{"text": "我們經常看到網上有很多的AI工作流,自全自動化工作流,7x24小時工作流。但是當我做不到的時候,我就會疑問說,他們是真的嗎?他們真的可以撒手不管,然後讓agent一直在做決策,在做運行嗎?當我真正自己實現了之後,我才發現這裡面的細節非常的多。那今天,我就帶大家親自拆解一條我的視頻製作工作流。現在看到的所有視頻,這個,這個,以及這個,都是這個工作流剪輯出來的。好,話不多說,那我們直接開始拆解。整個視頻製作分為六步,用痛點挖掘需求提取。根據自己的insights,形成觀點提取框架,要最後生成腳本。錄製視頻,剪輯視頻,製作,封面,以及最後發布全平台。但是我們並沒有用一個prompt把它交給agent,讓它全全去做。而是用這樣一個framework,兩個維度,把它所有的任務劃在四個項線裡面。而每一個項線,我會用不同的方式去做。那對於這種共同探索,或者一次性委託的,我們就不多追溯。基本上是通過和AI之間的聊天,或者是一次性的prompt完成。我們今天主要focus在流程化的協作和自動化執行這兩個項線裡面。比如說第一個,用戶痛點調研,以及全平台發布。這兩個重複性又非常高,對我的每次的判斷力又依賴很低。這種類型的任務,你就可以設置一個全自動化的工作流。對於用戶調研需求,我用的是Growbuck Skills,加上Notion的MCP,再加上Growbuck裡面的Routine,來實現這樣一份全自動化工作流。所以初始的設置是這樣的,你要寫一份完整的Skills,需要用戶調研的Skills,這裡包含你的內容要求,你的衡量標準,以及輸出的格式。然後同時給他開一些MCP Tools,比如說,我想看報告是在Notion裡面,用Notion Template。所以你給他連接到Notion的MCP,並且告訴他用人生的Block Template去使用。然後最後給他設置一個Routine,Routine就是讓他主動的,按時的,或者周期性的去做一件事情。那這個Routine我就用它,讓他自動的來去掃描他所生成的Notion的Research報告,的Comments。如果有新的Comments,他就會去Address,Spone一個新的Session或者Task來去做。所以他是通過這三個功能來實現全自動化,而我和Aden之間的協作是通過Notion。那接下來我們來看,流程化協作的這個象限。就是對你的判斷性依賴又很高,同時重複性又很高的這種工作,我們應該怎麼做?首先我們要選擇一個基礎能力,就是需要有一個工具或者是引擎,它可以去自動編輯實品。那這裡我選擇的是Hyperfree,如果大家需要,我可以再出一期單獨講解為什麼我會選擇Hyperfree,以及它的一些Tactics。所以你把第一步引擎選定了之後,工具選定了之後,你要先去試用它,讓它就用它原生的這個引擎,它自帶的一些Skills,你直接直出一板效果。結果你會發現出出來的效果,可能完全不是你想要的樣子。舉個具體的例子,我的視頻裡面有大量的屏幕錄製。當我講到界面上的某個功能的時候,我希望觀眾能立刻知道現在應該看哪裡。而我想要的效果通常是這樣的,框出重點區域,把其他部分壓按,再把重點區域抬出來,加上一個簡短的標籤,這樣觀眾的注意力就能跟著我的講解來走。那這套表達本身不複雜,但如果每次都做,我每次都要重新解釋一遍。而AI每次的執行,它都不一樣。所以我希望把這套處理方式,沉澱成一個可用的一個Recipe,再加上它實現效果的Block。這樣,Recipe告訴我什麼時候用,該怎麼用。而Block呢,直接把框選壓按,抬起,標籤,這一系列的動作,給它實現起來。那模型負責什麼?模型就負責去理解這一次的扣播腳本。結合我的講解和路評,它需要判斷,這段是不是該出,局部交,聚焦,應該出在哪個區域,什麼時候出現,標籤寫在哪裡。所以這時候你大概已經了解了,什麼時候該選擇Harness。Harness不是一個簡單的,淨它的文件,Markdown文件,而是它是一個動態的系統,這裡面包含了Recipe,包含規則,裝的一些動態效果,兩標準,同時它還包含了自我演化的一個功能。當這次新的視頻裡面出現一些新的Pattern,它就會再沉澱到你的這個Harness裡面。所以你的Harness最后變成了Cumulate的你的一個資產庫。如果大家很感興趣,這部分到底來怎麼來做,我也可以把我的VideoFactory,Distill一下,再Open Source出來,這樣可以大家一起來使用。那最後你可能會問,整個這個製作流程分了這麼多步,那每一步之間的協作,是不是還得你來Drive呀?不是的,我跟它之間的協作,就是用Notion來協作。整個協作非常簡單,我錄製好視頻,把內容Title沾添到Notion裡面。Agent會定時的去掃描,如果發現內容的Title已經有了,它會調用Descript MCP去直接下載我的錄取視頻,加上Transcription。然後它就可以自動的kick off整個剪輯的過程。當它完成之後,它會GrabBot告訴我,然後進行Review。當我Review通過之後,這個時候是通過聊天,會自動的通知其他的Agent來進行浪施到其他的平台。所以整個過程就是Routine,加上Agent和Agent之間自帶的Communication,這裡我使用的是GrabBot。同時,Skills,再加上各式各樣的MCP,你用的Notion MCP,Descript MCP等等,就把整個的過程串起來。那最後,我們WrapUp,AI整個工作流就是在不斷的區分你任務的屬性,以及你和Agent之間的分工,而剩下的都是可以通過各式各樣的Routine加上Tools,可以把它自動化流程下來。那好,今天我們就分享到這兒,下次再見,我是Gin。", "segments": [{"id": 0, "seek": 0, "start": 0.0, "end": 6.0, "text": "我們經常看到網上有很多的AI工作流,自全自動化工作流,7x24小時工作流。", "tokens": [50365, 5884, 15145, 11279, 18032, 26352, 5708, 2412, 20778, 1546, 48698, 41315, 27854, 11, 9722, 11319, 9722, 12572, 23756, 41315, 27854, 11, 22, 87, 7911, 46997, 41315, 27854, 1543, 50665], "temperature": 0, "avg_logprob": -0.16408782958984375, "compression_ratio": 1.289902280130293, "no_speech_prob": 9.042384029045447e-12}, {"id": 1, "seek": 0, "start": 6.0, "end": 15.0, "text": "但是當我做不到的時候,我就會疑問說,他們是真的嗎?他們真的可以撒手不管,然後讓agent一直在做決策,在做運行嗎?", "tokens": [50665, 11189, 13118, 1654, 10907, 35728, 20643, 11, 22020, 6236, 18140, 239, 11361, 4622, 11, 20486, 1541, 8034, 7434, 30, 20486, 8034, 6723, 20559, 240, 11389, 45376, 11, 10213, 21195, 559, 317, 34448, 3581, 10907, 33540, 7973, 244, 11, 3581, 10907, 32772, 8082, 7434, 30, 51115], "temperature": 0, "avg_logprob": -0.16408782958984375, "compression_ratio": 1.289902280130293, "no_speech_prob": 9.042384029045447e-12}, {"id": 2, "seek": 0, "start": 15.0, "end": 20.0, "text": "當我真正自己實現了之後,我才發現這裡面的細節非常的多。", "tokens": [51115, 13118, 1654, 6303, 