{ "text": "大家好 欢迎来到林姐说AI\n这期视频跟大家分享一下\n如何用AI原生的方法\n将Cloud Code接入第三方的模型\n比如说国内的低成本的模型\n像Deepseek的V4\n大家看到这是我的Cloud的桌面端的GUI界面\n这个是我和他的聊天\n这里有个很神奇的地方\n看一下这里的模型选择\n则是Deep Seek的V4 Pro Flash各个版本的模型接入\n有没有一种穿越的感觉\n你看而且Deep Seek也可以快速的给我回复\n并且给我高质量的答复\n最近我主力的运营的运行的中心是Codex\n它是我执行非常多落地任务的手脚\n我是使用GPT加Codex的方式来完成整个的路\n但是呢\nCodex也有一些坏处\n我的这个模型的套餐\n已经开到100刀Pro Lite的这个档位了\n有可能马上又要升到200刀\n但是因为网络不是特别稳定\n它经常针对一些常任务会失败\n就是一直会出现这个reconnect这个界面啊\n任务经常会失败\n特别是网络不太稳定的时候\n这个时候呢\n我的这个用完Codex的时候呢\n我就需要再去用一下Cloud Code去挥挥雪\n用Cloud Code去接入我的这些比较便宜一些模型\n这样子我就有一个左右护法了\n一个Codex一个Cloud Code加第三方模型\n再加上GPT\n让我整个的运行项目的效率变得更高\n有一个平衡点吧\n而且用起来真的挺爽的\n这一期我会给大家分享这个方法\n而且我今天分享的方法是非常AI原生化的\n有的人曾经在我的节目里面留言说\n那你都用国内的第三方的一些模型了\n那你还为什么要接Cloud Code呢\n这不是多此一举吗\n我告诉你非常有必要\n首先啊\n你用过Cloud Code\n你就知道Cloud Code\n它的Harness整个的工作流的框架是非常强的\n它之前不是还把自己整个托管的体系拿出来卖吗\n这个其实就是它的底气\n它里面是有一个很好的规划过的agent loop的\n它可以去自己去规划任务\n要用工具读取结果继续执行\nCloud Code它整个的文件系统的能力\n包括终端的能力都是非常强的\n它现在整体的估值这么高\n每年的这个AR这么高都是有道理的\n它是真的强\n而且在Cloud Code这个桌面端里面的应用\n普通的知识工作者或者普通的职场人士\n也能够很好的去使用\nCodex它的桌面段引起了一个新的风潮\n就是其实在未来\n这个CLI的这个模式是更加先进的\n而且它是精确的\n我也非常提倡大家去使用CLI的模式\n但是对于大部分的普通的工作者而言\n这个GUI的模式还是大家更适应的模式\n而且这个Codex现在真的是非常流行\n大家对这样的工作方式也是比较适应的\n所以在Cloud Code的桌面端\n也去接入第三方的模型\n它中间有一个能力\n就是Cowork\nCowork区别于Cloud Code\n就是写代码的能力\n它相当于是给大家普通的这种工作者\n也可以利用这个模型框架\n去解决一些具体的自主工作的问题\n并且能够返回完成品\nCloud Code本身会更加偏开发者终端和一些代码项目和工程项目\n那Cloud Code Cowork会这种集群还有封群的模式\n会更加适合普通的办公任务本地文件\n还有应用操作和交付成果\n我自己在使用的时候\n我觉得它的整个的Harness框架加上Deep Seek V4\n或者类似这样的能力的模型\n它的整个的产出结果还是不错的\n而且它可以和GPT去协同\n它最后给出的结果\nGPT也认可了\n可见它在这样的框架下的交付能力生成结果\n是比较好的\n质量是相对高的\n而且我自己的使用体验下来呢\n我会感觉啊\n现在Cloud Code和Codex的\n它的整个的Harness框架是非常强的\n他们两家最近的动作也是最多的\n我有一个非常超前的想象\n我觉得可以用Cloud Code和Harness这两家的Harness框架\n可以把它抽取出来去调用任何一家的大模型能力\n我们可以把它处理成这样\n有了这样比较好的Harness框架\n我们可以在顶层去调动不同厂商的大模型\n而且我们可以在顶层统一的去管理各个agent的Skier MCP\n哇我觉得真的是非常多的\n这个可以玩的东西\n好说回来啊\n我再跟大家切实的去分享\n怎么去做这样的接入\n把Cloud Code的GUI桌面界面\n去接入我们的Deep Seek V4\n我是以Deep Seek为例\n其实各家的这个模型\n都是可以接入的\n我这里讲的AI原生的方法\n就是用Codex\n给它具体的接入文档\n让Codex成为手脚\n来帮你使用CC-Switch这样的工具\n把Cloud Code的 Desktop的桌面端的界面\n接入像DeepSeek、阿里云或者CodingPlan\n或者其他的一些成本比较低的国内的模型\n这里我是给了他一个文档的\n给大家看一下我的这个国内模型\n接入的交接手册\n相当于这个文档\n是我从另外一个安装的过程中抽取出来了\n实际上这个交接手册\n这个执行的命令\n你也可以和GPT去沟通\n你跟他说我有这样的想法\n我要把Cloud Code接入\n这些非官方的模型应该怎么做\n你跟GPT对话很多轮之后\n让他给你给Codex的执行命令\n这里呢\n我是把我整个的安装过程\n抽象抽取了出来\n他会把那些坑都避雷掉\n另外一个人用了这个MD文档\n这个MarkDot文档\n就可以直接来用了\n机器能够快速的识别\n知道上下文\n然后知道你的目标\n你看这里的适用对象\n就是说一台全新的Mac mini\n由Codex远程执行安装和配置\n目标是让Cloud进入这样的模式\n通过CC switch本地路由去接入第三方的模型\n那个就是CC switch的界面\n一开始没用过的\n你可能都不知道怎么配\nCodex它本来有computer use这样的能力\n它可以直接帮你去配置和设置\n把那个模型窗口的命名设置\n大家刚刚看到的这里的Cloud Code\n这里的模型名称全部帮你改掉\n你可以是完全智能化AI原生的方式\n所以说这个过程才是真正的AI原生\n我们要想把AI用得好\n必须有这样的心态和这样的思路\n而不是再像传统的方式那样\n去哪里下载这个软件\n然后怎么配置一步步的跟着走\n你知不知道如果说中间遇到一些卡点\n你一步步截图跟GPT沟通\n或者跟其他模型沟通来来回回多少轮\n这里你用computer use\n用context就直接帮你搞定了\n会非常爽\n给大家快速的看一下这个文档\n首先会讲讲最终的效果\n要达到的目标\n然后Mac mini的前置条件\n然后接着就要安装Cloud Code的桌面端\n然后要安装CC switch\n下面是怎么写入这个Deep Seek的配置\n你看\n只要把这个文档丢给他\n我觉得哪怕你的电脑配置和我的不一样\n你也可以快速的完成\n即便没有你和GPT沟通出来\n也会有\n只不过说轮次多一点\n卡点多一点\n但是都比你那种原来传统的\n这种安装方式或者是接入方式\n要审视的多\n大家看一下我和Codex的沟通界面\n你看我就给他了这个文档 然后我让他参考我这个文档来给我安装 一开始我让他安装的是\nCode Wedding\n阿玲的Code Wedding\n但是Code Wedding它有个问题\n因为它是用来写Code的\n所以它是没有实施的搜索能力的\n大家可以看到\n这个其实就是Code Wedding的本身的回复\n你看我问的是特朗普2026年的访华\n你看他告诉我他的数据是截至2025年5月的\n但是我接入Deep6之后\n它其实它是有实施搜索能力的\n所以说这个方法就会非常的AI原生化\n我这里只是举了一个例子\n实际上大家在日常的工作生活中\n应该想办法把自己的所有的工作的模块\n都尽可能的AI原生化\n这个AI原生是说是你一个互相交互的东西\n你既要把你知道的去交互给他\n让他知道你的状态\n也要尽量避免自己直接输入东西\n说错了直接输出东西\n就是你要尽量让输出是AI给你的\n然后这个AI给你的东西\n你再交给这个另一个AI\nAI之间他们互相的这个可理解性比你更强\n长期这样的培养之后\n你的AI的思维会很强\n你会更能和AI对话\n你看这里对话就一轮\n我给他那个文档\n后面我要做的就是给他我阿里云的coding plan的API key\nAPI key这里要打码\n后面他这里就告诉我我的阿里云的\nCloud的coding plan已经接入成功了这里告诉我了\n然后呢\n但是当时我直接用的时候我发现他会有一些小bug\n比如说他这里的\n这个名字\n这里是没有改的\n然后呢我这个反馈给他\n同时我就直接启动了这个computer use\n他就直接帮我选项接入\n你要知道如果说没有context这样的辅助\n你走到这个接入的界面有多难\n有可能一个卡点就能卡你非常久\n大家现在看到的界面就是computer use\n通过cc switch的设置来接入cloud的桌面端的整个过程\n然后很快的阿里云的coding plan我接入完成之后\n我接着给他的是deep seek的apikey\nAPI Key这里还是要打一下码\n然后Deep Seek的V4的模型也介入成功了\n特别爽\n我觉得以后大家在处理问题的时候\n都要保持这样的AI原生的思路和思维\n然后我再把我这一轮的动作再抽象一下\n这样子我就形成了一个新的指南\n就是把前一版的文档\n没有抽取到没有覆盖到的东西\n我再把它抽取出来\n形成了一个更加完整的手册\n你看它的清晰度就会更高了\n而这一版的MD的文档\n在交接给下一个人来安装\n国内模型的接入的时候\n它中间遇到的卡点会更少\n这个就是不断迭代的一个过程\n怎么样我介绍的这个方法\n是不是特别的轻量级\n相关的文档可以到我的\n同名公众号里面去找\n而且我相信这种桌面终端\n也是大部分人都能够接受\n并且快速上手的一种方式\n国内的模型确实便宜\n我调用了好多轮\n用了半天\nDeepseek上面API接口才花了10块钱\n赶紧去试试吧\n记得订阅林姐说AI\n给我的节目点赞评论\n我们下期再见了\n拜拜", "language": "zh", "duration": 744.455, "segments": [ { "start": 0.0, "end": 2.0, "text": "大家好 欢迎来到林姐说AI", "chunk": 0, "language": "zh", "avg_logprob": -0.17782982535984204, "no_speech_prob": 2.564582227138068e-11, "compression_ratio": 1.2159090909090908 }, { "start": 2.0, "end": 4.0, "text": "这期视频跟大家分享一下", "chunk": 0, "language": "zh", "avg_logprob": -0.17782982535984204, "no_speech_prob": 2.564582227138068e-11, "compression_ratio": 1.2159090909090908 }, { "start": 4.0, "end": 7.0, "text": "如何用AI原生的方法", "chunk": 0, "language": "zh", "avg_logprob": -0.17782982535984204, "no_speech_prob": 2.564582227138068e-11, "compression_ratio": 1.2159090909090908 }, { "start": 7.0, "end": 10.5, "text": "将Cloud Code接入第三方的模型", "chunk": 0, "language": "zh", "avg_logprob": -0.17782982535984204, "no_speech_prob": 2.564582227138068e-11, "compression_ratio": 1.2159090909090908 }, { "start": 10.5, "end": 14.0, "text": "比如说国内的低成本的模型", "chunk": 0, "language": "zh", "avg_logprob": -0.17782982535984204, "no_speech_prob": 2.564582227138068e-11, "compression_ratio": 1.2159090909090908 }, { "start": 14.0, "end": 15.5, "text": "像Deepseek的V4", "chunk": 0, "language": "zh", "avg_logprob": -0.17782982535984204, "no_speech_prob": 2.564582227138068e-11, "compression_ratio": 1.2159090909090908 }, { "start": 15.5, "end": 21.0, "text": "大家看到这是我的Cloud的桌面端的GUI界面", "chunk": 0, "language": "zh", "avg_logprob": -0.17782982535984204, "no_speech_prob": 2.564582227138068e-11, "compression_ratio": 1.2159090909090908 }, { "start": 21.0, "end": 23.0, "text": "这个是我和他的聊天", "chunk": 0, "language": "zh", "avg_logprob": -0.17782982535984204, "no_speech_prob": 2.564582227138068e-11, "compression_ratio": 1.2159090909090908 }, { "start": 23.0, "end": 25.0, "text": "这里有个很神奇的地方", "chunk": 0, "language": "zh", "avg_logprob": -0.17782982535984204, "no_speech_prob": 2.564582227138068e-11, "compression_ratio": 1.2159090909090908 }, { "start": 25.0, "end": 27.0, "text": "看一下这里的模型选择", "chunk": 0, "language": "zh", "avg_logprob": -0.17782982535984204, "no_speech_prob": 2.564582227138068e-11, "compression_ratio": 1.2159090909090908 }, { "start": 27.0, "end": 33.14, "text": "则是Deep Seek的V4 Pro Flash各个版本的模型接入", "chunk": 0, "language": "zh", "avg_logprob": -0.11194757372140884, "no_speech_prob": 3.290576838788084e-11, "compression_ratio": 1.1660516605166051 }, { "start": 33.14, "end": 35.78, "text": "有没有一种穿越的感觉", "chunk": 0, "language": "zh", "avg_logprob": -0.11194757372140884, "no_speech_prob": 3.290576838788084e-11, "compression_ratio": 1.1660516605166051 }, { "start": 35.78, "end": 39.92, "text": "你看而且Deep Seek也可以快速的给我回复", "chunk": 0, "language": "zh", "avg_logprob": -0.11194757372140884, "no_speech_prob": 3.290576838788084e-11, "compression_ratio": 1.1660516605166051 }, { "start": 39.92, "end": 42.68, "text": "并且给我高质量的答复", "chunk": 0, "language": "zh", "avg_logprob": -0.11194757372140884, "no_speech_prob": 3.290576838788084e-11, "compression_ratio": 1.1660516605166051 }, { "start": 42.68, "end": 46.98, "text": "最近我主力的运营的运行的中心是Codex", "chunk": 0, "language": "zh", "avg_logprob": -0.11194757372140884, "no_speech_prob": 3.290576838788084e-11, "compression_ratio": 1.1660516605166051 }, { "start": 46.98, "end": 51.14, "text": "它是我执行非常多落地任务的手脚", "chunk": 0, "language": "zh", "avg_logprob": -0.11194757372140884, "no_speech_prob": 3.290576838788084e-11, "compression_ratio": 1.1660516605166051 }, { "start": 51.14, "end": 55.46, "text": "我是使用GPT加Codex的方式来完成整个的路", "chunk": 0, "language": "zh", "avg_logprob": -0.11194757372140884, "no_speech_prob": 