prob": 1.4329096936771535e-11,
      "compression_ratio": 1.255663430420712,
      "source": "whisper"
    },
    {
      "start": 288.02,
      "end": 290.22,
      "text": "而是选择吃自己的狗粮",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.06696993188013005,
      "no_speech_prob": 1.2210848651661443e-11,
      "compression_ratio": 1.2259136212624584,
      "source": "whisper"
    },
    {
      "start": 290.22,
      "end": 294.32,
      "text": "把机器学习研发本身作为第一个实验战场",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.06696993188013005,
      "no_speech_prob": 1.2210848651661443e-11,
      "compression_ratio": 1.2259136212624584,
      "source": "whisper"
    },
    {
      "start": 294.32,
      "end": 297.58,
      "text": "也就是说他们要用AI去自动设计",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.06696993188013005,
      "no_speech_prob": 1.2210848651661443e-11,
      "compression_ratio": 1.2259136212624584,
      "source": "whisper"
    },
    {
      "start": 297.58,
      "end": 298.56,
      "text": "自动搜索",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.06696993188013005,
      "no_speech_prob": 1.2210848651661443e-11,
      "compression_ratio": 1.2259136212624584,
      "source": "whisper"
    },
    {
      "start": 298.56,
      "end": 302.28,
      "text": "自动验证下一代超越Transformer的模型架构",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.06696993188013005,
      "no_speech_prob": 1.2210848651661443e-11,
      "compression_ratio": 1.2259136212624584,
      "source": "whisper"
    },
    {
      "start": 302.28,
      "end": 304.14,
      "text": "很多朋友可能会问",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.06696993188013005,
      "no_speech_prob": 1.2210848651661443e-11,
      "compression_ratio": 1.2259136212624584,
      "source": "whisper"
    },
    {
      "start": 304.14,
      "end": 306.84,
      "text": "用AI自动写代码做研究",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.06696993188013005,
      "no_speech_prob": 1.2210848651661443e-11,
      "compression_ratio": 1.2259136212624584,
      "source": "whisper"
    },
    {
      "start": 306.84,
      "end": 310.14,
      "text": "之前日本Sakana AI的项目不是也做过吗",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.06696993188013005,
      "no_speech_prob": 1.2210848651661443e-11,
      "compression_ratio": 1.2259136212624584,
      "source": "whisper"
    },
    {
      "start": 310.14,
      "end": 313.18,
      "text": "这正是我们接下来要拆解的关键",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.06696993188013005,
      "no_speech_prob": 1.2210848651661443e-11,
      "compression_ratio": 1.2259136212624584,
      "source": "whisper"
    },
    {
      "start": 313.18,
      "end": 315.74,
      "text": "Sakana AI的项目本质上是一个",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.06696993188013005,
      "no_speech_prob": 1.2210848651661443e-11,
      "compression_ratio": 1.2259136212624584,
      "source": "whisper"
    },
    {
      "start": 315.74,
      "end": 318.1,
      "text": "应用层的agent流程封装",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1313201758512266,
      "no_speech_prob": 1.5682779727632656e-11,
      "compression_ratio": 1.0838509316770186,
      "source": "whisper"
    },
    {
      "start": 318.1,
      "end": 322.7,
      "text": "它是拿现成的GPT4O或者Cloud外包一层prompt",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1313201758512266,
      "no_speech_prob": 1.5682779727632656e-11,
      "compression_ratio": 1.0838509316770186,
      "source": "whisper"
    },
    {
      "start": 322.7,
      "end": 326.3,
      "text": "去模拟科学家写代码跑实验写论文",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1313201758512266,
      "no_speech_prob": 1.5682779727632656e-11,
      "compression_ratio": 1.0838509316770186,
      "source": "whisper"
    },
    {
      "start": 326.3,
      "end": 331.38,
      "text": "它的上限完全被底下调用的通用大模型给封死了",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1313201758512266,
      "no_speech_prob": 1.5682779727632656e-11,
      "compression_ratio": 1.0838509316770186,
      "source": "whisper"
    },
    {
      "start": 331.38,
      "end": 335.12,
      "text": "而Discovery Loop完全是另一个维度的存在",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1313201758512266,
      "no_speech_prob": 1.5682779727632656e-11,
      "compression_ratio": 1.0838509316770186,
      "source": "whisper"
    },
    {
      "start": 335.12,
      "end": 337.68,
      "text": "Jeff Dean和桑杰·格瓦马特",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1313201758512266,
      "no_speech_prob": 1.5682779727632656e-11,
      "compression_ratio": 1.0838509316770186,
      "source": "whisper"
    },
    {
      "start": 337.68,
      "end": 342.54,
      "text": "是写出MapReduce Bigtable的超级系统架构师",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1313201758512266,
      "no_speech_prob": 1.5682779727632656e-11,
      "compression_ratio": 1.0838509316770186,
      "source": "whisper"
    },
    {
      "start": 342.54,
      "end": 345.3,
      "text": "李光莲是AutoML的鼻祖",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.1313201758512266,
      "no_speech_prob": 1.5682779727632656e-11,
      "compression_ratio": 1.0838509316770186,
      "source": "whisper"
    },
    {
      "start": 345.3,
      "end": 347.02,
      "text": "奥里尔尔维尼亚尔斯",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.06588362622004683,
      "no_speech_prob": 1.3315545714653876e-11,
      "compression_ratio": 1.2972972972972974,
      "source": "whisper"
    },
    {
      "start": 347.02,
      "end": 350.58,
      "text": "是顶尖的神经网络搜索与多模态专家",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.06588362622004683,
      "no_speech_prob": 1.3315545714653876e-11,
      "compression_ratio": 1.2972972972972974,
      "source": "whisper"
    },
    {
      "start": 350.58,
      "end": 353.96,
      "text": "他们要做的是从底层分布式计算",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.06588362622004683,
      "no_speech_prob": 1.3315545714653876e-11,
      "compression_ratio": 1.2972972972972974,
      "source": "whisper"
    },
    {
      "start": 353.96,
      "end": 355.52,
      "text": "硬件算力调度",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.06588362622004683,
      "no_speech_prob": 1.3315545714653876e-11,
      "compression_ratio": 1.2972972972972974,
      "source": "whisper"
    },
    {
      "start": 355.52,
      "end": 360.12,
      "text": "神经架构搜索到算法闭环的系统级重构",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.06588362622004683,
      "no_speech_prob": 1.3315545714653876e-11,
      "compression_ratio": 1.2972972972972974,
      "source": "whisper"
    },
    {
      "start": 360.12,
      "end": 362.18,
      "text": "他们不是在用AI写论文",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.06588362622004683,
      "no_speech_prob": 1.3315545714653876e-11,
      "compression_ratio": 1.2972972972972974,
      "source": "whisper"
    },
    {
      "start": 362.18,
      "end": 364.66,
      "text": "他们是在用超级计算机集群",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.06588362622004683,
      "no_speech_prob": 1.3315545714653876e-11,
      "compression_ratio": 1.2972972972972974,
      "source": "whisper"
    },
    {
      "start": 364.66,
      "end": 369.26,
      "text": "让AI去做数以万计的硬件级与架构级实验",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.06588362622004683,
      "no_speech_prob": 1.3315545714653876e-11,
      "compression_ratio": 1.2972972972972974,
      "source": "whisper"
    },
    {
      "start": 369.26,
      "end": 373.0,
      "text": "直接在算法底层寻找下一个Transformer",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.06588362622004683,
      "no_speech_prob": 1.3315545714653876e-11,
      "compression_ratio": 1.2972972972972974,
      "source": "whisper"
    },
    {
      "start": 373.0,
      "end": 375.26,
      "text": "那么回到我们开头的悬念",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.06588362622004683,