15789, 17645, 10376, 9581, 2289, 26362, 11, 1654, 18888, 38927, 18771, 8833, 1546, 41588, 27694, 48263, 6392, 1543, 51365], "temperature": 0, "avg_logprob": -0.16408782958984375, "compression_ratio": 1.289902280130293, "no_speech_prob": 9.042384029045447e-12}, {"id": 3, "seek": 0, "start": 20.0, "end": 25.0, "text": "那今天,我就帶大家親自拆解一條我的視頻製作工作流。", "tokens": [51365, 4184, 12074, 11, 22020, 32379, 6868, 42117, 9722, 6852, 228, 17278, 2257, 31106, 14200, 27333, 29940, 8083, 121, 11914, 41315, 27854, 1543, 51615], "temperature": 0, "avg_logprob": -0.16408782958984375, "compression_ratio": 1.289902280130293, "no_speech_prob": 9.042384029045447e-12}, {"id": 4, "seek": 2500, "start": 25.0, "end": 32.0, "text": "現在看到的所有視頻,這個,這個,以及這個,都是這個工作流剪輯出來的。", "tokens": [50365, 12648, 18032, 1546, 39300, 27333, 29940, 11, 6287, 11, 6287, 11, 40282, 6287, 11, 22796, 6287, 41315, 27854, 5935, 103, 12290, 107, 29741, 1546, 1543, 50715], "temperature": 0, "avg_logprob": -0.15335415319069146, "compression_ratio": 1.1802575107296138, "no_speech_prob": 1.0789724148863034e-11}, {"id": 5, "seek": 2500, "start": 32.0, "end": 34.0, "text": "好,話不多說,那我們直接開始拆解。", "tokens": [50715, 2131, 11, 11103, 33364, 4622, 11, 46714, 43297, 21017, 6852, 228, 17278, 1543, 50815], "temperature": 0, "avg_logprob": -0.15335415319069146, "compression_ratio": 1.1802575107296138, "no_speech_prob": 1.0789724148863034e-11}, {"id": 6, "seek": 2500, "start": 41.0, "end": 45.0, "text": "整個視頻製作分為六步,用痛點挖掘需求提取。", "tokens": [51165, 27662, 3338, 27333, 29940, 8083, 121, 11914, 6627, 6344, 30566, 31429, 11, 9254, 28737, 8216, 8501, 244, 6900, 246, 32535, 32718, 20949, 29436, 1543, 51365], "temperature": 0, "avg_logprob": -0.15335415319069146, "compression_ratio": 1.1802575107296138, "no_speech_prob": 1.0789724148863034e-11}, {"id": 7, "seek": 2500, "start": 45.0, "end": 50.0, "text": "根據自己的insights,形成觀點提取框架,要最後生成腳本。", "tokens": [51365, 31337, 36841, 17645, 1546, 1292, 5761, 11, 30900, 11336, 25179, 8216, 20949, 29436, 21416, 228, 7360, 114, 11, 4275, 20578, 8244, 11336, 49595, 8802, 1543, 51615], "temperature": 0, "avg_logprob": -0.15335415319069146, "compression_ratio": 1.1802575107296138, "no_speech_prob": 1.0789724148863034e-11}, {"id": 8, "seek": 5000, "start": 50.0, "end": 57.0, "text": "錄製視頻,剪輯視頻,製作,封面,以及最後發布全平台。", "tokens": [50365, 9567, 226, 8083, 121, 27333, 29940, 11, 5935, 103, 12290, 107, 27333, 29940, 11, 8083, 121, 11914, 11, 1530, 223, 8833, 11, 40282, 20578, 14637, 34688, 11319, 16716, 15433, 1543, 50715], "temperature": 0, "avg_logprob": -0.09704830676694459, "compression_ratio": 1.3125, "no_speech_prob": 1.7467761612155464e-11}, {"id": 9, "seek": 5000, "start": 57.0, "end": 61.0, "text": "但是我們並沒有用一個prompt把它交給agent,讓它全全去做。", "tokens": [50715, 11189, 5884, 35934, 6963, 9254, 8990, 28722, 662, 42061, 28455, 17798, 559, 317, 11, 21195, 11284, 11319, 11319, 6734, 10907, 1543, 50915], "temperature": 0, "avg_logprob": -0.09704830676694459, "compression_ratio": 1.3125, "no_speech_prob": 1.7467761612155464e-11}, {"id": 10, "seek": 5000, "start": 61.0, "end": 67.0, "text": "而是用這樣一個framework,兩個維度,把它所有的任務劃在四個項線裡面。", "tokens": [50915, 11070, 1541, 9254, 8377, 8990, 17265, 1902, 11, 34623, 9261, 255, 13127, 11, 42061, 39300, 1546, 26443, 39517, 5087, 225, 3581, 19425, 3338, 8313, 227, 31699, 32399, 1543, 51215], "temperature": 0, "avg_logprob": -0.09704830676694459, "compression_ratio": 1.3125, "no_speech_prob": 1.7467761612155464e-11}, {"id": 11, "seek": 5000, "start": 67.0, "end": 70.0, "text": "而每一個項線,我會用不同的方式去做。", "tokens": [51215, 