3.290576838788084e-11, "compression_ratio": 1.1660516605166051 }, { "start": 55.46, "end": 56.5, "text": "但是呢", "chunk": 0, "language": "zh", "avg_logprob": -0.11969818904482085, "no_speech_prob": 1.6441421141211165e-11, "compression_ratio": 1.3369565217391304 }, { "start": 56.5, "end": 58.16, "text": "Codex也有一些坏处", "chunk": 0, "language": "zh", "avg_logprob": -0.11969818904482085, "no_speech_prob": 1.6441421141211165e-11, "compression_ratio": 1.3369565217391304 }, { "start": 58.16, "end": 60.22, "text": "我的这个模型的套餐", "chunk": 0, "language": "zh", "avg_logprob": -0.11969818904482085, "no_speech_prob": 1.6441421141211165e-11, "compression_ratio": 1.3369565217391304 }, { "start": 60.22, "end": 63.74, "text": "已经开到100刀Pro Lite的这个档位了", "chunk": 0, "language": "zh", "avg_logprob": -0.11969818904482085, "no_speech_prob": 1.6441421141211165e-11, "compression_ratio": 1.3369565217391304 }, { "start": 63.74, "end": 65.92, "text": "有可能马上又要升到200刀", "chunk": 0, "language": "zh", "avg_logprob": -0.11969818904482085, "no_speech_prob": 1.6441421141211165e-11, "compression_ratio": 1.3369565217391304 }, { "start": 65.92, "end": 68.78, "text": "但是因为网络不是特别稳定", "chunk": 0, "language": "zh", "avg_logprob": -0.11969818904482085, "no_speech_prob": 1.6441421141211165e-11, "compression_ratio": 1.3369565217391304 }, { "start": 68.78, "end": 72.5, "text": "它经常针对一些常任务会失败", "chunk": 0, "language": "zh", "avg_logprob": -0.11969818904482085, "no_speech_prob": 1.6441421141211165e-11, "compression_ratio": 1.3369565217391304 }, { "start": 72.5, "end": 76.22, "text": "就是一直会出现这个reconnect这个界面啊", "chunk": 0, "language": "zh", "avg_logprob": -0.11969818904482085, "no_speech_prob": 1.6441421141211165e-11, "compression_ratio": 1.3369565217391304 }, { "start": 76.22, "end": 77.34, "text": "任务经常会失败", "chunk": 0, "language": "zh", "avg_logprob": -0.11969818904482085, "no_speech_prob": 1.6441421141211165e-11, "compression_ratio": 1.3369565217391304 }, { "start": 77.34, "end": 79.92, "text": "特别是网络不太稳定的时候", "chunk": 0, "language": "zh", "avg_logprob": -0.11969818904482085, "no_speech_prob": 1.6441421141211165e-11, "compression_ratio": 1.3369565217391304 }, { "start": 79.92, "end": 81.18, "text": "这个时候呢", "chunk": 0, "language": "zh", "avg_logprob": -0.11969818904482085, "no_speech_prob": 1.6441421141211165e-11, "compression_ratio": 1.3369565217391304 }, { "start": 81.18, "end": 83.76, "text": "我的这个用完Codex的时候呢", "chunk": 0, "language": "zh", "avg_logprob": -0.11969818904482085, "no_speech_prob": 1.6441421141211165e-11, "compression_ratio": 1.3369565217391304 }, { "start": 83.76, "end": 87.6, "text": "我就需要再去用一下Cloud Code去挥挥雪", "chunk": 0, "language": "zh", "avg_logprob": -0.120979497130488, "no_speech_prob": 2.7623766121753768e-11, "compression_ratio": 1.3277591973244147 }, { "start": 87.6, "end": 92.52, "text": "用Cloud Code去接入我的这些比较便宜一些模型", "chunk": 0, "language": "zh", "avg_logprob": -0.120979497130488, "no_speech_prob": 2.7623766121753768e-11, "compression_ratio": 1.3277591973244147 }, { "start": 92.52, "end": 95.12, "text": "这样子我就有一个左右护法了", "chunk": 0, "language": "zh", "avg_logprob": -0.120979497130488, "no_speech_prob": 2.7623766121753768e-11, "compression_ratio": 1.3277591973244147 }, { "start": 95.12, "end": 98.32, "text": "一个Codex一个Cloud Code加第三方模型", "chunk": 0, "language": "zh", "avg_logprob": -0.120979497130488, "no_speech_prob": 2.7623766121753768e-11, "compression_ratio": 1.3277591973244147 }, { "start": 98.32, "end": 99.6, "text": "再加上GPT", "chunk": 0, "language": "zh", "avg_logprob": -0.120979497130488, "no_speech_prob": 2.7623766121753768e-11, "compression_ratio": 1.3277591973244147 }, { "start": 99.6, "end": 103.52, "text": "让我整个的运行项目的效率变得更高", "chunk": 0, "language": "zh", "avg_logprob": -0.120979497130488, "no_speech_prob": 2.7623766121753768e-11, "compression_ratio": 1.3277591973244147 }, { "start": 103.52, "end": 104.68, "text": "有一个平衡点吧", "chunk": 0, "language": "zh", "avg_logprob": -0.120979497130488, "no_speech_prob": 2.7623766121753768e-11, "compression_ratio": 1.3277591973244147 }, { "start": 104.68, "end": 106.52, "text": "而且用起来真的挺爽的", "chunk": 0, "language": "zh", "avg_logprob": -0.120979497130488, "no_speech_prob": 2.7623766121753768e-11, "compression_ratio": 1.3277591973244147 }, { "start": 106.52, "end": 109.08, "text": "这一期我会给大家分享这个方法", "chunk": 0, "language": "zh", "avg_logprob": -0.120979497130488, "no_speech_prob": 2.7623766121753768e-11, "compression_ratio": 1.3277591973244147 }, { "start": 109.08, "end": 113.64, "text": "而且我今天分享的方法是非常AI原生化的", "chunk": 0, "language": "zh", "avg_logprob": -0.120979497130488, "no_speech_prob": 2.7623766121753768e-11, "compression_ratio": 1.3277591973244147 }, { "start": 113.64, "end": 116.72, "text": "有的人曾经在我的节目里面留言说", "chunk": 0, "language": "zh", "avg_logprob": -0.11732429265975952, "no_speech_prob": 2.9242139959473334e-11, "compression_ratio": 1.2591362126245846 }, { "start": 116.72, "end": 120.42, "text": "那你都用国内的第三方的一些模型了", "chunk": 0, "language": "zh", "avg_logprob": -0.11732429265975952, "no_speech_prob": 2.9242139959473334e-11, "compression_ratio": 1.2591362126245846 }, { "start": 120.42, "end": 122.82, "text": "那你还为什么要接Cloud Code呢", "chunk": 0, "language": "zh", "avg_logprob": -0.11732429265975952, "no_speech_prob": 2.9242139959473334e-11, "compression_ratio": 1.2591362126245846 }, { "start": 122.82, "end": 124.68, "text": "这不是多此一举吗", "chunk": 0, "language": "zh", "avg_logprob": -0.11732429265975952, "no_speech_prob": 2.9242139959473334e-11, "compression_ratio": 1.2591362126245846 }, { "start": 124.68, "end": 127.26, "text": "我告诉你非常有必要", "chunk": 0, "language": "zh", "avg_logprob": -0.11732429265975952, "no_speech_prob": 2.9242139959473334e-11, "compression_ratio": 1.2591362126245846 }, { "start": 127.26, "end": 127.84, "text": "首先啊", "chunk": 0, "language": "zh", "avg_logprob": -0.11732429265975952, "no_speech_prob": 2.9242139959473334e-11, "compression_ratio": 1.2591362126245846 }, { "start": 127.84, "end": 129.0, "text": "你用过Cloud Code", "chunk": 0, "language": "zh", "avg_logprob": -0.11732429265975952, "no_speech_prob": 2.9242139959473334e-11, "compression_ratio": 1.2591362126245846 }, { "start": 129.0, "end": 130.3, "text": "你就知道Cloud Code", "chunk": 0, "language": "zh", "avg_logprob": -0.11732429265975952, "no_speech_prob": 2.9242139959473334e-11, "compression_ratio": 1.2591362126245846 }, { "start": 130.3, "end": 135.62, "text": "它的Harness整个的工作流的框架是非常强的", "chunk": 0, "language": "zh", "avg_logprob": -0.11732429265975952, "no_speech_prob": 2.9242139959473334e-11, "compression_ratio": 1.2591362126245846 }, { "start": 135.62, "end": 139.52, "text": "它之前不是还把自己整个托管的体系拿出来卖吗", "chunk": 0, "language": "zh", "avg_logprob": -0.11732429265975952, "no_speech_prob": 2.9242139959473334e-11, "compression_ratio": 1.2591362126245846 }, { "start": 139.52, "end": 142.0, "text": "这个其实就是它的底气", "chunk": 0, "language": "zh", "avg_logprob": -0.11732429265975952, "no_speech_prob": 2.9242139959473334e-11, "compression_ratio": 1.2591362126245846 }, { "start": 142.0, "end": 146.62, "text": "它里面是有一个很好的规划过的agent loop的", "chunk": 0, "language": "zh", "avg_logprob": -0.10768770815721199, "no_speech_prob": 1.4557148889093874e-11, "compression_ratio": 1.275735294117647 }, { "start": 146.62, "end": 149.48, "text": "它可以去自己去规划任务", "chunk": 0, "language": "zh", "avg_logprob": -0.10768770815721199, "no_speech_prob": 1.4557148889093874e-11, "compression_ratio": 1.275735294117647 }, { "start": 149.48, "end": 152.64, "text": "要用工具读取结果继续执行", "chunk": 0, "language": "zh", "avg_logprob": -0.10768770815721199, "no_speech_prob": 1.4557148889093874e-11, "compression_ratio": 1.275735294117647 }, { "start": 152.64, "end": 156.6, "text": "Cloud Code它整个的文件系统的能力", "chunk": 0, "language": "zh", "avg_logprob": -0.10768770815721199, "no_speech_prob": 1.4557148889093874e-11, "compression_ratio": 1.275735294117647 }, { "start": 156.6, "end": 