      "no_speech_prob": 1.3315545714653876e-11,
      "compression_ratio": 1.2972972972972974,
      "source": "whisper"
    },
    {
      "start": 375.26,
      "end": 379.54,
      "text": "就是谷歌为什么不把这个超级团队留在内部",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.09870078549747578,
      "no_speech_prob": 1.2463616076707762e-11,
      "compression_ratio": 1.1772151898734178,
      "source": "whisper"
    },
    {
      "start": 379.54,
      "end": 383.34,
      "text": "而是支持他们独立成立一家公共利益公司呢",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.09870078549747578,
      "no_speech_prob": 1.2463616076707762e-11,
      "compression_ratio": 1.1772151898734178,
      "source": "whisper"
    },
    {
      "start": 383.34,
      "end": 386.26,
      "text": "因为谷歌算透了一笔账",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.09870078549747578,
      "no_speech_prob": 1.2463616076707762e-11,
      "compression_ratio": 1.1772151898734178,
      "source": "whisper"
    },
    {
      "start": 386.26,
      "end": 389.72,
      "text": "像AI自主研发下一代AI这种前沿探索",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.09870078549747578,
      "no_speech_prob": 1.2463616076707762e-11,
      "compression_ratio": 1.1772151898734178,
      "source": "whisper"
    },
    {
      "start": 389.72,
      "end": 391.1,
      "text": "风险极高",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.09870078549747578,
      "no_speech_prob": 1.2463616076707762e-11,
      "compression_ratio": 1.1772151898734178,
      "source": "whisper"
    },
    {
      "start": 391.1,
      "end": 392.34,
      "text": "烧钱极大",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.09870078549747578,
      "no_speech_prob": 1.2463616076707762e-11,
      "compression_ratio": 1.1772151898734178,
      "source": "whisper"
    },
    {
      "start": 392.34,
      "end": 393.64,
      "text": "失败率也极高",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.09870078549747578,
      "no_speech_prob": 1.2463616076707762e-11,
      "compression_ratio": 1.1772151898734178,
      "source": "whisper"
    },
    {
      "start": 393.64,
      "end": 395.72,
      "text": "如果放在谷歌Digemine内部",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.09870078549747578,
      "no_speech_prob": 1.2463616076707762e-11,
      "compression_ratio": 1.1772151898734178,
      "source": "whisper"
    },
    {
      "start": 395.72,
      "end": 400.42,
      "text": "研究员们每天要面对华尔街的财报审查和季度KPI",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.09870078549747578,
      "no_speech_prob": 1.2463616076707762e-11,
      "compression_ratio": 1.1772151898734178,
      "source": "whisper"
    },
    {
      "start": 400.42,
      "end": 403.7,
      "text": "就会被无限牵扯精力去搞产品适配",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.09870078549747578,
      "no_speech_prob": 1.2463616076707762e-11,
      "compression_ratio": 1.1772151898734178,
      "source": "whisper"
    },
    {
      "start": 403.7,
      "end": 406.26,
      "text": "通过把JFDN团队外部化",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07854664325714111,
      "no_speech_prob": 1.3266140790058056e-11,
      "compression_ratio": 1.1830985915492958,
      "source": "whisper"
    },
    {
      "start": 406.26,
      "end": 408.74,
      "text": "谷歌既让这一帮顶尖天才",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07854664325714111,
      "no_speech_prob": 1.3266140790058056e-11,
      "compression_ratio": 1.1830985915492958,
      "source": "whisper"
    },
    {
      "start": 408.74,
      "end": 412.28,
      "text": "摆脱了大公司病和商业汇报的枷锁",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07854664325714111,
      "no_speech_prob": 1.3266140790058056e-11,
      "compression_ratio": 1.1830985915492958,
      "source": "whisper"
    },
    {
      "start": 412.28,
      "end": 417.14,
      "text": "又能通过初始投资方和独家GoCloud算力的身份",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07854664325714111,
      "no_speech_prob": 1.3266140790058056e-11,
      "compression_ratio": 1.1830985915492958,
      "source": "whisper"
    },
    {
      "start": 417.14,
      "end": 421.84,
      "text": "牢牢锁定未来潜在的技术回流与云服务收益",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07854664325714111,
      "no_speech_prob": 1.3266140790058056e-11,
      "compression_ratio": 1.1830985915492958,
      "source": "whisper"
    },
    {
      "start": 421.84,
      "end": 426.24,
      "text": "这一招叫做前沿高风险战略的外部结偶",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07854664325714111,
      "no_speech_prob": 1.3266140790058056e-11,
      "compression_ratio": 1.1830985915492958,
      "source": "whisper"
    },
    {
      "start": 426.24,
      "end": 431.44,
      "text": "如果说JFDN的离职是把远期高风险探索拆分了出去",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.07854664325714111,
      "no_speech_prob": 1.3266140790058056e-11,
      "compression_ratio": 1.1830985915492958,
      "source": "whisper"
    },
    {
      "start": 431.44,
      "end": 433.32,
      "text": "那么接下来发生的事情",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15692661025307394,
      "no_speech_prob": 1.9602788142725203e-11,
      "compression_ratio": 1.128440366972477,
      "source": "whisper"
    },
    {
      "start": 433.32,
      "end": 436.74,
      "text": "才是对Jamlight产品生态影响最深远",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15692661025307394,
      "no_speech_prob": 1.9602788142725203e-11,
      "compression_ratio": 1.128440366972477,
      "source": "whisper"
    },
    {
      "start": 436.74,
      "end": 438.24,
      "text": "也是最暴力的一环",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15692661025307394,
      "no_speech_prob": 1.9602788142725203e-11,
      "compression_ratio": 1.128440366972477,
      "source": "whisper"
    },
    {
      "start": 438.24,
      "end": 441.14,
      "text": "接掌Google D.My的全局率颖的",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15692661025307394,
      "no_speech_prob": 1.9602788142725203e-11,
      "compression_ratio": 1.128440366972477,
      "source": "whisper"
    },
    {
      "start": 441.14,
      "end": 442.7,
      "text": "科雷·卡伍克·乔武",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15692661025307394,
      "no_speech_prob": 1.9602788142725203e-11,
      "compression_ratio": 1.128440366972477,
      "source": "whisper"
    },
    {
      "start": 442.7,
      "end": 444.56,
      "text": "是一个与Demis Hossabies",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15692661025307394,
      "no_speech_prob": 1.9602788142725203e-11,
      "compression_ratio": 1.128440366972477,
      "source": "whisper"
    },
    {
      "start": 444.56,
      "end": 447.16,
      "text": "风格截然不同的工程技术派",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15692661025307394,
      "no_speech_prob": 1.9602788142725203e-11,
      "compression_ratio": 1.128440366972477,
      "source": "whisper"
    },
    {
      "start": 447.16,
      "end": 451.48,
      "text": "Demis是国际象棋神童和神经科学家",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15692661025307394,
      "no_speech_prob": 1.9602788142725203e-11,
      "compression_ratio": 1.128440366972477,
      "source": "whisper"
    },
    {
      "start": 451.48,
      "end": 453.4,
      "text": "他充满着理想主义",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15692661025307394,
      "no_speech_prob": 1.9602788142725203e-11,
      "compression_ratio": 1.128440366972477,
      "source": "whisper"
    },
    {
      "start": 453.4,
      "end": 456.18,
      "text": "而科雷是把深度强化学习",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15692661025307394,
      "no_speech_prob": 1.9602788142725203e-11,
      "compression_ratio": 1.128440366972477,
      "source": "whisper"
    },
    {
      "start": 456.18,
      "end": 458.66,
      "text": "真正推向AlphaGo破局",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15692661025307394,