11070, 23664, 8990, 8313, 227, 31699, 11, 1654, 6236, 9254, 47123, 1546, 9249, 27584, 6734, 10907, 1543, 51365], "temperature": 0, "avg_logprob": -0.09704830676694459, "compression_ratio": 1.3125, "no_speech_prob": 1.7467761612155464e-11}, {"id": 12, "seek": 5000, "start": 70.0, "end": 74.0, "text": "那對於這種共同探索,或者一次性委託的,我們就不多追溯。", "tokens": [51365, 4184, 2855, 19488, 33005, 15408, 13089, 6900, 95, 7732, 95, 11, 31148, 27505, 21686, 44230, 5396, 245, 1546, 11, 5884, 3111, 33364, 39947, 16084, 107, 1543, 51565], "temperature": 0, "avg_logprob": -0.09704830676694459, "compression_ratio": 1.3125, "no_speech_prob": 1.7467761612155464e-11}, {"id": 13, "seek": 5000, "start": 74.0, "end": 78.0, "text": "基本上是通過和AI之間的聊天,或者是一次性的prompt完成。", "tokens": [51565, 37946, 5708, 1541, 19550, 8816, 12565, 48698, 9574, 11016, 1546, 40096, 6135, 11, 31148, 1541, 27505, 21686, 1546, 28722, 662, 41509, 1543, 51765], "temperature": 0, "avg_logprob": -0.09704830676694459, "compression_ratio": 1.3125, "no_speech_prob": 1.7467761612155464e-11}, {"id": 14, "seek": 7800, "start": 78.0, "end": 85.0, "text": "我們今天主要focus在流程化的協作和自動化執行這兩個項線裡面。", "tokens": [50365, 5884, 12074, 13557, 4275, 69, 15206, 3581, 27854, 29649, 23756, 1546, 5322, 242, 11914, 12565, 9722, 12572, 23756, 16262, 115, 8082, 2664, 34623, 8313, 227, 31699, 32399, 1543, 50715], "temperature": 0, "avg_logprob": -0.13767604827880858, "compression_ratio": 1.0821256038647342, "no_speech_prob": 1.559898737957255e-11}, {"id": 15, "seek": 7800, "start": 85.0, "end": 95.0, "text": "比如說第一個,用戶痛點調研,以及全平台發布。", "tokens": [50715, 36757, 4622, 40760, 11, 9254, 1486, 114, 28737, 8216, 32943, 23230, 242, 11, 40282, 11319, 16716, 15433, 14637, 34688, 1543, 51215], "temperature": 0, "avg_logprob": -0.13767604827880858, "compression_ratio": 1.0821256038647342, "no_speech_prob": 1.559898737957255e-11}, {"id": 16, "seek": 7800, "start": 95.0, "end": 100.0, "text": "這兩個重複性又非常高,對我的每次的判斷力又依賴很低。", "tokens": [51215, 2664, 34623, 12624, 164, 97, 229, 21686, 17047, 14392, 12979, 11, 2855, 14200, 23664, 9487, 1546, 2437, 97, 4307, 115, 13486, 17047, 3254, 251, 164, 7062, 4563, 41377, 1543, 51465], "temperature": 0, "avg_logprob": -0.13767604827880858, "compression_ratio": 1.0821256038647342, "no_speech_prob": 1.559898737957255e-11}, {"id": 17, "seek": 10000, "start": 100.0, "end": 104.0, "text": "這種類型的任務,你就可以設置一個全自動化的工作流。", "tokens": [50365, 33005, 40092, 39823, 1546, 26443, 39517, 11, 2166, 36781, 39035, 34719, 8990, 11319, 9722, 12572, 23756, 1546, 41315, 27854, 1543, 50565], "temperature": 0, "avg_logprob": -0.1685164891756498, "compression_ratio": 1.1942148760330578, "no_speech_prob": 1.4342153333013474e-11}, {"id": 18, "seek": 10000, "start": 104.0, "end": 109.0, "text": "對於用戶調研需求,我用的是Growbuck Skills,加上Notion的MCP,", "tokens": [50565, 2855, 19488, 9254, 1486, 114, 32943, 23230, 242, 32535, 32718, 11, 1654, 9254, 24620, 38, 1892, 65, 1134, 27856, 11, 9990, 5708, 15208, 313, 1546, 44, 20049, 11, 50815], "temperature": 0, "avg_logprob": -0.1685164891756498, "compression_ratio": 1.1942148760330578, "no_speech_prob": 1.4342153333013474e-11}, {"id": 19, "seek": 10000, "start": 109.0, "end": 114.0, "text": "再加上Growbuck裡面的Routine,來實現這樣一份全自動化工作流。", "tokens": [50815, 8623, 9990, 5708, 38, 1892, 65, 1134, 32399, 1546, 49, 45075, 11, 3763, 10376, 9581, 8377, 2257, 36266, 11319, 9722, 12572, 23756, 41315, 27854, 1543, 51065], "temperature": 0, "avg_logprob": -0.1685164891756498, "compression_ratio": 1.1942148760330578, "no_speech_prob": 