160.0, "text": "包括终端的能力都是非常强的", "chunk": 0, "language": "zh", "avg_logprob": -0.10768770815721199, "no_speech_prob": 1.4557148889093874e-11, "compression_ratio": 1.275735294117647 }, { "start": 160.0, "end": 162.92, "text": "它现在整体的估值这么高", "chunk": 0, "language": "zh", "avg_logprob": -0.10768770815721199, "no_speech_prob": 1.4557148889093874e-11, "compression_ratio": 1.275735294117647 }, { "start": 162.92, "end": 166.3, "text": "每年的这个AR这么高都是有道理的", "chunk": 0, "language": "zh", "avg_logprob": -0.10768770815721199, "no_speech_prob": 1.4557148889093874e-11, "compression_ratio": 1.275735294117647 }, { "start": 166.3, "end": 167.76, "text": "它是真的强", "chunk": 0, "language": "zh", "avg_logprob": -0.10768770815721199, "no_speech_prob": 1.4557148889093874e-11, "compression_ratio": 1.275735294117647 }, { "start": 167.76, "end": 171.12, "text": "而且在Cloud Code这个桌面端里面的应用", "chunk": 0, "language": "zh", "avg_logprob": -0.10768770815721199, "no_speech_prob": 1.4557148889093874e-11, "compression_ratio": 1.275735294117647 }, { "start": 171.12, "end": 175.14, "text": "普通的知识工作者或者普通的职场人士", "chunk": 0, "language": "zh", "avg_logprob": -0.07208879229048608, "no_speech_prob": 2.106929561374038e-11, "compression_ratio": 1.3971631205673758 }, { "start": 175.14, "end": 177.4, "text": "也能够很好的去使用", "chunk": 0, "language": "zh", "avg_logprob": -0.07208879229048608, "no_speech_prob": 2.106929561374038e-11, "compression_ratio": 1.3971631205673758 }, { "start": 177.4, "end": 180.88, "text": "Codex它的桌面段引起了一个新的风潮", "chunk": 0, "language": "zh", "avg_logprob": -0.07208879229048608, "no_speech_prob": 2.106929561374038e-11, "compression_ratio": 1.3971631205673758 }, { "start": 180.88, "end": 182.56, "text": "就是其实在未来", "chunk": 0, "language": "zh", "avg_logprob": -0.07208879229048608, "no_speech_prob": 2.106929561374038e-11, "compression_ratio": 1.3971631205673758 }, { "start": 182.56, "end": 185.36, "text": "这个CLI的这个模式是更加先进的", "chunk": 0, "language": "zh", "avg_logprob": -0.07208879229048608, "no_speech_prob": 2.106929561374038e-11, "compression_ratio": 1.3971631205673758 }, { "start": 185.36, "end": 186.76, "text": "而且它是精确的", "chunk": 0, "language": "zh", "avg_logprob": -0.07208879229048608, "no_speech_prob": 2.106929561374038e-11, "compression_ratio": 1.3971631205673758 }, { "start": 186.76, "end": 190.68, "text": "我也非常提倡大家去使用CLI的模式", "chunk": 0, "language": "zh", "avg_logprob": -0.07208879229048608, "no_speech_prob": 2.106929561374038e-11, "compression_ratio": 1.3971631205673758 }, { "start": 190.68, "end": 193.86, "text": "但是对于大部分的普通的工作者而言", "chunk": 0, "language": "zh", "avg_logprob": -0.07208879229048608, "no_speech_prob": 2.106929561374038e-11, "compression_ratio": 1.3971631205673758 }, { "start": 193.86, "end": 196.94, "text": "这个GUI的模式还是大家更适应的模式", "chunk": 0, "language": "zh", "avg_logprob": -0.07208879229048608, "no_speech_prob": 2.106929561374038e-11, "compression_ratio": 1.3971631205673758 }, { "start": 196.94, "end": 199.72, "text": "而且这个Codex现在真的是非常流行", "chunk": 0, "language": "zh", "avg_logprob": -0.07208879229048608, "no_speech_prob": 2.106929561374038e-11, "compression_ratio": 1.3971631205673758 }, { "start": 199.72, "end": 202.74, "text": "大家对这样的工作方式也是比较适应的", "chunk": 0, "language": "zh", "avg_logprob": -0.07828542921278211, "no_speech_prob": 1.605223592937577e-11, "compression_ratio": 1.2589928057553956 }, { "start": 202.74, "end": 205.58, "text": "所以在Cloud Code的桌面端", "chunk": 0, "language": "zh", "avg_logprob": -0.07828542921278211, "no_speech_prob": 1.605223592937577e-11, "compression_ratio": 1.2589928057553956 }, { "start": 205.58, "end": 207.98, "text": "也去接入第三方的模型", "chunk": 0, "language": "zh", "avg_logprob": -0.07828542921278211, "no_speech_prob": 1.605223592937577e-11, "compression_ratio": 1.2589928057553956 }, { "start": 207.98, "end": 210.22, "text": "它中间有一个能力", "chunk": 0, "language": "zh", "avg_logprob": -0.07828542921278211, "no_speech_prob": 1.605223592937577e-11, "compression_ratio": 1.2589928057553956 }, { "start": 210.22, "end": 211.14, "text": "就是Cowork", "chunk": 0, "language": "zh", "avg_logprob": -0.07828542921278211, "no_speech_prob": 1.605223592937577e-11, "compression_ratio": 1.2589928057553956 }, { "start": 211.14, "end": 213.96, "text": "Cowork区别于Cloud Code", "chunk": 0, "language": "zh", "avg_logprob": -0.07828542921278211, "no_speech_prob": 1.605223592937577e-11, "compression_ratio": 1.2589928057553956 }, { "start": 213.96, "end": 215.34, "text": "就是写代码的能力", "chunk": 0, "language": "zh", "avg_logprob": -0.07828542921278211, "no_speech_prob": 1.605223592937577e-11, "compression_ratio": 1.2589928057553956 }, { "start": 215.34, "end": 218.6, "text": "它相当于是给大家普通的这种工作者", "chunk": 0, "language": "zh", "avg_logprob": -0.07828542921278211, "no_speech_prob": 1.605223592937577e-11, "compression_ratio": 1.2589928057553956 }, { "start": 218.6, "end": 221.22, "text": "也可以利用这个模型框架", "chunk": 0, "language": "zh", "avg_logprob": -0.07828542921278211, "no_speech_prob": 1.605223592937577e-11, "compression_ratio": 1.2589928057553956 }, { "start": 221.22, "end": 225.22, "text": "去解决一些具体的自主工作的问题", "chunk": 0, "language": "zh", "avg_logprob": -0.07828542921278211, "no_speech_prob": 1.605223592937577e-11, "compression_ratio": 1.2589928057553956 }, { "start": 225.22, "end": 227.56, "text": "并且能够返回完成品", "chunk": 0, "language": "zh", "avg_logprob": -0.07828542921278211, "no_speech_prob": 1.605223592937577e-11, "compression_ratio": 1.2589928057553956 }, { "start": 227.56, "end": 233.2, "text": "Cloud Code本身会更加偏开发者终端和一些代码项目和工程项目", "chunk": 0, "language": "zh", "avg_logprob": -0.10887247895541256, "no_speech_prob": 2.5038757525686073e-11, "compression_ratio": 1.224755700325733 }, { "start": 233.2, "end": 237.98, "text": "那Cloud Code Cowork会这种集群还有封群的模式", "chunk": 0, "language": "zh", "avg_logprob": -0.10887247895541256, "no_speech_prob": 2.5038757525686073e-11, "compression_ratio": 1.224755700325733 }, { "start": 237.98, "end": 242.04, "text": "会更加适合普通的办公任务本地文件", "chunk": 0, "language": "zh", "avg_logprob": -0.10887247895541256, "no_speech_prob": 2.5038757525686073e-11, "compression_ratio": 1.224755700325733 }, { "start": 242.04, "end": 243.72, "text": "还有应用操作和交付成果", "chunk": 0, "language": "zh", "avg_logprob": -0.10887247895541256, "no_speech_prob": 2.5038757525686073e-11, "compression_ratio": 1.224755700325733 }, { "start": 243.72, "end": 245.42, "text": "我自己在使用的时候", "chunk": 0, "language": "zh", "avg_logprob": -0.10887247895541256, "no_speech_prob": 2.5038757525686073e-11, "compression_ratio": 1.224755700325733 }, { "start": 245.42, "end": 250.34, "text": "我觉得它的整个的Harness框架加上Deep Seek V4", "chunk": 0, "language": "zh", "avg_logprob": -0.10887247895541256, "no_speech_prob": 2.5038757525686073e-11, "compression_ratio": 1.224755700325733 }, { "start": 250.34, "end": 252.74, "text": "或者类似这样的能力的模型", "chunk": 0, "language": "zh", "avg_logprob": -0.10887247895541256, "no_speech_prob": 2.5038757525686073e-11, "compression_ratio": 1.224755700325733 }, { "start": 252.74, "end": 255.08, "text": "它的整个的产出结果还是不错的", "chunk": 0, "language": "zh", "avg_logprob": -0.10887247895541256, "no_speech_prob": 2.5038757525686073e-11, "compression_ratio": 1.224755700325733 }, { "start": 255.08, "end": 257.86, "text": "而且它可以和GPT去协同", "chunk": 0, "language": "zh", "avg_logprob": -0.10238349519926926, "no_speech_prob": 1.5845696282879018e-11, "compression_ratio": 1.2114093959731544 }, { "start": 257.86, "end": 259.76, "text": "它最后给出的结果", "chunk": 0, "language": "zh", "avg_logprob": -0.10238349519926926, "no_speech_prob": 1.5845696282879018e-11, "compression_ratio": 1.2114093959731544 }, { "start": 259.76, "end": 261.08, "text": "GPT也认可了", "chunk": 0, "language": "zh", "avg_logprob": -0.10238349519926926, "no_speech_prob": 1.5845696282879018e-11, "compression_ratio": 1.2114093959731544 }, { "start": 261.08, "end": 265.66, "text": "可见它在这样的框架下的交付能力生成结果", "chunk": 