      "no_speech_prob": 1.9602788142725203e-11,
      "compression_ratio": 1.128440366972477,
      "source": "whisper"
    },
    {
      "start": 458.66,
      "end": 460.6,
      "text": "把WaveNet实施语音",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.15692661025307394,
      "no_speech_prob": 1.9602788142725203e-11,
      "compression_ratio": 1.128440366972477,
      "source": "whisper"
    },
    {
      "start": 460.6,
      "end": 465.62,
      "text": "真正变成极易手机里语音引擎的超级工程操盘手",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14291413128376007,
      "no_speech_prob": 1.1798275897922927e-11,
      "compression_ratio": 1.0661157024793388,
      "source": "whisper"
    },
    {
      "start": 465.62,
      "end": 467.76,
      "text": "在科雷掌舵之后",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14291413128376007,
      "no_speech_prob": 1.1798275897922927e-11,
      "compression_ratio": 1.0661157024793388,
      "source": "whisper"
    },
    {
      "start": 467.76,
      "end": 470.64,
      "text": "Google G-Mind释放出了一个极其明确的信号",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14291413128376007,
      "no_speech_prob": 1.1798275897922927e-11,
      "compression_ratio": 1.0661157024793388,
      "source": "whisper"
    },
    {
      "start": 470.64,
      "end": 473.06,
      "text": "就是全面放弃学术包袱",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14291413128376007,
      "no_speech_prob": 1.1798275897922927e-11,
      "compression_ratio": 1.0661157024793388,
      "source": "whisper"
    },
    {
      "start": 473.06,
      "end": 477.24,
      "text": "转向极致工程眼镜与大规模商业化暴击",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14291413128376007,
      "no_speech_prob": 1.1798275897922927e-11,
      "compression_ratio": 1.0661157024793388,
      "source": "whisper"
    },
    {
      "start": 477.24,
      "end": 478.7,
      "text": "在他的主持下",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14291413128376007,
      "no_speech_prob": 1.1798275897922927e-11,
      "compression_ratio": 1.0661157024793388,
      "source": "whisper"
    },
    {
      "start": 478.7,
      "end": 479.96,
      "text": "Gamlight的底层架构",
      "chunk": 0,
      "language": "zh",
      "avg_logprob": -0.14291413128376007,
      "no_speech_prob": 1.1798275897922927e-11,
      "compression_ratio": 1.0661157024793388,
      "source": "whisper"
    },
    {
      "start": 483.1,
      "end": 493.72,
      "text": "首先,最根本的變化是,徹底打破對預训訓參數崇拜的單位幻想,全面轉向後訓練與推理期計算的深度融合。",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.1606137752532959,
      "no_speech_prob": 8.276624177683267e-12,
      "compression_ratio": 1.173374613003096,
      "source": "whisper"
    },
    {
      "start": 493.72,
      "end": 497.8,
      "text": "在过去两年 行业盲目堆叠预训练参数",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09004640845613107,
      "no_speech_prob": 1.3113482522086883e-11,
      "compression_ratio": 1.1476510067114094,
      "source": "whisper"
    },
    {
      "start": 497.8,
      "end": 501.72,
      "text": "但预训练越往后 编辑效应递减越严重",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09004640845613107,
      "no_speech_prob": 1.3113482522086883e-11,
      "compression_ratio": 1.1476510067114094,
      "source": "whisper"
    },
    {
      "start": 501.72,
      "end": 504.1,
      "text": "算力成本高到令人发指",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09004640845613107,
      "no_speech_prob": 1.3113482522086883e-11,
      "compression_ratio": 1.1476510067114094,
      "source": "whisper"
    },
    {
      "start": 504.1,
      "end": 507.9,
      "text": "柯雷结合他在DQN强化学习领域的积累",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09004640845613107,
      "no_speech_prob": 1.3113482522086883e-11,
      "compression_ratio": 1.1476510067114094,
      "source": "whisper"
    },
    {
      "start": 507.9,
      "end": 511.28,
      "text": "给Gemla引入了过程奖励模型PRM",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09004640845613107,
      "no_speech_prob": 1.3113482522086883e-11,
      "compression_ratio": 1.1476510067114094,
      "source": "whisper"
    },
    {
      "start": 511.28,
      "end": 514.16,
      "text": "与蒙特卡洛书搜索MCTS",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09004640845613107,
      "no_speech_prob": 1.3113482522086883e-11,
      "compression_ratio": 1.1476510067114094,
      "source": "whisper"
    },
    {
      "start": 514.16,
      "end": 517.08,
      "text": "打个比方 传统的语言模型",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09004640845613107,
      "no_speech_prob": 1.3113482522086883e-11,
      "compression_ratio": 1.1476510067114094,
      "source": "whisper"
    },
    {
      "start": 517.08,
      "end": 519.82,
      "text": "就像在考场上没有任何草稿纸",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09004640845613107,
      "no_speech_prob": 1.3113482522086883e-11,
      "compression_ratio": 1.1476510067114094,
      "source": "whisper"
    },
    {
      "start": 519.82,
      "end": 521.78,
      "text": "靠直觉备答案的考生",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09004640845613107,
      "no_speech_prob": 1.3113482522086883e-11,
      "compression_ratio": 1.1476510067114094,
      "source": "whisper"
    },
    {
      "start": 521.78,
      "end": 524.18,
      "text": "遇到简单的题那是没问题的",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10655518668038505,
      "no_speech_prob": 1.1940416537459253e-11,
      "compression_ratio": 1.2013201320132014,
      "source": "whisper"
    },
    {
      "start": 524.18,
      "end": 527.52,
      "text": "但遇到复杂逻辑推导或者长流程编程",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10655518668038505,
      "no_speech_prob": 1.1940416537459253e-11,
      "compression_ratio": 1.2013201320132014,
      "source": "whisper"
    },
    {
      "start": 527.52,
      "end": 530.52,
      "text": "中间要是错一步后面就全乱了",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10655518668038505,
      "no_speech_prob": 1.1940416537459253e-11,
      "compression_ratio": 1.2013201320132014,
      "source": "whisper"
    },
    {
      "start": 530.52,
      "end": 533.46,
      "text": "而PRM与MCTS的融合",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10655518668038505,
      "no_speech_prob": 1.1940416537459253e-11,
      "compression_ratio": 1.2013201320132014,
      "source": "whisper"
    },
    {
      "start": 533.46,
      "end": 538.0,
      "text": "相当于给GEMLA配了一张无线扩展的虚拟草稿纸",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10655518668038505,
      "no_speech_prob": 1.1940416537459253e-11,
      "compression_ratio": 1.2013201320132014,
      "source": "whisper"
    },
    {
      "start": 538.0,
      "end": 539.92,
      "text": "和一位微观监考官",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10655518668038505,
      "no_speech_prob": 1.1940416537459253e-11,
      "compression_ratio": 1.2013201320132014,
      "source": "whisper"
    },
    {
      "start": 539.92,
      "end": 541.52,
      "text": "在推理的每一步",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10655518668038505,
      "no_speech_prob": 1.1940416537459253e-11,
      "compression_ratio": 1.2013201320132014,
      "source": "whisper"
    },
    {
      "start": 541.52,
      "end": 546.08,
      "text": "PRM都会对它的逻辑合理性进行毫秒级评分",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10655518668038505,
      "no_speech_prob": 1.1940416537459253e-11,
      "compression_ratio": 1.2013201320132014,
      "source": "whisper"
    },
    {
      "start": 546.08,
      "end": 550.0,
      "text": "同时MCTS数搜索会在后台预判",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10655518668038505,
      "no_speech_prob": 1.1940416537459253e-11,
      "compression_ratio": 1.2013201320132014,
      "source": "whisper"
    },
    {
      "start": 550.0,
      "end": 551.72,
      "text": "未来几十步的走向",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10655518668038505,