1.4342153333013474e-11}, {"id": 20, "seek": 10000, "start": 114.0, "end": 118.0, "text": "所以初始的設置是這樣的,你要寫一份完整的Skills,", "tokens": [51065, 7239, 28727, 15476, 1546, 39035, 34719, 1541, 39425, 11, 32085, 4510, 104, 2257, 36266, 14128, 27662, 1546, 50, 74, 2565, 11, 51265], "temperature": 0, "avg_logprob": -0.1685164891756498, "compression_ratio": 1.1942148760330578, "no_speech_prob": 1.4342153333013474e-11}, {"id": 21, "seek": 11800, "start": 118.0, "end": 124.0, "text": "需要用戶調研的Skills,這裡包含你的內容要求,你的衡量標準,以及輸出的格式。", "tokens": [50365, 35748, 9254, 1486, 114, 32943, 23230, 242, 1546, 50, 74, 2565, 11, 18771, 23305, 2392, 104, 18961, 28472, 25750, 4275, 32718, 11, 18961, 9890, 94, 26748, 46299, 28220, 11, 40282, 12290, 116, 7781, 1546, 30921, 27584, 1543, 50665], "temperature": 0, "avg_logprob": -0.1589977191044734, "compression_ratio": 1.1395348837209303, "no_speech_prob": 1.7306597127619838e-11}, {"id": 22, "seek": 11800, "start": 124.0, "end": 131.0, "text": "然後同時給他開一些MCP Tools,比如說,我想看報告是在Notion裡面,用Notion Template。", "tokens": [50665, 10213, 13089, 6611, 17798, 5000, 8949, 38515, 44, 20049, 30302, 11, 36757, 4622, 11, 25246, 4200, 17803, 16846, 1541, 3581, 15208, 313, 32399, 11, 9254, 15208, 313, 39563, 473, 1543, 51015], "temperature": 0, "avg_logprob": -0.1589977191044734, "compression_ratio": 1.1395348837209303, "no_speech_prob": 1.7306597127619838e-11}, {"id": 23, "seek": 11800, "start": 131.0, "end": 137.0, "text": "所以你給他連接到Notion的MCP,並且告訴他用人生的Block Template去使用。", "tokens": [51015, 7239, 2166, 17798, 5000, 36252, 14468, 4511, 15208, 313, 1546, 44, 20049, 11, 35934, 20334, 31080, 5000, 9254, 4035, 8244, 1546, 33, 4102, 39563, 473, 6734, 22982, 9254, 1543, 51315], "temperature": 0, "avg_logprob": -0.1589977191044734, "compression_ratio": 1.1395348837209303, "no_speech_prob": 1.7306597127619838e-11}, {"id": 24, "seek": 13700, "start": 137.0, "end": 144.0, "text": "然後最後給他設置一個Routine,Routine就是讓他主動的,按時的,或者周期性的去做一件事情。", "tokens": [50365, 10213, 20578, 17798, 5000, 39035, 34719, 8990, 49, 45075, 11, 49, 45075, 5620, 21195, 5000, 13557, 12572, 1546, 11, 26613, 6611, 1546, 11, 31148, 37390, 16786, 21686, 1546, 6734, 10907, 2257, 49459, 1543, 50715], "temperature": 0, "avg_logprob": -0.17241782612270778, "compression_ratio": 1.3206349206349206, "no_speech_prob": 1.6250185225219482e-11}, {"id": 25, "seek": 13700, "start": 144.0, "end": 150.0, "text": "那這個Routine我就用它,讓他自動的來去掃描他所生成的Notion的Research報告,", "tokens": [50715, 4184, 6287, 49, 45075, 22020, 9254, 11284, 11, 21195, 5000, 9722, 12572, 1546, 3763, 6734, 6900, 225, 11673, 237, 5000, 5966, 8244, 11336, 1546, 15208, 313, 1546, 49, 10177, 17803, 16846, 11, 51015], "temperature": 0, "avg_logprob": -0.17241782612270778, "compression_ratio": 1.3206349206349206, "no_speech_prob": 1.6250185225219482e-11}, {"id": 26, "seek": 13700, "start": 150.0, "end": 154.0, "text": "的Comments。如果有新的Comments,他就會去Address,", "tokens": [51015, 1546, 14627, 1117, 1543, 13119, 2412, 12560, 1546, 14627, 1117, 11, 5000, 47778, 6734, 36189, 735, 11, 51215], "temperature": 0, "avg_logprob": -0.17241782612270778, "compression_ratio": 1.3206349206349206, "no_speech_prob": 1.6250185225219482e-11}, {"id": 27, "seek": 13700, "start": 154.0, "end": 157.0, "text": "Spone一個新的Session或者Task來去做。", "tokens": [51215, 14014, 546, 8990, 12560, 1546, 50, 4311, 31148, 51, 3863, 3763, 6734, 10907, 1543, 51365], "temperature": 0, "avg_logprob": -0.17241782612270778, "compression_ratio": 1.3206349206349206, "no_speech_prob": 1.6250185225219482e-11}, {"id": 28, "seek": 13700, "start": 