0, "language": "zh", "avg_logprob": -0.10238349519926926, "no_speech_prob": 1.5845696282879018e-11, "compression_ratio": 1.2114093959731544 }, { "start": 265.66, "end": 266.96, "text": "是比较好的", "chunk": 0, "language": "zh", "avg_logprob": -0.10238349519926926, "no_speech_prob": 1.5845696282879018e-11, "compression_ratio": 1.2114093959731544 }, { "start": 266.96, "end": 268.7, "text": "质量是相对高的", "chunk": 0, "language": "zh", "avg_logprob": -0.10238349519926926, "no_speech_prob": 1.5845696282879018e-11, "compression_ratio": 1.2114093959731544 }, { "start": 268.7, "end": 271.38, "text": "而且我自己的使用体验下来呢", "chunk": 0, "language": "zh", "avg_logprob": -0.10238349519926926, "no_speech_prob": 1.5845696282879018e-11, "compression_ratio": 1.2114093959731544 }, { "start": 271.38, "end": 272.58, "text": "我会感觉啊", "chunk": 0, "language": "zh", "avg_logprob": -0.10238349519926926, "no_speech_prob": 1.5845696282879018e-11, "compression_ratio": 1.2114093959731544 }, { "start": 272.58, "end": 275.42, "text": "现在Cloud Code和Codex的", "chunk": 0, "language": "zh", "avg_logprob": -0.10238349519926926, "no_speech_prob": 1.5845696282879018e-11, "compression_ratio": 1.2114093959731544 }, { "start": 275.42, "end": 278.06, "text": "它的整个的Harness框架是非常强的", "chunk": 0, "language": "zh", "avg_logprob": -0.10238349519926926, "no_speech_prob": 1.5845696282879018e-11, "compression_ratio": 1.2114093959731544 }, { "start": 278.06, "end": 281.28, "text": "他们两家最近的动作也是最多的", "chunk": 0, "language": "zh", "avg_logprob": -0.10238349519926926, "no_speech_prob": 1.5845696282879018e-11, "compression_ratio": 1.2114093959731544 }, { "start": 281.28, "end": 284.36, "text": "我有一个非常超前的想象", "chunk": 0, "language": "zh", "avg_logprob": -0.10238349519926926, "no_speech_prob": 1.5845696282879018e-11, "compression_ratio": 1.2114093959731544 }, { "start": 284.36, "end": 291.02, "text": "我觉得可以用Cloud Code和Harness这两家的Harness框架", "chunk": 0, "language": "zh", "avg_logprob": -0.0723419189453125, "no_speech_prob": 1.3153127892406857e-11, "compression_ratio": 1.3417721518987342 }, { "start": 291.02, "end": 296.36, "text": "可以把它抽取出来去调用任何一家的大模型能力", "chunk": 0, "language": "zh", "avg_logprob": -0.0723419189453125, "no_speech_prob": 1.3153127892406857e-11, "compression_ratio": 1.3417721518987342 }, { "start": 296.36, "end": 297.94, "text": "我们可以把它处理成这样", "chunk": 0, "language": "zh", "avg_logprob": -0.0723419189453125, "no_speech_prob": 1.3153127892406857e-11, "compression_ratio": 1.3417721518987342 }, { "start": 297.94, "end": 300.44, "text": "有了这样比较好的Harness框架", "chunk": 0, "language": "zh", "avg_logprob": -0.0723419189453125, "no_speech_prob": 1.3153127892406857e-11, "compression_ratio": 1.3417721518987342 }, { "start": 300.44, "end": 304.84, "text": "我们可以在顶层去调动不同厂商的大模型", "chunk": 0, "language": "zh", "avg_logprob": -0.0723419189453125, "no_speech_prob": 1.3153127892406857e-11, "compression_ratio": 1.3417721518987342 }, { "start": 304.84, "end": 311.28, "text": "而且我们可以在顶层统一的去管理各个agent的Skier MCP", "chunk": 0, "language": "zh", "avg_logprob": -0.0723419189453125, "no_speech_prob": 1.3153127892406857e-11, "compression_ratio": 1.3417721518987342 }, { "start": 311.28, "end": 313.96, "text": "哇我觉得真的是非常多的", "chunk": 0, "language": "zh", "avg_logprob": -0.1063194965970689, "no_speech_prob": 1.3418904876161264e-11, "compression_ratio": 1.226027397260274 }, { "start": 313.96, "end": 315.32, "text": "这个可以玩的东西", "chunk": 0, "language": "zh", "avg_logprob": -0.1063194965970689, "no_speech_prob": 1.3418904876161264e-11, "compression_ratio": 1.226027397260274 }, { "start": 315.32, "end": 316.84, "text": "好说回来啊", "chunk": 0, "language": "zh", "avg_logprob": -0.1063194965970689, "no_speech_prob": 1.3418904876161264e-11, "compression_ratio": 1.226027397260274 }, { "start": 316.84, "end": 319.04, "text": "我再跟大家切实的去分享", "chunk": 0, "language": "zh", "avg_logprob": -0.1063194965970689, "no_speech_prob": 1.3418904876161264e-11, "compression_ratio": 1.226027397260274 }, { "start": 319.04, "end": 321.36, "text": "怎么去做这样的接入", "chunk": 0, "language": "zh", "avg_logprob": -0.1063194965970689, "no_speech_prob": 1.3418904876161264e-11, "compression_ratio": 1.226027397260274 }, { "start": 321.36, "end": 323.58, "text": "把Cloud Code的GUI桌面界面", "chunk": 0, "language": "zh", "avg_logprob": -0.1063194965970689, "no_speech_prob": 1.3418904876161264e-11, "compression_ratio": 1.226027397260274 }, { "start": 323.58, "end": 325.54, "text": "去接入我们的Deep Seek V4", "chunk": 0, "language": "zh", "avg_logprob": -0.1063194965970689, "no_speech_prob": 1.3418904876161264e-11, "compression_ratio": 1.226027397260274 }, { "start": 325.54, "end": 327.14, "text": "我是以Deep Seek为例", "chunk": 0, "language": "zh", "avg_logprob": -0.1063194965970689, "no_speech_prob": 1.3418904876161264e-11, "compression_ratio": 1.226027397260274 }, { "start": 327.14, "end": 329.06, "text": "其实各家的这个模型", "chunk": 0, "language": "zh", "avg_logprob": -0.1063194965970689, "no_speech_prob": 1.3418904876161264e-11, "compression_ratio": 1.226027397260274 }, { "start": 329.06, "end": 330.28, "text": "都是可以接入的", "chunk": 0, "language": "zh", "avg_logprob": -0.1063194965970689, "no_speech_prob": 1.3418904876161264e-11, "compression_ratio": 1.226027397260274 }, { "start": 330.28, "end": 333.42, "text": "我这里讲的AI原生的方法", "chunk": 0, "language": "zh", "avg_logprob": -0.1063194965970689, "no_speech_prob": 1.3418904876161264e-11, "compression_ratio": 1.226027397260274 }, { "start": 333.42, "end": 335.2, "text": "就是用Codex", "chunk": 0, "language": "zh", "avg_logprob": -0.1063194965970689, "no_speech_prob": 1.3418904876161264e-11, "compression_ratio": 1.226027397260274 }, { "start": 335.2, "end": 337.62, "text": "给它具体的接入文档", "chunk": 0, "language": "zh", "avg_logprob": -0.1063194965970689, "no_speech_prob": 1.3418904876161264e-11, "compression_ratio": 1.226027397260274 }, { "start": 337.62, "end": 339.88, "text": "让Codex成为手脚", "chunk": 0, "language": "zh", "avg_logprob": -0.1063194965970689, "no_speech_prob": 1.3418904876161264e-11, "compression_ratio": 1.226027397260274 }, { "start": 339.88, "end": 344.0, "text": "来帮你使用CC-Switch这样的工具", "chunk": 0, "language": "zh", "avg_logprob": -0.11990885702979485, "no_speech_prob": 1.3631097119104485e-11, "compression_ratio": 1.293929712460064 }, { "start": 344.0, "end": 348.26, "text": "把Cloud Code的 Desktop的桌面端的界面", "chunk": 0, "language": "zh", "avg_logprob": -0.11990885702979485, "no_speech_prob": 1.3631097119104485e-11, "compression_ratio": 1.293929712460064 }, { "start": 348.26, "end": 351.12, "text": "接入像DeepSeek、阿里云或者CodingPlan", "chunk": 0, "language": "zh", "avg_logprob": -0.11990885702979485, "no_speech_prob": 1.3631097119104485e-11, "compression_ratio": 1.293929712460064 }, { "start": 351.12, "end": 355.54, "text": "或者其他的一些成本比较低的国内的模型", "chunk": 0, "language": "zh", "avg_logprob": -0.11990885702979485, "no_speech_prob": 1.3631097119104485e-11, "compression_ratio": 1.293929712460064 }, { "start": 355.54, "end": 357.82, "text": "这里我是给了他一个文档的", "chunk": 0, "language": "zh", "avg_logprob": -0.11990885702979485, "no_speech_prob": 1.3631097119104485e-11, "compression_ratio": 1.293929712460064 }, { "start": 357.82, "end": 360.52, "text": "给大家看一下我的这个国内模型", "chunk": 0, "language": "zh", "avg_logprob": -0.11990885702979485, "no_speech_prob": 1.3631097119104485e-11, "compression_ratio": 1.293929712460064 }, { "start": 360.52, "end": 361.72, "text": "接入的交接手册", "chunk": 0, "language": "zh", "avg_logprob": -0.11990885702979485, "no_speech_prob": 1.3631097119104485e-11, "compression_ratio": 1.293929712460064 }, { "start": 361.72, "end": 363.38, "text": "相当于这个文档", "chunk": 0, "language": "zh", "avg_logprob": -0.11990885702979485, "no_speech_prob": 1.3631097119104485e-11, "compression_ratio": 1.293929712460064 }, { "start": 363.38, "end": 366.32, "text": "是我从另外一个安装的过程中抽取出来了", "chunk": 0, "language": "zh", "avg_logprob": -0.11990885702979485, "no_speech_prob": 