      "no_speech_prob": 1.1940416537459253e-11,
      "compression_ratio": 1.2013201320132014,
      "source": "whisper"
    },
    {
      "start": 551.72,
      "end": 554.72,
      "text": "如果发现某个分支走向死胡同",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.11916495686736554,
      "no_speech_prob": 1.2553218013688916e-11,
      "compression_ratio": 1.1719745222929936,
      "source": "whisper"
    },
    {
      "start": 554.72,
      "end": 559.14,
      "text": "模型会在推理阶段自动撤回并重新搜索最右键",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.11916495686736554,
      "no_speech_prob": 1.2553218013688916e-11,
      "compression_ratio": 1.1719745222929936,
      "source": "whisper"
    },
    {
      "start": 559.14,
      "end": 564.12,
      "text": "这意味着在不用无限膨胀预讯链参数的前提下",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.11916495686736554,
      "no_speech_prob": 1.2553218013688916e-11,
      "compression_ratio": 1.1719745222929936,
      "source": "whisper"
    },
    {
      "start": 564.12,
      "end": 566.24,
      "text": "GEMLINE处理高难度编程",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.11916495686736554,
      "no_speech_prob": 1.2553218013688916e-11,
      "compression_ratio": 1.1719745222929936,
      "source": "whisper"
    },
    {
      "start": 566.24,
      "end": 569.62,
      "text": "复杂数学推导和A乘的规划的成功率",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.11916495686736554,
      "no_speech_prob": 1.2553218013688916e-11,
      "compression_ratio": 1.1719745222929936,
      "source": "whisper"
    },
    {
      "start": 569.62,
      "end": 571.42,
      "text": "称指数级提升",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.11916495686736554,
      "no_speech_prob": 1.2553218013688916e-11,
      "compression_ratio": 1.1719745222929936,
      "source": "whisper"
    },
    {
      "start": 571.42,
      "end": 573.96,
      "text": "紧接着第二种转变在于",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.11916495686736554,
      "no_speech_prob": 1.2553218013688916e-11,
      "compression_ratio": 1.1719745222929936,
      "source": "whisper"
    },
    {
      "start": 573.96,
      "end": 576.56,
      "text": "系统级推理效率的极限拉链",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.11916495686736554,
      "no_speech_prob": 1.2553218013688916e-11,
      "compression_ratio": 1.1719745222929936,
      "source": "whisper"
    },
    {
      "start": 576.56,
      "end": 580.66,
      "text": "那就是构建GEMINI Spark端云协同分层架构",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.11916495686736554,
      "no_speech_prob": 1.2553218013688916e-11,
      "compression_ratio": 1.1719745222929936,
      "source": "whisper"
    },
    {
      "start": 580.66,
      "end": 584.26,
      "text": "这里稍微提一下我个人对Spark的使用体验",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10818703595329733,
      "no_speech_prob": 1.8144085192628445e-11,
      "compression_ratio": 1.2054380664652569,
      "source": "whisper"
    },
    {
      "start": 584.26,
      "end": 586.04,
      "text": "虽然它还是Beta版本",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10818703595329733,
      "no_speech_prob": 1.8144085192628445e-11,
      "compression_ratio": 1.2054380664652569,
      "source": "whisper"
    },
    {
      "start": 586.04,
      "end": 587.26,
      "text": "但先说结论",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10818703595329733,
      "no_speech_prob": 1.8144085192628445e-11,
      "compression_ratio": 1.2054380664652569,
      "source": "whisper"
    },
    {
      "start": 587.26,
      "end": 589.1,
      "text": "我觉得它超级好用",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10818703595329733,
      "no_speech_prob": 1.8144085192628445e-11,
      "compression_ratio": 1.2054380664652569,
      "source": "whisper"
    },
    {
      "start": 589.1,
      "end": 590.74,
      "text": "我会单独出一期视频",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10818703595329733,
      "no_speech_prob": 1.8144085192628445e-11,
      "compression_ratio": 1.2054380664652569,
      "source": "whisper"
    },
    {
      "start": 590.74,
      "end": 594.42,
      "text": "聊一下Gemla Spark和Hermis Agent的配合使用",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10818703595329733,
      "no_speech_prob": 1.8144085192628445e-11,
      "compression_ratio": 1.2054380664652569,
      "source": "whisper"
    },
    {
      "start": 594.42,
      "end": 596.78,
      "text": "作为Google订阅的长期体验者",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10818703595329733,
      "no_speech_prob": 1.8144085192628445e-11,
      "compression_ratio": 1.2054380664652569,
      "source": "whisper"
    },
    {
      "start": 596.78,
      "end": 600.54,
      "text": "我明显感觉最近Gemla的使用限额提高了",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10818703595329733,
      "no_speech_prob": 1.8144085192628445e-11,
      "compression_ratio": 1.2054380664652569,
      "source": "whisper"
    },
    {
      "start": 600.54,
      "end": 603.88,
      "text": "他们虽然没有发布像Fable 5这类的模型",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10818703595329733,
      "no_speech_prob": 1.8144085192628445e-11,
      "compression_ratio": 1.2054380664652569,
      "source": "whisper"
    },
    {
      "start": 603.88,
      "end": 606.4,
      "text": "但是无论是Gemla 3.6 Flash",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10818703595329733,
      "no_speech_prob": 1.8144085192628445e-11,
      "compression_ratio": 1.2054380664652569,
      "source": "whisper"
    },
    {
      "start": 606.4,
      "end": 609.44,
      "text": "还是Gemla 3.1 Pro都变得更聪明了",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10818703595329733,
      "no_speech_prob": 1.8144085192628445e-11,
      "compression_ratio": 1.2054380664652569,
      "source": "whisper"
    },
    {
      "start": 609.44,
      "end": 611.74,
      "text": "好 我们继续这个第二重转变",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10342338315902218,
      "no_speech_prob": 1.636373328506302e-11,
      "compression_ratio": 1.1561461794019934,
      "source": "whisper"
    },
    {
      "start": 611.74,
      "end": 614.26,
      "text": "在商业高并发场景中",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10342338315902218,
      "no_speech_prob": 1.636373328506302e-11,
      "compression_ratio": 1.1561461794019934,
      "source": "whisper"
    },
    {
      "start": 614.26,
      "end": 616.42,
      "text": "用户问一句今天天气怎么样",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10342338315902218,
      "no_speech_prob": 1.636373328506302e-11,
      "compression_ratio": 1.1561461794019934,
      "source": "whisper"
    },
    {
      "start": 616.42,
      "end": 620.66,
      "text": "如果都要调用一次千亿参数的单体大模型",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10342338315902218,
      "no_speech_prob": 1.636373328506302e-11,
      "compression_ratio": 1.1561461794019934,
      "source": "whisper"
    },
    {
      "start": 620.66,
      "end": 623.82,
      "text": "显存和电力开销能把大厂烧穿",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10342338315902218,
      "no_speech_prob": 1.636373328506302e-11,
      "compression_ratio": 1.1561461794019934,
      "source": "whisper"
    },
    {
      "start": 623.82,
      "end": 626.22,
      "text": "克雷推动的Gemlet Spark架构",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10342338315902218,
      "no_speech_prob": 1.636373328506302e-11,
      "compression_ratio": 1.1561461794019934,
      "source": "whisper"
    },
    {
      "start": 626.22,
      "end": 630.88,
      "text": "把整套AI系统做成了像特种部队一样的分层调度",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10342338315902218,
      "no_speech_prob": 1.636373328506302e-11,
      "compression_ratio": 1.1561461794019934,
      "source": "whisper"
    },
    {
      "start": 630.88,
      "end": 635.84,