157.0, "end": 160.0, "text": "所以他是通過這三個功能來實現全自動化,", "tokens": [51365, 7239, 47046, 19550, 8816, 2664, 10960, 3338, 32311, 8225, 3763, 10376, 9581, 11319, 9722, 12572, 23756, 11, 51515], "temperature": 0, "avg_logprob": -0.17241782612270778, "compression_ratio": 1.3206349206349206, "no_speech_prob": 1.6250185225219482e-11}, {"id": 29, "seek": 13700, "start": 160.0, "end": 163.0, "text": "而我和Aden之間的協作是通過Notion。", "tokens": [51515, 11070, 1654, 12565, 32, 1556, 9574, 11016, 1546, 5322, 242, 11914, 1541, 19550, 8816, 15208, 313, 1543, 51665], "temperature": 0, "avg_logprob": -0.17241782612270778, "compression_ratio": 1.3206349206349206, "no_speech_prob": 1.6250185225219482e-11}, {"id": 30, "seek": 16300, "start": 163.0, "end": 167.0, "text": "那接下來我們來看,流程化協作的這個象限。", "tokens": [50365, 4184, 40012, 5884, 3763, 4200, 11, 27854, 29649, 23756, 5322, 242, 11914, 1546, 6287, 45007, 43446, 1543, 50565], "temperature": 0, "avg_logprob": -0.1730268769345041, "compression_ratio": 1.2322097378277153, "no_speech_prob": 1.700810672855546e-11}, {"id": 31, "seek": 16300, "start": 167.0, "end": 172.0, "text": "就是對你的判斷性依賴又很高,同時重複性又很高的這種工作,", "tokens": [50565, 5620, 2855, 18961, 2437, 97, 4307, 115, 21686, 3254, 251, 164, 7062, 17047, 4563, 12979, 11, 13089, 6611, 12624, 164, 97, 229, 21686, 17047, 4563, 12979, 1546, 33005, 41315, 11, 50815], "temperature": 0, "avg_logprob": -0.1730268769345041, "compression_ratio": 1.2322097378277153, "no_speech_prob": 1.700810672855546e-11}, {"id": 32, "seek": 16300, "start": 172.0, "end": 173.0, "text": "我們應該怎麼做?", "tokens": [50815, 5884, 26087, 11614, 10907, 30, 50865], "temperature": 0, "avg_logprob": -0.1730268769345041, "compression_ratio": 1.2322097378277153, "no_speech_prob": 1.700810672855546e-11}, {"id": 33, "seek": 16300, "start": 180.0, "end": 184.0, "text": "首先我們要選擇一個基礎能力,就是需要有一個工具或者是引擎,", "tokens": [51215, 36490, 5884, 4275, 21625, 13167, 229, 8990, 26008, 14971, 236, 8225, 13486, 11, 5620, 35748, 2412, 8990, 23323, 39806, 31148, 1541, 41301, 13167, 236, 11, 51415], "temperature": 0, "avg_logprob": -0.1730268769345041, "compression_ratio": 1.2322097378277153, "no_speech_prob": 1.700810672855546e-11}, {"id": 34, "seek": 16300, "start": 184.0, "end": 187.0, "text": "它可以去自動編輯實品。", "tokens": [51415, 11284, 6723, 6734, 9722, 12572, 12608, 101, 12290, 107, 10376, 30246, 1543, 51565], "temperature": 0, "avg_logprob": -0.1730268769345041, "compression_ratio": 1.2322097378277153, "no_speech_prob": 1.700810672855546e-11}, {"id": 35, "seek": 16300, "start": 187.0, "end": 190.0, "text": "那這裡我選擇的是Hyperfree,如果大家需要,", "tokens": [51565, 4184, 18771, 1654, 21625, 13167, 229, 24620, 38792, 610, 10792, 11, 13119, 6868, 35748, 11, 51715], "temperature": 0, "avg_logprob": -0.1730268769345041, "compression_ratio": 1.2322097378277153, "no_speech_prob": 1.700810672855546e-11}, {"id": 36, "seek": 19000, "start": 190.0, "end": 194.0, "text": "我可以再出一期單獨講解為什麼我會選擇Hyperfree,", "tokens": [50365, 1654, 6723, 8623, 7781, 2257, 16786, 24227, 163, 18304, 11932, 17278, 18987, 1654, 6236, 21625, 13167, 229, 38792, 610, 10792, 11, 50565], "temperature": 0, "avg_logprob": -0.1456775665283203, "compression_ratio": 1.2699386503067485, "no_speech_prob": 1.5578146411732163e-11}, {"id": 37, "seek": 19000, "start": 194.0, "end": 196.0, "text": "以及它的一些Tactics。", "tokens": [50565, 40282, 45224, 38515, 51, 578, 1167, 1543, 50665], "temperature": 0, "avg_logprob": -0.1456775665283203, "compression_ratio": 1.2699386503067485, "no_speech_prob": 1.5578146411732163e-11}, {"id": 