1.3631097119104485e-11, "compression_ratio": 1.293929712460064 }, { "start": 366.32, "end": 368.4, "text": "实际上这个交接手册", "chunk": 0, "language": "zh", "avg_logprob": -0.11990885702979485, "no_speech_prob": 1.3631097119104485e-11, "compression_ratio": 1.293929712460064 }, { "start": 368.4, "end": 369.4, "text": "这个执行的命令", "chunk": 0, "language": "zh", "avg_logprob": -0.11990885702979485, "no_speech_prob": 1.3631097119104485e-11, "compression_ratio": 1.293929712460064 }, { "start": 369.4, "end": 372.24, "text": "你也可以和GPT去沟通", "chunk": 0, "language": "zh", "avg_logprob": -0.11422332587269689, "no_speech_prob": 2.8822595352084868e-11, "compression_ratio": 1.2419825072886297 }, { "start": 372.24, "end": 373.9, "text": "你跟他说我有这样的想法", "chunk": 0, "language": "zh", "avg_logprob": -0.11422332587269689, "no_speech_prob": 2.8822595352084868e-11, "compression_ratio": 1.2419825072886297 }, { "start": 373.9, "end": 375.78, "text": "我要把Cloud Code接入", "chunk": 0, "language": "zh", "avg_logprob": -0.11422332587269689, "no_speech_prob": 2.8822595352084868e-11, "compression_ratio": 1.2419825072886297 }, { "start": 375.78, "end": 378.16, "text": "这些非官方的模型应该怎么做", "chunk": 0, "language": "zh", "avg_logprob": -0.11422332587269689, "no_speech_prob": 2.8822595352084868e-11, "compression_ratio": 1.2419825072886297 }, { "start": 378.16, "end": 380.6, "text": "你跟GPT对话很多轮之后", "chunk": 0, "language": "zh", "avg_logprob": -0.11422332587269689, "no_speech_prob": 2.8822595352084868e-11, "compression_ratio": 1.2419825072886297 }, { "start": 380.6, "end": 383.96, "text": "让他给你给Codex的执行命令", "chunk": 0, "language": "zh", "avg_logprob": -0.11422332587269689, "no_speech_prob": 2.8822595352084868e-11, "compression_ratio": 1.2419825072886297 }, { "start": 383.96, "end": 384.7, "text": "这里呢", "chunk": 0, "language": "zh", "avg_logprob": -0.11422332587269689, "no_speech_prob": 2.8822595352084868e-11, "compression_ratio": 1.2419825072886297 }, { "start": 384.7, "end": 387.28, "text": "我是把我整个的安装过程", "chunk": 0, "language": "zh", "avg_logprob": -0.11422332587269689, "no_speech_prob": 2.8822595352084868e-11, "compression_ratio": 1.2419825072886297 }, { "start": 387.28, "end": 388.86, "text": "抽象抽取了出来", "chunk": 0, "language": "zh", "avg_logprob": -0.11422332587269689, "no_speech_prob": 2.8822595352084868e-11, "compression_ratio": 1.2419825072886297 }, { "start": 388.86, "end": 390.88, "text": "他会把那些坑都避雷掉", "chunk": 0, "language": "zh", "avg_logprob": -0.11422332587269689, "no_speech_prob": 2.8822595352084868e-11, "compression_ratio": 1.2419825072886297 }, { "start": 390.88, "end": 393.6, "text": "另外一个人用了这个MD文档", "chunk": 0, "language": "zh", "avg_logprob": -0.11422332587269689, "no_speech_prob": 2.8822595352084868e-11, "compression_ratio": 1.2419825072886297 }, { "start": 393.6, "end": 394.38, "text": "这个MarkDot文档", "chunk": 0, "language": "zh", "avg_logprob": -0.11422332587269689, "no_speech_prob": 2.8822595352084868e-11, "compression_ratio": 1.2419825072886297 }, { "start": 394.38, "end": 395.8, "text": "就可以直接来用了", "chunk": 0, "language": "zh", "avg_logprob": -0.11422332587269689, "no_speech_prob": 2.8822595352084868e-11, "compression_ratio": 1.2419825072886297 }, { "start": 395.8, "end": 397.32, "text": "机器能够快速的识别", "chunk": 0, "language": "zh", "avg_logprob": -0.11422332587269689, "no_speech_prob": 2.8822595352084868e-11, "compression_ratio": 1.2419825072886297 }, { "start": 397.32, "end": 397.98, "text": "知道上下文", "chunk": 0, "language": "zh", "avg_logprob": -0.11422332587269689, "no_speech_prob": 2.8822595352084868e-11, "compression_ratio": 1.2419825072886297 }, { "start": 397.98, "end": 398.92, "text": "然后知道你的目标", "chunk": 0, "language": "zh", "avg_logprob": -0.11422332587269689, "no_speech_prob": 2.8822595352084868e-11, "compression_ratio": 1.2419825072886297 }, { "start": 398.92, "end": 400.98, "text": "你看这里的适用对象", "chunk": 0, "language": "zh", "avg_logprob": -0.08483867512808906, "no_speech_prob": 2.634368417853139e-11, "compression_ratio": 1.249190938511327 }, { "start": 400.98, "end": 403.1, "text": "就是说一台全新的Mac mini", "chunk": 0, "language": "zh", "avg_logprob": -0.08483867512808906, "no_speech_prob": 2.634368417853139e-11, "compression_ratio": 1.249190938511327 }, { "start": 403.1, "end": 406.28, "text": "由Codex远程执行安装和配置", "chunk": 0, "language": "zh", "avg_logprob": -0.08483867512808906, "no_speech_prob": 2.634368417853139e-11, "compression_ratio": 1.249190938511327 }, { "start": 406.28, "end": 408.74, "text": "目标是让Cloud进入这样的模式", "chunk": 0, "language": "zh", "avg_logprob": -0.08483867512808906, "no_speech_prob": 2.634368417853139e-11, "compression_ratio": 1.249190938511327 }, { "start": 408.74, "end": 413.7, "text": "通过CC switch本地路由去接入第三方的模型", "chunk": 0, "language": "zh", "avg_logprob": -0.08483867512808906, "no_speech_prob": 2.634368417853139e-11, "compression_ratio": 1.249190938511327 }, { "start": 413.7, "end": 415.7, "text": "那个就是CC switch的界面", "chunk": 0, "language": "zh", "avg_logprob": -0.08483867512808906, "no_speech_prob": 2.634368417853139e-11, "compression_ratio": 1.249190938511327 }, { "start": 415.7, "end": 417.1, "text": "一开始没用过的", "chunk": 0, "language": "zh", "avg_logprob": -0.08483867512808906, "no_speech_prob": 2.634368417853139e-11, "compression_ratio": 1.249190938511327 }, { "start": 417.1, "end": 418.3, "text": "你可能都不知道怎么配", "chunk": 0, "language": "zh", "avg_logprob": -0.08483867512808906, "no_speech_prob": 2.634368417853139e-11, "compression_ratio": 1.249190938511327 }, { "start": 418.3, "end": 422.32, "text": "Codex它本来有computer use这样的能力", "chunk": 0, "language": "zh", "avg_logprob": -0.08483867512808906, "no_speech_prob": 2.634368417853139e-11, "compression_ratio": 1.249190938511327 }, { "start": 422.32, "end": 425.64, "text": "它可以直接帮你去配置和设置", "chunk": 0, "language": "zh", "avg_logprob": -0.08483867512808906, "no_speech_prob": 2.634368417853139e-11, "compression_ratio": 1.249190938511327 }, { "start": 425.64, "end": 428.44, "text": "把那个模型窗口的命名设置", "chunk": 0, "language": "zh", "avg_logprob": -0.08483867512808906, "no_speech_prob": 2.634368417853139e-11, "compression_ratio": 1.249190938511327 }, { "start": 428.44, "end": 431.92, "text": "大家刚刚看到的这里的Cloud Code", "chunk": 0, "language": "zh", "avg_logprob": -0.10079373082806987, "no_speech_prob": 1.7482163286453023e-11, "compression_ratio": 1.2389705882352942 }, { "start": 431.92, "end": 434.58, "text": "这里的模型名称全部帮你改掉", "chunk": 0, "language": "zh", "avg_logprob": -0.10079373082806987, "no_speech_prob": 1.7482163286453023e-11, "compression_ratio": 1.2389705882352942 }, { "start": 434.58, "end": 439.4, "text": "你可以是完全智能化AI原生的方式", "chunk": 0, "language": "zh", "avg_logprob": -0.10079373082806987, "no_speech_prob": 1.7482163286453023e-11, "compression_ratio": 1.2389705882352942 }, { "start": 439.4, "end": 442.9, "text": "所以说这个过程才是真正的AI原生", "chunk": 0, "language": "zh", "avg_logprob": -0.10079373082806987, "no_speech_prob": 1.7482163286453023e-11, "compression_ratio": 1.2389705882352942 }, { "start": 442.9, "end": 445.42, "text": "我们要想把AI用得好", "chunk": 0, "language": "zh", "avg_logprob": -0.10079373082806987, "no_speech_prob": 1.7482163286453023e-11, "compression_ratio": 1.2389705882352942 }, { "start": 445.42, "end": 449.0, "text": "必须有这样的心态和这样的思路", "chunk": 0, "language": "zh", "avg_logprob": -0.10079373082806987, "no_speech_prob": 1.7482163286453023e-11, "compression_ratio": 1.2389705882352942 }, { "start": 449.0, "end": 452.36, "text": "而不是再像传统的方式那样", "chunk": 0, "language": "zh", "avg_logprob": -0.10079373082806987, "no_speech_prob": 1.7482163286453023e-11, "compression_ratio": 1.2389705882352942 }, { "start": 452.36, "end": 454.56, "text": "去哪里下载这个软件", "chunk": 0, "language": "zh", "avg_logprob": -0.10079373082806987, "no_speech_prob": 1.7482163286453023e-11, "compression_ratio": 1.2389705882352942 }, { "start": 454.56, "end": 456.5, "text": "然后怎么配置一步步的跟着走", "chunk": 