      "text": "前端部署及轻量及低延迟的Gemlet Spark代理模型",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10342338315902218,
      "no_speech_prob": 1.636373328506302e-11,
      "compression_ratio": 1.1561461794019934,
      "source": "whisper"
    },
    {
      "start": 635.84,
      "end": 637.8,
      "text": "负责高频意图识别",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.11245080649134624,
      "no_speech_prob": 1.4191400658780662e-11,
      "compression_ratio": 1.1551724137931034,
      "source": "whisper"
    },
    {
      "start": 637.8,
      "end": 640.24,
      "text": "上下门过滤和轻量任务",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.11245080649134624,
      "no_speech_prob": 1.4191400658780662e-11,
      "compression_ratio": 1.1551724137931034,
      "source": "whisper"
    },
    {
      "start": 640.24,
      "end": 641.82,
      "text": "只有当前线判定",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.11245080649134624,
      "no_speech_prob": 1.4191400658780662e-11,
      "compression_ratio": 1.1551724137931034,
      "source": "whisper"
    },
    {
      "start": 641.82,
      "end": 644.18,
      "text": "涉及复杂长链条逻辑的时候",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.11245080649134624,
      "no_speech_prob": 1.4191400658780662e-11,
      "compression_ratio": 1.1551724137931034,
      "source": "whisper"
    },
    {
      "start": 644.18,
      "end": 647.56,
      "text": "才会无缝调动云端期间大模型",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.11245080649134624,
      "no_speech_prob": 1.4191400658780662e-11,
      "compression_ratio": 1.1551724137931034,
      "source": "whisper"
    },
    {
      "start": 647.56,
      "end": 648.96,
      "text": "这种分层架构",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.11245080649134624,
      "no_speech_prob": 1.4191400658780662e-11,
      "compression_ratio": 1.1551724137931034,
      "source": "whisper"
    },
    {
      "start": 648.96,
      "end": 651.86,
      "text": "直接把单位Token成本打了下来",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.11245080649134624,
      "no_speech_prob": 1.4191400658780662e-11,
      "compression_ratio": 1.1551724137931034,
      "source": "whisper"
    },
    {
      "start": 651.86,
      "end": 655.34,
      "text": "这才是支撑数亿并发的商业硬实力",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.11245080649134624,
      "no_speech_prob": 1.4191400658780662e-11,
      "compression_ratio": 1.1551724137931034,
      "source": "whisper"
    },
    {
      "start": 655.34,
      "end": 656.9,
      "text": "而第三种变化",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.11245080649134624,
      "no_speech_prob": 1.4191400658780662e-11,
      "compression_ratio": 1.1551724137931034,
      "source": "whisper"
    },
    {
      "start": 656.9,
      "end": 659.46,
      "text": "则是科雷将他的看家本领",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.11245080649134624,
      "no_speech_prob": 1.4191400658780662e-11,
      "compression_ratio": 1.1551724137931034,
      "source": "whisper"
    },
    {
      "start": 659.46,
      "end": 662.5,
      "text": "Vivnet原生音频与流逝生成",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.11245080649134624,
      "no_speech_prob": 1.4191400658780662e-11,
      "compression_ratio": 1.1551724137931034,
      "source": "whisper"
    },
    {
      "start": 662.5,
      "end": 666.24,
      "text": "深度融合到了詹姆賴的全模态底座中",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10335779702791603,
      "no_speech_prob": 1.4790635730066448e-11,
      "compression_ratio": 1.28719723183391,
      "source": "whisper"
    },
    {
      "start": 666.24,
      "end": 668.02,
      "text": "以往的多模态AI",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10335779702791603,
      "no_speech_prob": 1.4790635730066448e-11,
      "compression_ratio": 1.28719723183391,
      "source": "whisper"
    },
    {
      "start": 668.02,
      "end": 671.42,
      "text": "往往是文本模型外挂语音识别",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10335779702791603,
      "no_speech_prob": 1.4790635730066448e-11,
      "compression_ratio": 1.28719723183391,
      "source": "whisper"
    },
    {
      "start": 671.42,
      "end": 673.96,
      "text": "与语音合成模块的拼接车",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10335779702791603,
      "no_speech_prob": 1.4790635730066448e-11,
      "compression_ratio": 1.28719723183391,
      "source": "whisper"
    },
    {
      "start": 673.96,
      "end": 676.12,
      "text": "延迟高 语调生硬",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10335779702791603,
      "no_speech_prob": 1.4790635730066448e-11,
      "compression_ratio": 1.28719723183391,
      "source": "whisper"
    },
    {
      "start": 676.12,
      "end": 678.12,
      "text": "而柯雷主导下的詹姆賴",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10335779702791603,
      "no_speech_prob": 1.4790635730066448e-11,
      "compression_ratio": 1.28719723183391,
      "source": "whisper"
    },
    {
      "start": 678.12,
      "end": 681.44,
      "text": "实现了真正的原生全模态流逝交互",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10335779702791603,
      "no_speech_prob": 1.4790635730066448e-11,
      "compression_ratio": 1.28719723183391,
      "source": "whisper"
    },
    {
      "start": 681.44,
      "end": 683.28,
      "text": "不仅能实现毫秒级",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10335779702791603,
      "no_speech_prob": 1.4790635730066448e-11,
      "compression_ratio": 1.28719723183391,
      "source": "whisper"
    },
    {
      "start": 683.28,
      "end": 685.84,
      "text": "实时音视频双向对打",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10335779702791603,
      "no_speech_prob": 1.4790635730066448e-11,
      "compression_ratio": 1.28719723183391,
      "source": "whisper"
    },
    {
      "start": 685.84,
      "end": 687.64,
      "text": "还能在听懂说话的同时",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10335779702791603,
      "no_speech_prob": 1.4790635730066448e-11,
      "compression_ratio": 1.28719723183391,
      "source": "whisper"
    },
    {
      "start": 687.64,
      "end": 689.18,
      "text": "自主执行代码",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10335779702791603,
      "no_speech_prob": 1.4790635730066448e-11,
      "compression_ratio": 1.28719723183391,
      "source": "whisper"
    },
    {
      "start": 689.18,
      "end": 691.58,
      "text": "甚至在手机屏幕上动态生成",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10335779702791603,
      "no_speech_prob": 1.4790635730066448e-11,
      "compression_ratio": 1.28719723183391,
      "source": "whisper"
    },
    {
      "start": 691.58,
      "end": 693.6,
      "text": "UI界面与交互空间",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09496749787640994,
      "no_speech_prob": 1.5565684158280746e-11,
      "compression_ratio": 1.2108626198083068,
      "source": "whisper"
    },
    {
      "start": 693.6,
      "end": 695.6,
      "text": "这已经不是在聊天了",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09496749787640994,
      "no_speech_prob": 1.5565684158280746e-11,
      "compression_ratio": 1.2108626198083068,
      "source": "whisper"
    },
    {
      "start": 695.6,
      "end": 700.58,
      "text": "这是直接把自然语言变成了实时软件生成的界面",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09496749787640994,
      "no_speech_prob": 1.5565684158280746e-11,
      "compression_ratio": 1.2108626198083068,
      "source": "whisper"
    },
    {
      "start": 700.58,
      "end": 702.8,
      "text": "那底层架构演进的再漂亮",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09496749787640994,
      "no_speech_prob": 1.5565684158280746e-11,
      "compression_ratio": 1.2108626198083068,
      "source": "whisper"
    },
    {
      "start": 702.8,
      "end": 705.0,
      "text": "如果不能转化为商业流水",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09496749787640994,
      "no_speech_prob": 1.5565684158280746e-11,