38, "seek": 19000, "start": 196.0, "end": 200.0, "text": "所以你把第一步引擎選定了之後,工具選定了之後,", "tokens": [50665, 7239, 2166, 16075, 18049, 31429, 41301, 13167, 236, 21625, 12088, 2289, 26362, 11, 23323, 39806, 21625, 12088, 2289, 26362, 11, 50865], "temperature": 0, "avg_logprob": -0.1456775665283203, "compression_ratio": 1.2699386503067485, "no_speech_prob": 1.5578146411732163e-11}, {"id": 39, "seek": 19000, "start": 200.0, "end": 202.0, "text": "你要先去試用它,", "tokens": [50865, 32085, 10108, 6734, 22099, 9254, 11284, 11, 50965], "temperature": 0, "avg_logprob": -0.1456775665283203, "compression_ratio": 1.2699386503067485, "no_speech_prob": 1.5578146411732163e-11}, {"id": 40, "seek": 19000, "start": 202.0, "end": 204.0, "text": "讓它就用它原生的這個引擎,", "tokens": [50965, 21195, 11284, 3111, 9254, 11284, 19683, 8244, 1546, 6287, 41301, 13167, 236, 11, 51065], "temperature": 0, "avg_logprob": -0.1456775665283203, "compression_ratio": 1.2699386503067485, "no_speech_prob": 1.5578146411732163e-11}, {"id": 41, "seek": 19000, "start": 204.0, "end": 206.0, "text": "它自帶的一些Skills,", "tokens": [51065, 11284, 9722, 32379, 1546, 38515, 50, 74, 2565, 11, 51165], "temperature": 0, "avg_logprob": -0.1456775665283203, "compression_ratio": 1.2699386503067485, "no_speech_prob": 1.5578146411732163e-11}, {"id": 42, "seek": 19000, "start": 206.0, "end": 208.0, "text": "你直接直出一板效果。", "tokens": [51165, 2166, 43297, 16186, 7781, 2257, 43664, 43076, 9319, 1543, 51265], "temperature": 0, "avg_logprob": -0.1456775665283203, "compression_ratio": 1.2699386503067485, "no_speech_prob": 1.5578146411732163e-11}, {"id": 43, "seek": 19000, "start": 208.0, "end": 210.0, "text": "結果你會發現出出來的效果,", "tokens": [51265, 35181, 2166, 6236, 38927, 7781, 29741, 1546, 43076, 9319, 11, 51365], "temperature": 0, "avg_logprob": -0.1456775665283203, "compression_ratio": 1.2699386503067485, "no_speech_prob": 1.5578146411732163e-11}, {"id": 44, "seek": 19000, "start": 210.0, "end": 212.0, "text": "可能完全不是你想要的樣子。", "tokens": [51365, 16657, 37100, 7296, 38386, 4275, 1546, 6180, 7626, 1543, 51465], "temperature": 0, "avg_logprob": -0.1456775665283203, "compression_ratio": 1.2699386503067485, "no_speech_prob": 1.5578146411732163e-11}, {"id": 45, "seek": 19000, "start": 212.0, "end": 213.0, "text": "舉個具體的例子,", "tokens": [51465, 42828, 3338, 39806, 23987, 1546, 17797, 7626, 11, 51515], "temperature": 0, "avg_logprob": -0.1456775665283203, "compression_ratio": 1.2699386503067485, "no_speech_prob": 1.5578146411732163e-11}, {"id": 46, "seek": 19000, "start": 213.0, "end": 215.0, "text": "我的視頻裡面有大量的屏幕錄製。", "tokens": [51515, 14200, 27333, 29940, 32399, 2412, 3582, 26748, 1546, 9636, 237, 40211, 9567, 226, 8083, 121, 1543, 51615], "temperature": 0, "avg_logprob": -0.1456775665283203, "compression_ratio": 1.2699386503067485, "no_speech_prob": 1.5578146411732163e-11}, {"id": 47, "seek": 21500, "start": 215.0, "end": 218.0, "text": "當我講到界面上的某個功能的時候,", "tokens": [50365, 13118, 1654, 11932, 4511, 21185, 8833, 5708, 1546, 17238, 238, 3338, 32311, 8225, 20643, 11, 50515], "temperature": 0, "avg_logprob": -0.09931665585364824, "compression_ratio": 1.3245901639344262, "no_speech_prob": 1.827246340346811e-11}, {"id": 48, "seek": 21500, "start": 218.0, "end": 221.0, "text": "我希望觀眾能立刻知道現在應該看哪裡。", "tokens": [50515, 1654, 29955, 39368, 8225, 24409, 45500, 7758, 12648, 26087, 4200, 37054, 1543, 50665], "temperature": 0, "avg_logprob": -0.09931665585364824, "compression_ratio": 1.3245901639344262, "no_speech_prob": 1.827246340346811e-11}, {"id": 49, "seek": 21500, "start": 