0, "language": "zh", "avg_logprob": -0.10079373082806987, "no_speech_prob": 1.7482163286453023e-11, "compression_ratio": 1.2389705882352942 }, { "start": 456.5, "end": 459.1, "text": "你知不知道如果说中间遇到一些卡点", "chunk": 0, "language": "zh", "avg_logprob": -0.11760819753011068, "no_speech_prob": 2.595616083178598e-11, "compression_ratio": 1.186335403726708 }, { "start": 459.1, "end": 461.56, "text": "你一步步截图跟GPT沟通", "chunk": 0, "language": "zh", "avg_logprob": -0.11760819753011068, "no_speech_prob": 2.595616083178598e-11, "compression_ratio": 1.186335403726708 }, { "start": 461.56, "end": 464.5, "text": "或者跟其他模型沟通来来回回多少轮", "chunk": 0, "language": "zh", "avg_logprob": -0.11760819753011068, "no_speech_prob": 2.595616083178598e-11, "compression_ratio": 1.186335403726708 }, { "start": 464.5, "end": 466.22, "text": "这里你用computer use", "chunk": 0, "language": "zh", "avg_logprob": -0.11760819753011068, "no_speech_prob": 2.595616083178598e-11, "compression_ratio": 1.186335403726708 }, { "start": 466.22, "end": 469.16, "text": "用context就直接帮你搞定了", "chunk": 0, "language": "zh", "avg_logprob": -0.11760819753011068, "no_speech_prob": 2.595616083178598e-11, "compression_ratio": 1.186335403726708 }, { "start": 469.16, "end": 470.46, "text": "会非常爽", "chunk": 0, "language": "zh", "avg_logprob": -0.11760819753011068, "no_speech_prob": 2.595616083178598e-11, "compression_ratio": 1.186335403726708 }, { "start": 470.46, "end": 472.36, "text": "给大家快速的看一下这个文档", "chunk": 0, "language": "zh", "avg_logprob": -0.11760819753011068, "no_speech_prob": 2.595616083178598e-11, "compression_ratio": 1.186335403726708 }, { "start": 472.36, "end": 474.58, "text": "首先会讲讲最终的效果", "chunk": 0, "language": "zh", "avg_logprob": -0.11760819753011068, "no_speech_prob": 2.595616083178598e-11, "compression_ratio": 1.186335403726708 }, { "start": 474.58, "end": 475.44, "text": "要达到的目标", "chunk": 0, "language": "zh", "avg_logprob": -0.11760819753011068, "no_speech_prob": 2.595616083178598e-11, "compression_ratio": 1.186335403726708 }, { "start": 475.44, "end": 478.3, "text": "然后Mac mini的前置条件", "chunk": 0, "language": "zh", "avg_logprob": -0.11760819753011068, "no_speech_prob": 2.595616083178598e-11, "compression_ratio": 1.186335403726708 }, { "start": 478.3, "end": 482.4, "text": "然后接着就要安装Cloud Code的桌面端", "chunk": 0, "language": "zh", "avg_logprob": -0.11760819753011068, "no_speech_prob": 2.595616083178598e-11, "compression_ratio": 1.186335403726708 }, { "start": 482.4, "end": 484.58, "text": "然后要安装CC switch", "chunk": 0, "language": "zh", "avg_logprob": -0.11760819753011068, "no_speech_prob": 2.595616083178598e-11, "compression_ratio": 1.186335403726708 }, { "start": 484.58, "end": 487.86, "text": "下面是怎么写入这个Deep Seek的配置", "chunk": 0, "language": "zh", "avg_logprob": -0.11728088783495354, "no_speech_prob": 2.2620107176241078e-11, "compression_ratio": 1.2197802197802199 }, { "start": 487.86, "end": 489.08, "text": "你看", "chunk": 0, "language": "zh", "avg_logprob": -0.11728088783495354, "no_speech_prob": 2.2620107176241078e-11, "compression_ratio": 1.2197802197802199 }, { "start": 489.08, "end": 490.96, "text": "只要把这个文档丢给他", "chunk": 0, "language": "zh", "avg_logprob": -0.11728088783495354, "no_speech_prob": 2.2620107176241078e-11, "compression_ratio": 1.2197802197802199 }, { "start": 490.96, "end": 494.1, "text": "我觉得哪怕你的电脑配置和我的不一样", "chunk": 0, "language": "zh", "avg_logprob": -0.11728088783495354, "no_speech_prob": 2.2620107176241078e-11, "compression_ratio": 1.2197802197802199 }, { "start": 494.1, "end": 496.02, "text": "你也可以快速的完成", "chunk": 0, "language": "zh", "avg_logprob": -0.11728088783495354, "no_speech_prob": 2.2620107176241078e-11, "compression_ratio": 1.2197802197802199 }, { "start": 496.02, "end": 499.4, "text": "即便没有你和GPT沟通出来", "chunk": 0, "language": "zh", "avg_logprob": -0.11728088783495354, "no_speech_prob": 2.2620107176241078e-11, "compression_ratio": 1.2197802197802199 }, { "start": 499.4, "end": 500.5, "text": "也会有", "chunk": 0, "language": "zh", "avg_logprob": -0.11728088783495354, "no_speech_prob": 2.2620107176241078e-11, "compression_ratio": 1.2197802197802199 }, { "start": 500.5, "end": 503.26, "text": "只不过说轮次多一点", "chunk": 0, "language": "zh", "avg_logprob": -0.11728088783495354, "no_speech_prob": 2.2620107176241078e-11, "compression_ratio": 1.2197802197802199 }, { "start": 503.26, "end": 504.28, "text": "卡点多一点", "chunk": 0, "language": "zh", "avg_logprob": -0.11728088783495354, "no_speech_prob": 2.2620107176241078e-11, "compression_ratio": 1.2197802197802199 }, { "start": 504.28, "end": 508.26, "text": "但是都比你那种原来传统的", "chunk": 0, "language": "zh", "avg_logprob": -0.11728088783495354, "no_speech_prob": 2.2620107176241078e-11, "compression_ratio": 1.2197802197802199 }, { "start": 508.26, "end": 511.1, "text": "这种安装方式或者是接入方式", "chunk": 0, "language": "zh", "avg_logprob": -0.11728088783495354, "no_speech_prob": 2.2620107176241078e-11, "compression_ratio": 1.2197802197802199 }, { "start": 511.1, "end": 512.28, "text": "要审视的多", "chunk": 0, "language": "zh", "avg_logprob": -0.11728088783495354, "no_speech_prob": 2.2620107176241078e-11, "compression_ratio": 1.2197802197802199 }, { "start": 512.28, "end": 515.1, "text": "大家看一下我和Codex的沟通界面", "chunk": 0, "language": "zh", "avg_logprob": -0.1080259382724762, "no_speech_prob": 6.85236806532985e-12, "compression_ratio": 0.7884615384615384 }, { "start": 516.05, "end": 524.05, "text": "你看我就给他了这个文档 然后我让他参考我这个文档来给我安装 一开始我让他安装的是", "chunk": 1, "language": "zh", "avg_logprob": -0.1713516827925895, "no_speech_prob": 1.9244435905951818e-11, "compression_ratio": 1.2763157894736843 }, { "start": 524.05, "end": 526.33, "text": "Code Wedding", "chunk": 1, "language": "zh", "avg_logprob": -0.11745822429656982, "no_speech_prob": 3.1309357884090616e-11, "compression_ratio": 1.3854961832061068 }, { "start": 526.33, "end": 527.43, "text": "阿玲的Code Wedding", "chunk": 1, "language": "zh", "avg_logprob": -0.11745822429656982, "no_speech_prob": 3.1309357884090616e-11, "compression_ratio": 1.3854961832061068 }, { "start": 527.43, "end": 529.27, "text": "但是Code Wedding它有个问题", "chunk": 1, "language": "zh", "avg_logprob": -0.11745822429656982, "no_speech_prob": 3.1309357884090616e-11, "compression_ratio": 1.3854961832061068 }, { "start": 529.27, "end": 531.21, "text": "因为它是用来写Code的", "chunk": 1, "language": "zh", "avg_logprob": -0.11745822429656982, "no_speech_prob": 3.1309357884090616e-11, "compression_ratio": 1.3854961832061068 }, { "start": 531.21, "end": 534.65, "text": "所以它是没有实施的搜索能力的", "chunk": 1, "language": "zh", "avg_logprob": -0.11745822429656982, "no_speech_prob": 3.1309357884090616e-11, "compression_ratio": 1.3854961832061068 }, { "start": 534.65, "end": 535.47, "text": "大家可以看到", "chunk": 1, "language": "zh", "avg_logprob": -0.11745822429656982, "no_speech_prob": 3.1309357884090616e-11, "compression_ratio": 1.3854961832061068 }, { "start": 535.47, "end": 539.91, "text": "这个其实就是Code Wedding的本身的回复", "chunk": 1, "language": "zh", "avg_logprob": -0.11745822429656982, "no_speech_prob": 3.1309357884090616e-11, "compression_ratio": 1.3854961832061068 }, { "start": 539.91, "end": 543.43, "text": "你看我问的是特朗普2026年的访华", "chunk": 1, "language": "zh", "avg_logprob": -0.11745822429656982, "no_speech_prob": 3.1309357884090616e-11, "compression_ratio": 1.3854961832061068 }, { "start": 543.43, "end": 548.53, "text": "你看他告诉我他的数据是截至2025年5月的", "chunk": 1, "language": "zh", "avg_logprob": -0.11745822429656982, "no_speech_prob": 3.1309357884090616e-11, "compression_ratio": 1.3854961832061068 }, { "start": 548.53, "end": 550.29, "text": "但是我接入Deep6之后", "chunk": 1, "language": "zh", "avg_logprob": -0.11745822429656982, "no_speech_prob": 3.1309357884090616e-11, "compression_ratio": 1.3854961832061068 }, { "start": 550.29, "end": 553.23, "text": "它其实它是有实施搜索能力的", "chunk": 1, "language": "zh", "avg_logprob": -0.11745822429656982, "no_speech_prob": 