      "compression_ratio": 1.2108626198083068,
      "source": "whisper"
    },
    {
      "start": 705.0,
      "end": 707.28,
      "text": "在今天也是一张废纸",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09496749787640994,
      "no_speech_prob": 1.5565684158280746e-11,
      "compression_ratio": 1.2108626198083068,
      "source": "whisper"
    },
    {
      "start": 707.28,
      "end": 710.9,
      "text": "科雷·卡伍克·乔鲁这次最让人惊悚的调整",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09496749787640994,
      "no_speech_prob": 1.5565684158280746e-11,
      "compression_ratio": 1.2108626198083068,
      "source": "whisper"
    },
    {
      "start": 710.9,
      "end": 712.68,
      "text": "在于他不仅管科研",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09496749787640994,
      "no_speech_prob": 1.5565684158280746e-11,
      "compression_ratio": 1.2108626198083068,
      "source": "whisper"
    },
    {
      "start": 712.68,
      "end": 717.6,
      "text": "还同时兼任了Gemline产品与开发者生态的最高指挥官",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09496749787640994,
      "no_speech_prob": 1.5565684158280746e-11,
      "compression_ratio": 1.2108626198083068,
      "source": "whisper"
    },
    {
      "start": 717.6,
      "end": 719.94,
      "text": "在过去的大厂体制里",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09496749787640994,
      "no_speech_prob": 1.5565684158280746e-11,
      "compression_ratio": 1.2108626198083068,
      "source": "whisper"
    },
    {
      "start": 719.94,
      "end": 724.36,
      "text": "科研实验室和商品产品部门是完全脱节的",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09205443208867853,
      "no_speech_prob": 1.3162270752486993e-11,
      "compression_ratio": 1.1729559748427674,
      "source": "whisper"
    },
    {
      "start": 724.36,
      "end": 725.58,
      "text": "实验室发了论文",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09205443208867853,
      "no_speech_prob": 1.3162270752486993e-11,
      "compression_ratio": 1.1729559748427674,
      "source": "whisper"
    },
    {
      "start": 725.58,
      "end": 728.78,
      "text": "产品部门要花大半年去沟通打包",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09205443208867853,
      "no_speech_prob": 1.3162270752486993e-11,
      "compression_ratio": 1.1729559748427674,
      "source": "whisper"
    },
    {
      "start": 728.78,
      "end": 730.74,
      "text": "而柯雷一把抓之后",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09205443208867853,
      "no_speech_prob": 1.3162270752486993e-11,
      "compression_ratio": 1.1729559748427674,
      "source": "whisper"
    },
    {
      "start": 730.74,
      "end": 733.82,
      "text": "直接祭出了商业落地的三大杀招",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09205443208867853,
      "no_speech_prob": 1.3162270752486993e-11,
      "compression_ratio": 1.1729559748427674,
      "source": "whisper"
    },
    {
      "start": 733.82,
      "end": 735.82,
      "text": "在终端入口上",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09205443208867853,
      "no_speech_prob": 1.3162270752486993e-11,
      "compression_ratio": 1.1729559748427674,
      "source": "whisper"
    },
    {
      "start": 735.82,
      "end": 739.0,
      "text": "他们开始加速清洗传统的语音助手",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09205443208867853,
      "no_speech_prob": 1.3162270752486993e-11,
      "compression_ratio": 1.1729559748427674,
      "source": "whisper"
    },
    {
      "start": 739.0,
      "end": 742.44,
      "text": "全面用GM来替换掉Android和VLOS上",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09205443208867853,
      "no_speech_prob": 1.3162270752486993e-11,
      "compression_ratio": 1.1729559748427674,
      "source": "whisper"
    },
    {
      "start": 742.44,
      "end": 744.56,
      "text": "那个指挥社闹钟的assistant",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09205443208867853,
      "no_speech_prob": 1.3162270752486993e-11,
      "compression_ratio": 1.1729559748427674,
      "source": "whisper"
    },
    {
      "start": 744.56,
      "end": 748.86,
      "text": "赋予他操作系统级的最高上下文读取权限",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09205443208867853,
      "no_speech_prob": 1.3162270752486993e-11,
      "compression_ratio": 1.1729559748427674,
      "source": "whisper"
    },
    {
      "start": 748.86,
      "end": 751.16,
      "text": "与跨应用API调用能力",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09605925067577487,
      "no_speech_prob": 1.148209825829749e-11,
      "compression_ratio": 1.1533101045296168,
      "source": "whisper"
    },
    {
      "start": 751.16,
      "end": 755.04,
      "text": "直接把它打造成操作系统的中央控制大脑",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09605925067577487,
      "no_speech_prob": 1.148209825829749e-11,
      "compression_ratio": 1.1533101045296168,
      "source": "whisper"
    },
    {
      "start": 755.04,
      "end": 758.52,
      "text": "比如在ValOS手表或者车载系统里",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09605925067577487,
      "no_speech_prob": 1.148209825829749e-11,
      "compression_ratio": 1.1533101045296168,
      "source": "whisper"
    },
    {
      "start": 758.52,
      "end": 762.34,
      "text": "你只需要说帮我安排今晚与张总的商务晚餐",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09605925067577487,
      "no_speech_prob": 1.148209825829749e-11,
      "compression_ratio": 1.1533101045296168,
      "source": "whisper"
    },
    {
      "start": 762.34,
      "end": 766.04,
      "text": "并准备一份他公司最新财报的背景简报",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09605925067577487,
      "no_speech_prob": 1.148209825829749e-11,
      "compression_ratio": 1.1533101045296168,
      "source": "whisper"
    },
    {
      "start": 766.04,
      "end": 770.16,
      "text": "Gamlight就会自动调取邮件和地图订餐厅",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09605925067577487,
      "no_speech_prob": 1.148209825829749e-11,
      "compression_ratio": 1.1533101045296168,
      "source": "whisper"
    },
    {
      "start": 770.16,
      "end": 774.44,
      "text": "同时在云端提取财报生成简报推送给你",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09605925067577487,
      "no_speech_prob": 1.148209825829749e-11,
      "compression_ratio": 1.1533101045296168,
      "source": "whisper"
    },
    {
      "start": 774.44,
      "end": 775.58,
      "text": "与此同时",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10293127542518708,
      "no_speech_prob": 1.1691482851294843e-11,
      "compression_ratio": 1.186206896551724,
      "source": "whisper"
    },
    {
      "start": 775.58,
      "end": 778.3,
      "text": "他们还大幅压缩了科研成果",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10293127542518708,
      "no_speech_prob": 1.1691482851294843e-11,
      "compression_ratio": 1.186206896551724,
      "source": "whisper"
    },
    {
      "start": 778.3,
      "end": 780.44,
      "text": "转化到商业API的时间",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10293127542518708,
      "no_speech_prob": 1.1691482851294843e-11,
      "compression_ratio": 1.186206896551724,
      "source": "whisper"
    },
    {
      "start": 780.44,
      "end": 784.44,
      "text": "在谷歌云Vertex AI和开发者API上",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10293127542518708,
      "no_speech_prob": 1.1691482851294843e-11,
      "compression_ratio": 1.186206896551724,
      "source": "whisper"
    },
    {
      "start": 784.44,
      "end": 787.14,
      "text": "前沿团队在Achic代码生成",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10293127542518708,
      "no_speech_prob": 1.1691482851294843e-11,
      "compression_ratio": 1.186206896551724,
      "source": "whisper"
    },
    {
      "start": 787.14,
      "end": 790.8,
      "text": "自动模型评估和推理加速上的突破",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10293127542518708,
      "no_speech_prob": 1.1691482851294843e-11,