221.0, "end": 223.0, "text": "而我想要的效果通常是這樣的,", "tokens": [50665, 11070, 25246, 4275, 1546, 43076, 9319, 19550, 11279, 1541, 39425, 11, 50765], "temperature": 0, "avg_logprob": -0.09931665585364824, "compression_ratio": 1.3245901639344262, "no_speech_prob": 1.827246340346811e-11}, {"id": 50, "seek": 21500, "start": 223.0, "end": 225.0, "text": "框出重點區域,", "tokens": [50765, 21416, 228, 7781, 12624, 8216, 44526, 16262, 253, 11, 50865], "temperature": 0, "avg_logprob": -0.09931665585364824, "compression_ratio": 1.3245901639344262, "no_speech_prob": 1.827246340346811e-11}, {"id": 51, "seek": 21500, "start": 225.0, "end": 227.0, "text": "把其他部分壓按,", "tokens": [50865, 16075, 9572, 5000, 32174, 45964, 26613, 11, 50965], "temperature": 0, "avg_logprob": -0.09931665585364824, "compression_ratio": 1.3245901639344262, "no_speech_prob": 1.827246340346811e-11}, {"id": 52, "seek": 21500, "start": 227.0, "end": 229.0, "text": "再把重點區域抬出來,", "tokens": [50965, 8623, 16075, 12624, 8216, 44526, 16262, 253, 7235, 105, 29741, 11, 51065], "temperature": 0, "avg_logprob": -0.09931665585364824, "compression_ratio": 1.3245901639344262, "no_speech_prob": 1.827246340346811e-11}, {"id": 53, "seek": 21500, "start": 229.0, "end": 231.0, "text": "加上一個簡短的標籤,", "tokens": [51065, 9990, 5708, 8990, 26612, 5881, 255, 1546, 46299, 22113, 97, 11, 51165], "temperature": 0, "avg_logprob": -0.09931665585364824, "compression_ratio": 1.3245901639344262, "no_speech_prob": 1.827246340346811e-11}, {"id": 54, "seek": 21500, "start": 231.0, "end": 233.0, "text": "這樣觀眾的注意力就能跟著我的講解來走。", "tokens": [51165, 8377, 39368, 1546, 40794, 13486, 3111, 8225, 9678, 19382, 14200, 11932, 17278, 3763, 9575, 1543, 51265], "temperature": 0, "avg_logprob": -0.09931665585364824, "compression_ratio": 1.3245901639344262, "no_speech_prob": 1.827246340346811e-11}, {"id": 55, "seek": 21500, "start": 233.0, "end": 236.0, "text": "那這套表達本身不複雜,", "tokens": [51265, 4184, 2664, 1881, 245, 17571, 33086, 8802, 19847, 1960, 164, 97, 229, 6306, 250, 11, 51415], "temperature": 0, "avg_logprob": -0.09931665585364824, "compression_ratio": 1.3245901639344262, "no_speech_prob": 1.827246340346811e-11}, {"id": 56, "seek": 21500, "start": 236.0, "end": 237.0, "text": "但如果每次都做,", "tokens": [51415, 8395, 13119, 23664, 9487, 7182, 10907, 11, 51465], "temperature": 0, "avg_logprob": -0.09931665585364824, "compression_ratio": 1.3245901639344262, "no_speech_prob": 1.827246340346811e-11}, {"id": 57, "seek": 21500, "start": 237.0, "end": 239.0, "text": "我每次都要重新解釋一遍。", "tokens": [51465, 1654, 23664, 9487, 7182, 4275, 12624, 12560, 17278, 5873, 233, 2257, 3330, 235, 1543, 51565], "temperature": 0, "avg_logprob": -0.09931665585364824, "compression_ratio": 1.3245901639344262, "no_speech_prob": 1.827246340346811e-11}, {"id": 58, "seek": 21500, "start": 239.0, "end": 242.0, "text": "而AI每次的執行,", "tokens": [51565, 11070, 48698, 23664, 9487, 1546, 16262, 115, 8082, 11, 51715], "temperature": 0, "avg_logprob": -0.09931665585364824, "compression_ratio": 1.3245901639344262, "no_speech_prob": 1.827246340346811e-11}, {"id": 59, "seek": 24200, "start": 242.0, "end": 243.0, "text": "它都不一樣。", "tokens": [50365, 11284, 7182, 1960, 22977, 1543, 50415], "temperature": 0, "avg_logprob": -0.17823032622641705, "compression_ratio": 1.278443113772455, "no_speech_prob": 1.656527519322548e-11}, {"id": 60, "seek": 24200, "start": 243.0, "end": 246.0, "text": "所以我希望把這套處理方式,", "tokens": [50415, 7239, 1654, 29955, 16075, 2664, 1881, 245, 39289, 13876, 9249, 27584, 11, 50565], "temperature": 