3.1309357884090616e-11, "compression_ratio": 1.3854961832061068 }, { "start": 553.23, "end": 558.17, "text": "所以说这个方法就会非常的AI原生化", "chunk": 1, "language": "zh", "avg_logprob": -0.05763951148695618, "no_speech_prob": 2.076183322485825e-11, "compression_ratio": 1.367816091954023 }, { "start": 558.17, "end": 559.99, "text": "我这里只是举了一个例子", "chunk": 1, "language": "zh", "avg_logprob": -0.05763951148695618, "no_speech_prob": 2.076183322485825e-11, "compression_ratio": 1.367816091954023 }, { "start": 559.99, "end": 563.45, "text": "实际上大家在日常的工作生活中", "chunk": 1, "language": "zh", "avg_logprob": -0.05763951148695618, "no_speech_prob": 2.076183322485825e-11, "compression_ratio": 1.367816091954023 }, { "start": 563.45, "end": 568.29, "text": "应该想办法把自己的所有的工作的模块", "chunk": 1, "language": "zh", "avg_logprob": -0.05763951148695618, "no_speech_prob": 2.076183322485825e-11, "compression_ratio": 1.367816091954023 }, { "start": 568.29, "end": 570.77, "text": "都尽可能的AI原生化", "chunk": 1, "language": "zh", "avg_logprob": -0.05763951148695618, "no_speech_prob": 2.076183322485825e-11, "compression_ratio": 1.367816091954023 }, { "start": 570.77, "end": 573.69, "text": "这个AI原生是说是你一个互相交互的东西", "chunk": 1, "language": "zh", "avg_logprob": -0.05763951148695618, "no_speech_prob": 2.076183322485825e-11, "compression_ratio": 1.367816091954023 }, { "start": 573.69, "end": 576.17, "text": "你既要把你知道的去交互给他", "chunk": 1, "language": "zh", "avg_logprob": -0.05763951148695618, "no_speech_prob": 2.076183322485825e-11, "compression_ratio": 1.367816091954023 }, { "start": 576.17, "end": 577.49, "text": "让他知道你的状态", "chunk": 1, "language": "zh", "avg_logprob": -0.05763951148695618, "no_speech_prob": 2.076183322485825e-11, "compression_ratio": 1.367816091954023 }, { "start": 577.49, "end": 581.73, "text": "也要尽量避免自己直接输入东西", "chunk": 1, "language": "zh", "avg_logprob": -0.05763951148695618, "no_speech_prob": 2.076183322485825e-11, "compression_ratio": 1.367816091954023 }, { "start": 581.73, "end": 584.37, "text": "说错了直接输出东西", "chunk": 1, "language": "zh", "avg_logprob": -0.09577601750691732, "no_speech_prob": 1.368619974295715e-11, "compression_ratio": 1.3060344827586208 }, { "start": 584.37, "end": 588.73, "text": "就是你要尽量让输出是AI给你的", "chunk": 1, "language": "zh", "avg_logprob": -0.09577601750691732, "no_speech_prob": 1.368619974295715e-11, "compression_ratio": 1.3060344827586208 }, { "start": 588.73, "end": 590.79, "text": "然后这个AI给你的东西", "chunk": 1, "language": "zh", "avg_logprob": -0.09577601750691732, "no_speech_prob": 1.368619974295715e-11, "compression_ratio": 1.3060344827586208 }, { "start": 590.79, "end": 593.61, "text": "你再交给这个另一个AI", "chunk": 1, "language": "zh", "avg_logprob": -0.09577601750691732, "no_speech_prob": 1.368619974295715e-11, "compression_ratio": 1.3060344827586208 }, { "start": 593.61, "end": 597.57, "text": "AI之间他们互相的这个可理解性比你更强", "chunk": 1, "language": "zh", "avg_logprob": -0.09577601750691732, "no_speech_prob": 1.368619974295715e-11, "compression_ratio": 1.3060344827586208 }, { "start": 597.57, "end": 600.17, "text": "长期这样的培养之后", "chunk": 1, "language": "zh", "avg_logprob": -0.09577601750691732, "no_speech_prob": 1.368619974295715e-11, "compression_ratio": 1.3060344827586208 }, { "start": 600.17, "end": 603.45, "text": "你的AI的思维会很强", "chunk": 1, "language": "zh", "avg_logprob": -0.09577601750691732, "no_speech_prob": 1.368619974295715e-11, "compression_ratio": 1.3060344827586208 }, { "start": 603.45, "end": 605.99, "text": "你会更能和AI对话", "chunk": 1, "language": "zh", "avg_logprob": -0.09577601750691732, "no_speech_prob": 1.368619974295715e-11, "compression_ratio": 1.3060344827586208 }, { "start": 605.99, "end": 608.57, "text": "你看这里对话就一轮", "chunk": 1, "language": "zh", "avg_logprob": -0.09577601750691732, "no_speech_prob": 1.368619974295715e-11, "compression_ratio": 1.3060344827586208 }, { "start": 608.57, "end": 610.07, "text": "我给他那个文档", "chunk": 1, "language": "zh", "avg_logprob": -0.09577601750691732, "no_speech_prob": 1.368619974295715e-11, "compression_ratio": 1.3060344827586208 }, { "start": 610.07, "end": 613.91, "text": "后面我要做的就是给他我阿里云的coding plan的API key", "chunk": 1, "language": "zh", "avg_logprob": -0.1633356185186477, "no_speech_prob": 1.950553607521499e-11, "compression_ratio": 1.411371237458194 }, { "start": 614.17, "end": 615.45, "text": "API key这里要打码", "chunk": 1, "language": "zh", "avg_logprob": -0.1633356185186477, "no_speech_prob": 1.950553607521499e-11, "compression_ratio": 1.411371237458194 }, { "start": 615.71, "end": 618.77, "text": "后面他这里就告诉我我的阿里云的", "chunk": 1, "language": "zh", "avg_logprob": -0.1633356185186477, "no_speech_prob": 1.950553607521499e-11, "compression_ratio": 1.411371237458194 }, { "start": 619.03, "end": 622.61, "text": "Cloud的coding plan已经接入成功了这里告诉我了", "chunk": 1, "language": "zh", "avg_logprob": -0.1633356185186477, "no_speech_prob": 1.950553607521499e-11, "compression_ratio": 1.411371237458194 }, { "start": 622.87, "end": 623.39, "text": "然后呢", "chunk": 1, "language": "zh", "avg_logprob": -0.1633356185186477, "no_speech_prob": 1.950553607521499e-11, "compression_ratio": 1.411371237458194 }, { "start": 623.63, "end": 627.23, "text": "但是当时我直接用的时候我发现他会有一些小bug", "chunk": 1, "language": "zh", "avg_logprob": -0.1633356185186477, "no_speech_prob": 1.950553607521499e-11, "compression_ratio": 1.411371237458194 }, { "start": 627.47, "end": 628.51, "text": "比如说他这里的", "chunk": 1, "language": "zh", "avg_logprob": -0.1633356185186477, "no_speech_prob": 1.950553607521499e-11, "compression_ratio": 1.411371237458194 }, { "start": 628.75, "end": 629.79, "text": "这个名字", "chunk": 1, "language": "zh", "avg_logprob": -0.1633356185186477, "no_speech_prob": 1.950553607521499e-11, "compression_ratio": 1.411371237458194 }, { "start": 630.03, "end": 631.07, "text": "这里是没有改的", "chunk": 1, "language": "zh", "avg_logprob": -0.1633356185186477, "no_speech_prob": 1.950553607521499e-11, "compression_ratio": 1.411371237458194 }, { "start": 631.31, "end": 633.11, "text": "然后呢我这个反馈给他", "chunk": 1, "language": "zh", "avg_logprob": -0.1633356185186477, "no_speech_prob": 1.950553607521499e-11, "compression_ratio": 1.411371237458194 }, { "start": 633.37, "end": 636.43, "text": "同时我就直接启动了这个computer use", "chunk": 1, "language": "zh", "avg_logprob": -0.1633356185186477, "no_speech_prob": 1.950553607521499e-11, "compression_ratio": 1.411371237458194 }, { "start": 636.69, "end": 639.25, "text": "他就直接帮我选项接入", "chunk": 1, "language": "zh", "avg_logprob": -0.1633356185186477, "no_speech_prob": 1.950553607521499e-11, "compression_ratio": 1.411371237458194 }, { "start": 639.25, "end": 642.55, "text": "你要知道如果说没有context这样的辅助", "chunk": 1, "language": "zh", "avg_logprob": -0.1665173748083282, "no_speech_prob": 1.847223762840855e-11, "compression_ratio": 1.1943462897526502 }, { "start": 642.55, "end": 646.47, "text": "你走到这个接入的界面有多难", "chunk": 1, "language": "zh", "avg_logprob": -0.1665173748083282, "no_speech_prob": 1.847223762840855e-11, "compression_ratio": 1.1943462897526502 }, { "start": 646.47, "end": 649.73, "text": "有可能一个卡点就能卡你非常久", "chunk": 1, "language": "zh", "avg_logprob": -0.1665173748083282, "no_speech_prob": 1.847223762840855e-11, "compression_ratio": 1.1943462897526502 }, { "start": 649.73, "end": 653.61, "text": "大家现在看到的界面就是computer use", "chunk": 1, "language": "zh", "avg_logprob": -0.1665173748083282, "no_speech_prob": 1.847223762840855e-11, "compression_ratio": 1.1943462897526502 }, { "start": 653.61, "end": 660.53, "text": "通过cc switch的设置来接入cloud的桌面端的整个过程", "chunk": 1, "language": "zh", "avg_logprob": -0.1665173748083282, "no_speech_prob": 1.847223762840855e-11, "compression_ratio": 1.1943462897526502 }, { "start": 660.53, "end": 664.59, "text": "然后很快的阿里云的coding plan我接入完成之后", "chunk": 1, "language": "zh", "avg_logprob": -0.1665173748083282, "no_speech_prob": 1.847223762840855e-11, "compression_ratio": 1.1943462897526502 }, { "start": 664.59, "end": 