      "compression_ratio": 1.186206896551724,
      "source": "whisper"
    },
    {
      "start": 790.8,
      "end": 793.56,
      "text": "能以周级别的速度直接部署上线",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10293127542518708,
      "no_speech_prob": 1.1691482851294843e-11,
      "compression_ratio": 1.186206896551724,
      "source": "whisper"
    },
    {
      "start": 793.56,
      "end": 795.72,
      "text": "让科研成果在第一秒",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10293127542518708,
      "no_speech_prob": 1.1691482851294843e-11,
      "compression_ratio": 1.186206896551724,
      "source": "whisper"
    },
    {
      "start": 795.72,
      "end": 797.68,
      "text": "直接转化为商业账单",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10293127542518708,
      "no_speech_prob": 1.1691482851294843e-11,
      "compression_ratio": 1.186206896551724,
      "source": "whisper"
    },
    {
      "start": 797.68,
      "end": 801.34,
      "text": "形成技术产品与算力的超级费轮",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10293127542518708,
      "no_speech_prob": 1.1691482851294843e-11,
      "compression_ratio": 1.186206896551724,
      "source": "whisper"
    },
    {
      "start": 801.34,
      "end": 802.74,
      "text": "更关键的是",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.10293127542518708,
      "no_speech_prob": 1.1691482851294843e-11,
      "compression_ratio": 1.186206896551724,
      "source": "whisper"
    },
    {
      "start": 802.74,
      "end": 805.96,
      "text": "谷歌正在悄无声息地改变自己的盈利模式",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09177576262375405,
      "no_speech_prob": 1.2611357270986279e-11,
      "compression_ratio": 1.212624584717608,
      "source": "whisper"
    },
    {
      "start": 805.96,
      "end": 810.0,
      "text": "当用户不再翻页点击搜索广告链接",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09177576262375405,
      "no_speech_prob": 1.2611357270986279e-11,
      "compression_ratio": 1.212624584717608,
      "source": "whisper"
    },
    {
      "start": 810.0,
      "end": 812.72,
      "text": "而是让Gamlet给出最终答案时",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09177576262375405,
      "no_speech_prob": 1.2611357270986279e-11,
      "compression_ratio": 1.212624584717608,
      "source": "whisper"
    },
    {
      "start": 812.72,
      "end": 814.02,
      "text": "谷歌的解法是",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09177576262375405,
      "no_speech_prob": 1.2611357270986279e-11,
      "compression_ratio": 1.212624584717608,
      "source": "whisper"
    },
    {
      "start": 814.02,
      "end": 817.3,
      "text": "直接占据agent决策路径的入口",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09177576262375405,
      "no_speech_prob": 1.2611357270986279e-11,
      "compression_ratio": 1.212624584717608,
      "source": "whisper"
    },
    {
      "start": 817.3,
      "end": 820.64,
      "text": "当Gamlet在每日简报或者个人助理里",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09177576262375405,
      "no_speech_prob": 1.2611357270986279e-11,
      "compression_ratio": 1.212624584717608,
      "source": "whisper"
    },
    {
      "start": 820.64,
      "end": 823.26,
      "text": "帮你规划行程推荐服务时",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09177576262375405,
      "no_speech_prob": 1.2611357270986279e-11,
      "compression_ratio": 1.212624584717608,
      "source": "whisper"
    },
    {
      "start": 823.26,
      "end": 825.44,
      "text": "它在生成方案的同时",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09177576262375405,
      "no_speech_prob": 1.2611357270986279e-11,
      "compression_ratio": 1.212624584717608,
      "source": "whisper"
    },
    {
      "start": 825.44,
      "end": 827.96,
      "text": "精准接入商业API入口",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09177576262375405,
      "no_speech_prob": 1.2611357270986279e-11,
      "compression_ratio": 1.212624584717608,
      "source": "whisper"
    },
    {
      "start": 827.96,
      "end": 831.5,
      "text": "谷歌由此从一个信息检索广告商",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09177576262375405,
      "no_speech_prob": 1.2611357270986279e-11,
      "compression_ratio": 1.212624584717608,
      "source": "whisper"
    },
    {
      "start": 831.5,
      "end": 834.64,
      "text": "变成了全局智能决策的开门人",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.08402070691508631,
      "no_speech_prob": 1.3626461070614937e-11,
      "compression_ratio": 1.1428571428571428,
      "source": "whisper"
    },
    {
      "start": 834.64,
      "end": 836.92,
      "text": "抽取服务分账和佣金",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.08402070691508631,
      "no_speech_prob": 1.3626461070614937e-11,
      "compression_ratio": 1.1428571428571428,
      "source": "whisper"
    },
    {
      "start": 836.92,
      "end": 839.58,
      "text": "拆决员Jeff Dean的离职创业",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.08402070691508631,
      "no_speech_prob": 1.3626461070614937e-11,
      "compression_ratio": 1.1428571428571428,
      "source": "whisper"
    },
    {
      "start": 839.58,
      "end": 842.14,
      "text": "与柯磊掌舵下的Deep Mind重组",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.08402070691508631,
      "no_speech_prob": 1.3626461070614937e-11,
      "compression_ratio": 1.1428571428571428,
      "source": "whisper"
    },
    {
      "start": 842.14,
      "end": 846.44,
      "text": "我们终于能看清这场硅谷AI大戏的全貌了",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.08402070691508631,
      "no_speech_prob": 1.3626461070614937e-11,
      "compression_ratio": 1.1428571428571428,
      "source": "whisper"
    },
    {
      "start": 846.44,
      "end": 851.14,
      "text": "硅谷的AI战争正是告别了依靠概念和论文引用量",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.08402070691508631,
      "no_speech_prob": 1.3626461070614937e-11,
      "compression_ratio": 1.1428571428571428,
      "source": "whisper"
    },
    {
      "start": 851.14,
      "end": 853.78,
      "text": "获取高估值的学术竞赛阶段",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.08402070691508631,
      "no_speech_prob": 1.3626461070614937e-11,
      "compression_ratio": 1.1428571428571428,
      "source": "whisper"
    },
    {
      "start": 853.78,
      "end": 856.7,
      "text": "全面跨入了拼推理成本",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.08402070691508631,
      "no_speech_prob": 1.3626461070614937e-11,
      "compression_ratio": 1.1428571428571428,
      "source": "whisper"
    },
    {
      "start": 856.7,
      "end": 858.38,
      "text": "拼系统稳定性",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.08402070691508631,
      "no_speech_prob": 1.3626461070614937e-11,
      "compression_ratio": 1.1428571428571428,
      "source": "whisper"
    },
    {
      "start": 858.38,
      "end": 859.88,
      "text": "拼终端控制力",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.08402070691508631,
      "no_speech_prob": 1.3626461070614937e-11,
      "compression_ratio": 1.1428571428571428,
      "source": "whisper"
    },
    {
      "start": 859.88,
      "end": 863.92,
      "text": "以及拼商业造学能力的工程与商业深水去",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.0772941916028439,
      "no_speech_prob": 1.2027743384601663e-11,
      "compression_ratio": 1.2541254125412542,
      "source": "whisper"
    },
    {
      "start": 863.92,
      "end": 867.8,
      "text": "对于我们所有在做AI研发架构设计",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.0772941916028439,
      "no_speech_prob": 1.2027743384601663e-11,
      "compression_ratio": 1.2541254125412542,
      "source": "whisper"
    },
    {
      "start": 867.8,
      "end": 869.92,
      "text": "或者技术创业的朋友来说",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.0772941916028439,
      "no_speech_prob": 1.2027743384601663e-11,
      "compression_ratio": 1.2541254125412542,
      "source": "whisper"
    },
    {
      "start": 869.92,
      "end": 871.72,
      "text": "谷歌的这一系列辩证",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.0772941916028439,
      "no_speech_prob": 1.2027743384601663e-11,
      "compression_ratio": 1.2541254125412542,
      "source": "whisper"
    },
    {
      "start": 871.72,
      "end": 874.86,
      "text": "留下了三条极其宝贵的工程启发",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.0772941916028439,
      "no_speech_prob": 1.2027743384601663e-11,
      "compression_ratio": 1.2541254125412542,
      "source": "whisper"
    },
    {
      "start": 874.86,
      "end": 876.94,
      "text": "首先不要盲目迷信",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.0772941916028439,
      "no_speech_prob": 1.2027743384601663e-11,
      "compression_ratio": 1.2541254125412542,
      "source": "whisper"
    },
    {
      "start": 876.94,
      "end": 879.58,
      "text": "纯域训练参数的单位堆叠",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.0772941916028439,
      "no_speech_prob": 1.2027743384601663e-11,
      "compression_ratio": 1.2541254125412542,
      "source": "whisper"
    },
    {
      "start": 879.58,
      "end": 881.96,
      "text": "多去研究推理器计算",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.0772941916028439,
      "no_speech_prob": 1.2027743384601663e-11,
      "compression_ratio": 1.2541254125412542,
      "source": "whisper"
    },
    {
      "start": 881.96,
      "end": 886.12,
      "text": "通过过程奖励模型打分与蒙特卡洛数搜索",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.0772941916028439,
      "no_speech_prob": 1.2027743384601663e-11,
      "compression_ratio": 1.2541254125412542,
      "source": "whisper"
    },
    {
      "start": 886.12,
      "end": 889.3,
      "text": "让模型在推理阶段学会自我纠错",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.0772941916028439,
      "no_speech_prob": 1.2027743384601663e-11,
      "compression_ratio": 1.2541254125412542,
      "source": "whisper"
    },
    {
      "start": 889.3,
      "end": 893.12,
      "text": "这是低成本实现逻辑跨越的唯一解法",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09766130499501045,
      "no_speech_prob": 1.1086297678486456e-11,
      "compression_ratio": 1.1981424148606812,
      "source": "whisper"
    },
    {
      "start": 893.12,
      "end": 896.58,
      "text": "其次要学会拥抱端云分层架构",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09766130499501045,
      "no_speech_prob": 1.1086297678486456e-11,
      "compression_ratio": 1.1981424148606812,
      "source": "whisper"
    },
    {
      "start": 896.58,
      "end": 898.52,
      "text": "平气无论任务大小",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09766130499501045,
      "no_speech_prob": 1.1086297678486456e-11,
      "compression_ratio": 1.1981424148606812,
      "source": "whisper"
    },
    {
      "start": 898.52,
      "end": 901.62,
      "text": "一力调用千亿大模型的粗暴思维",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09766130499501045,
      "no_speech_prob": 1.1086297678486456e-11,
      "compression_ratio": 1.1981424148606812,
      "source": "whisper"
    },
    {
      "start": 901.62,
      "end": 903.4,
      "text": "学习Jamler Spark的设计",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09766130499501045,
      "no_speech_prob": 1.1086297678486456e-11,
      "compression_ratio": 1.1981424148606812,
      "source": "whisper"
    },
    {
      "start": 903.4,
      "end": 906.9,
      "text": "用轻量级模型做好前置意图路由",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09766130499501045,
      "no_speech_prob": 1.1086297678486456e-11,
      "compression_ratio": 1.1981424148606812,
      "source": "whisper"
    },
    {
      "start": 906.9,
      "end": 910.04,
      "text": "才能在高并发任务中守住成本底线",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09766130499501045,
      "no_speech_prob": 1.1086297678486456e-11,
      "compression_ratio": 1.1981424148606812,
      "source": "whisper"
    },
    {
      "start": 910.04,
      "end": 913.8,
      "text": "最后我们需要重新思考软件入口的定义",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09766130499501045,
      "no_speech_prob": 1.1086297678486456e-11,
      "compression_ratio": 1.1981424148606812,
      "source": "whisper"
    },
    {
      "start": 913.8,
      "end": 916.26,
      "text": "未来的软件竞争不再是争夺",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09766130499501045,
      "no_speech_prob": 1.1086297678486456e-11,
      "compression_ratio": 1.1981424148606812,
      "source": "whisper"
    },
    {
      "start": 916.26,
      "end": 918.42,
      "text": "手机屏幕上的APP图标位置",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09766130499501045,
      "no_speech_prob": 1.1086297678486456e-11,
      "compression_ratio": 1.1981424148606812,
      "source": "whisper"
    },
    {
      "start": 918.42,
      "end": 920.98,
      "text": "而是争夺能否被系统及agent",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09820140491832387,
      "no_speech_prob": 9.515414498006969e-12,
      "compression_ratio": 0.9366197183098591,
      "source": "whisper"
    },
    {
      "start": 920.98,
      "end": 923.16,
      "text": "API决策链所调用",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09820140491832387,
      "no_speech_prob": 9.515414498006969e-12,
      "compression_ratio": 0.9366197183098591,
      "source": "whisper"
    },
    {
      "start": 923.16,
      "end": 925.12,
      "text": "技术不相信幻觉",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09820140491832387,
      "no_speech_prob": 9.515414498006969e-12,
      "compression_ratio": 0.9366197183098591,
      "source": "whisper"
    },
    {
      "start": 925.12,
      "end": 926.72,
      "text": "商业不相信眼泪",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09820140491832387,
      "no_speech_prob": 9.515414498006969e-12,
      "compression_ratio": 0.9366197183098591,
      "source": "whisper"
    },
    {
      "start": 926.72,
      "end": 927.86,
      "text": "这里是All Insight",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09820140491832387,
      "no_speech_prob": 9.515414498006969e-12,
      "compression_ratio": 0.9366197183098591,
      "source": "whisper"
    },
    {
      "start": 927.86,
      "end": 928.76,
      "text": "我们下期见",
      "chunk": 1,
      "language": "zh",
      "avg_logprob": -0.09820140491832387,
      "no_speech_prob": 9.515414498006969e-12,
      "compression_ratio": 0.9366197183098591,
      "source": "whisper"
    }
  ],
  "transcript_selection": {
    "schema": "videonote.transcript_selection.v1",
    "status": "SELECTED",
    "requested_language": "zh",
    "duration": 929.100063,
    "reference_path": null,
    "reference_kind": null,
    "reference_language": null,
    "selected_source": "whisper_segmented",
    "decision_reason": "source_unavailable_or_untrusted_whisper_passed",
    "whisper_rejected": false,
    "source_profile": {
      "segment_count": 0,
      "duration_coverage": 0.0,
      "usable_full": false,
      "usable_partial": false,
      "hallucination_region_count": 0,
      "hallucination_regions": [],
      "language": {
        "requested": "zh",
        "matched": false
      }
    },
    "whisper_profile": {
      "segment_count": 330,
      "first_start": 0.0,
      "final_end": 928.76,
      "duration_coverage": 0.999634,
      "maximum_gap_seconds": 3.14,
      "inverted_count": 0,
      "language": {
        "requested": "zh",
        "matched": true,
        "latin": 788,
        "han": 3659,
        "kana": 0,
        "hangul": 0
      },
      "hallucination_region_count": 0,
      "hallucination_regions": [],
      "usable_full": true,
      "usable_partial": true
    },
    "comparison": {
      "window_seconds": 30.0,
      "compared_window_count": 0,
      "median_score": 0.0,
      "divergent_window_count": 0,
      "divergent_ratio": 0.0,
      "divergent_windows": []
    }
  }
}