667.33, "text": "我接着给他的是deep seek的apikey", "chunk": 1, "language": "zh", "avg_logprob": -0.1665173748083282, "no_speech_prob": 1.847223762840855e-11, "compression_ratio": 1.1943462897526502 }, { "start": 667.33, "end": 669.01, "text": "API Key这里还是要打一下码", "chunk": 1, "language": "zh", "avg_logprob": -0.08959874822132623, "no_speech_prob": 3.0206979279023827e-11, "compression_ratio": 1.2372881355932204 }, { "start": 669.01, "end": 672.29, "text": "然后Deep Seek的V4的模型也介入成功了", "chunk": 1, "language": "zh", "avg_logprob": -0.08959874822132623, "no_speech_prob": 3.0206979279023827e-11, "compression_ratio": 1.2372881355932204 }, { "start": 672.29, "end": 673.27, "text": "特别爽", "chunk": 1, "language": "zh", "avg_logprob": -0.08959874822132623, "no_speech_prob": 3.0206979279023827e-11, "compression_ratio": 1.2372881355932204 }, { "start": 673.27, "end": 675.85, "text": "我觉得以后大家在处理问题的时候", "chunk": 1, "language": "zh", "avg_logprob": -0.08959874822132623, "no_speech_prob": 3.0206979279023827e-11, "compression_ratio": 1.2372881355932204 }, { "start": 675.85, "end": 679.13, "text": "都要保持这样的AI原生的思路和思维", "chunk": 1, "language": "zh", "avg_logprob": -0.08959874822132623, "no_speech_prob": 3.0206979279023827e-11, "compression_ratio": 1.2372881355932204 }, { "start": 679.13, "end": 683.95, "text": "然后我再把我这一轮的动作再抽象一下", "chunk": 1, "language": "zh", "avg_logprob": -0.08959874822132623, "no_speech_prob": 3.0206979279023827e-11, "compression_ratio": 1.2372881355932204 }, { "start": 683.95, "end": 687.29, "text": "这样子我就形成了一个新的指南", "chunk": 1, "language": "zh", "avg_logprob": -0.08959874822132623, "no_speech_prob": 3.0206979279023827e-11, "compression_ratio": 1.2372881355932204 }, { "start": 687.29, "end": 689.67, "text": "就是把前一版的文档", "chunk": 1, "language": "zh", "avg_logprob": -0.08959874822132623, "no_speech_prob": 3.0206979279023827e-11, "compression_ratio": 1.2372881355932204 }, { "start": 689.67, "end": 692.65, "text": "没有抽取到没有覆盖到的东西", "chunk": 1, "language": "zh", "avg_logprob": -0.08959874822132623, "no_speech_prob": 3.0206979279023827e-11, "compression_ratio": 1.2372881355932204 }, { "start": 692.65, "end": 694.71, "text": "我再把它抽取出来", "chunk": 1, "language": "zh", "avg_logprob": -0.08959874822132623, "no_speech_prob": 3.0206979279023827e-11, "compression_ratio": 1.2372881355932204 }, { "start": 694.71, "end": 697.37, "text": "形成了一个更加完整的手册", "chunk": 1, "language": "zh", "avg_logprob": -0.0850427087984587, "no_speech_prob": 1.4122724691090216e-11, "compression_ratio": 1.3206896551724139 }, { "start": 697.37, "end": 699.71, "text": "你看它的清晰度就会更高了", "chunk": 1, "language": "zh", "avg_logprob": -0.0850427087984587, "no_speech_prob": 1.4122724691090216e-11, "compression_ratio": 1.3206896551724139 }, { "start": 699.71, "end": 702.33, "text": "而这一版的MD的文档", "chunk": 1, "language": "zh", "avg_logprob": -0.0850427087984587, "no_speech_prob": 1.4122724691090216e-11, "compression_ratio": 1.3206896551724139 }, { "start": 702.33, "end": 705.61, "text": "在交接给下一个人来安装", "chunk": 1, "language": "zh", "avg_logprob": -0.0850427087984587, "no_speech_prob": 1.4122724691090216e-11, "compression_ratio": 1.3206896551724139 }, { "start": 705.61, "end": 707.79, "text": "国内模型的接入的时候", "chunk": 1, "language": "zh", "avg_logprob": -0.0850427087984587, "no_speech_prob": 1.4122724691090216e-11, "compression_ratio": 1.3206896551724139 }, { "start": 707.79, "end": 710.75, "text": "它中间遇到的卡点会更少", "chunk": 1, "language": "zh", "avg_logprob": -0.0850427087984587, "no_speech_prob": 1.4122724691090216e-11, "compression_ratio": 1.3206896551724139 }, { "start": 710.75, "end": 713.73, "text": "这个就是不断迭代的一个过程", "chunk": 1, "language": "zh", "avg_logprob": -0.0850427087984587, "no_speech_prob": 1.4122724691090216e-11, "compression_ratio": 1.3206896551724139 }, { "start": 713.73, "end": 715.59, "text": "怎么样我介绍的这个方法", "chunk": 1, "language": "zh", "avg_logprob": -0.0850427087984587, "no_speech_prob": 1.4122724691090216e-11, "compression_ratio": 1.3206896551724139 }, { "start": 715.59, "end": 717.65, "text": "是不是特别的轻量级", "chunk": 1, "language": "zh", "avg_logprob": -0.0850427087984587, "no_speech_prob": 1.4122724691090216e-11, "compression_ratio": 1.3206896551724139 }, { "start": 717.65, "end": 719.51, "text": "相关的文档可以到我的", "chunk": 1, "language": "zh", "avg_logprob": -0.0850427087984587, "no_speech_prob": 1.4122724691090216e-11, "compression_ratio": 1.3206896551724139 }, { "start": 719.51, "end": 721.29, "text": "同名公众号里面去找", "chunk": 1, "language": "zh", "avg_logprob": -0.0850427087984587, "no_speech_prob": 1.4122724691090216e-11, "compression_ratio": 1.3206896551724139 }, { "start": 721.29, "end": 724.03, "text": "而且我相信这种桌面终端", "chunk": 1, "language": "zh", "avg_logprob": -0.0850427087984587, "no_speech_prob": 1.4122724691090216e-11, "compression_ratio": 1.3206896551724139 }, { "start": 724.03, "end": 726.91, "text": "也是大部分人都能够接受", "chunk": 1, "language": "zh", "avg_logprob": -0.13171128990236391, "no_speech_prob": 3.29256795439381e-11, "compression_ratio": 1.0920502092050208 }, { "start": 726.91, "end": 729.25, "text": "并且快速上手的一种方式", "chunk": 1, "language": "zh", "avg_logprob": -0.13171128990236391, "no_speech_prob": 3.29256795439381e-11, "compression_ratio": 1.0920502092050208 }, { "start": 729.25, "end": 731.35, "text": "国内的模型确实便宜", "chunk": 1, "language": "zh", "avg_logprob": -0.13171128990236391, "no_speech_prob": 3.29256795439381e-11, "compression_ratio": 1.0920502092050208 }, { "start": 731.35, "end": 732.85, "text": "我调用了好多轮", "chunk": 1, "language": "zh", "avg_logprob": -0.13171128990236391, "no_speech_prob": 3.29256795439381e-11, "compression_ratio": 1.0920502092050208 }, { "start": 732.85, "end": 733.93, "text": "用了半天", "chunk": 1, "language": "zh", "avg_logprob": -0.13171128990236391, "no_speech_prob": 3.29256795439381e-11, "compression_ratio": 1.0920502092050208 }, { "start": 733.93, "end": 736.47, "text": "Deepseek上面API接口才花了10块钱", "chunk": 1, "language": "zh", "avg_logprob": -0.13171128990236391, "no_speech_prob": 3.29256795439381e-11, "compression_ratio": 1.0920502092050208 }, { "start": 736.47, "end": 737.63, "text": "赶紧去试试吧", "chunk": 1, "language": "zh", "avg_logprob": -0.13171128990236391, "no_speech_prob": 3.29256795439381e-11, "compression_ratio": 1.0920502092050208 }, { "start": 737.63, "end": 739.47, "text": "记得订阅林姐说AI", "chunk": 1, "language": "zh", "avg_logprob": -0.13171128990236391, "no_speech_prob": 3.29256795439381e-11, "compression_ratio": 1.0920502092050208 }, { "start": 739.47, "end": 741.79, "text": "给我的节目点赞评论", "chunk": 1, "language": "zh", "avg_logprob": -0.13171128990236391, "no_speech_prob": 3.29256795439381e-11, "compression_ratio": 1.0920502092050208 }, { "start": 741.79, "end": 743.53, "text": "我们下期再见了", "chunk": 1, "language": "zh", "avg_logprob": -0.13171128990236391, "no_speech_prob": 3.29256795439381e-11, "compression_ratio": 1.0920502092050208 }, { "start": 743.53, "end": 743.89, "text": "拜拜", "chunk": 1, "language": "zh", "avg_logprob": -0.13171128990236391, "no_speech_prob": 3.29256795439381e-11, "compression_ratio": 1.0920502092050208 } ], "segmented_transcription": { "status": "SEGMENTED_TRANSCRIPT_OK", "created_at": "2026-07-28T10:48:58", "manifest": "/Users/sagawa/AI_WORK/video_notes/url/20260728_104832__Claude_Code_竟然能接_DeepSeek_我用_Codex_跑通了_CC_Switch/segmented_work/chunk_manifest.json", "audio": "/Users/sagawa/AI_WORK/video_notes/url/20260728_104832__Claude_Code_竟然能接_DeepSeek_我用_Codex_跑通了_CC_Switch/audio_16k_mono.wav", "audio_sha256": "847f85aed79db6ed1045cb95179e87a206ffe0093e192bf4f54b640f4e4c8ae9", "audio_duration": 744.455, "model": "mlx-community/whisper-large-v3-turbo", "language_hint": "zh", "condition_on_previous_text": false, "chunk_count": 2, "resumed_chunks": 0, "merged_segment_count": 275, "final_segment_end": 743.89, "tail_repeat_run": 1, "tail_repeat_text": "我觉得以后大家在处理问题的时候", "qc_ok": true, "qc_errors": [] } }