0, "avg_logprob": -0.17823032622641705, "compression_ratio": 1.278443113772455, "no_speech_prob": 1.656527519322548e-11}, {"id": 61, "seek": 24200, "start": 246.0, "end": 248.0, "text": "沉澱成一個可用的一個Recipe,", "tokens": [50565, 3308, 231, 25621, 109, 11336, 8990, 4429, 9254, 1546, 8990, 8524, 46687, 11, 50665], "temperature": 0, "avg_logprob": -0.17823032622641705, "compression_ratio": 1.278443113772455, "no_speech_prob": 1.656527519322548e-11}, {"id": 62, "seek": 24200, "start": 248.0, "end": 251.0, "text": "再加上它實現效果的Block。", "tokens": [50665, 8623, 9990, 5708, 11284, 10376, 9581, 43076, 9319, 1546, 33, 4102, 1543, 50815], "temperature": 0, "avg_logprob": -0.17823032622641705, "compression_ratio": 1.278443113772455, "no_speech_prob": 1.656527519322548e-11}, {"id": 63, "seek": 24200, "start": 251.0, "end": 254.0, "text": "這樣,Recipe告訴我什麼時候用,", "tokens": [50815, 8377, 11, 8524, 46687, 31080, 1654, 7598, 14010, 9254, 11, 50965], "temperature": 0, "avg_logprob": -0.17823032622641705, "compression_ratio": 1.278443113772455, "no_speech_prob": 1.656527519322548e-11}, {"id": 64, "seek": 24200, "start": 254.0, "end": 255.0, "text": "該怎麼用。", "tokens": [50965, 22033, 11614, 9254, 1543, 51015], "temperature": 0, "avg_logprob": -0.17823032622641705, "compression_ratio": 1.278443113772455, "no_speech_prob": 1.656527519322548e-11}, {"id": 65, "seek": 24200, "start": 255.0, "end": 256.0, "text": "而Block呢,", "tokens": [51015, 11070, 33, 4102, 6240, 11, 51065], "temperature": 0, "avg_logprob": -0.17823032622641705, "compression_ratio": 1.278443113772455, "no_speech_prob": 1.656527519322548e-11}, {"id": 66, "seek": 24200, "start": 256.0, "end": 257.0, "text": "直接把框選壓按,", "tokens": [51065, 43297, 16075, 21416, 228, 21625, 45964, 26613, 11, 51115], "temperature": 0, "avg_logprob": -0.17823032622641705, "compression_ratio": 1.278443113772455, "no_speech_prob": 1.656527519322548e-11}, {"id": 67, "seek": 24200, "start": 257.0, "end": 258.0, "text": "抬起,", "tokens": [51115, 7235, 105, 9147, 11, 51165], "temperature": 0, "avg_logprob": -0.17823032622641705, "compression_ratio": 1.278443113772455, "no_speech_prob": 1.656527519322548e-11}, {"id": 68, "seek": 24200, "start": 258.0, "end": 259.0, "text": "標籤,", "tokens": [51165, 46299, 22113, 97, 11, 51215], "temperature": 0, "avg_logprob": -0.17823032622641705, "compression_ratio": 1.278443113772455, "no_speech_prob": 1.656527519322548e-11}, {"id": 69, "seek": 24200, "start": 259.0, "end": 260.0, "text": "這一系列的動作,", "tokens": [51215, 32260, 25368, 43338, 1546, 12572, 11914, 11, 51265], "temperature": 0, "avg_logprob": -0.17823032622641705, "compression_ratio": 1.278443113772455, "no_speech_prob": 1.656527519322548e-11}, {"id": 70, "seek": 24200, "start": 260.0, "end": 261.0, "text": "給它實現起來。", "tokens": [51265, 17798, 11284, 10376, 9581, 21670, 1543, 51315], "temperature": 0, "avg_logprob": -0.17823032622641705, "compression_ratio": 1.278443113772455, "no_speech_prob": 1.656527519322548e-11}, {"id": 71, "seek": 24200, "start": 261.0, "end": 262.0, "text": "那模型負責什麼?", "tokens": [51315, 4184, 41908, 39823, 11561, 254, 11561, 105, 7598, 30, 51365], "temperature": 0, "avg_logprob": -0.17823032622641705, "compression_ratio": 1.278443113772455, "no_speech_prob": 1.656527519322548e-11}, {"id": 72, "seek": 24200, "start": 262.0, "end": 266.0, "text": "模型就負責去理解這一次的扣播腳本。", "tokens": [51365, 41908, 39823, 3111, 11561, 254, 11561, 105, 6734, 13876, 17278, 2664, 27505, 1546, 3416, 96, 49993, 49595, 8802, 1543, 51565], "temperature": 0, "avg_logprob": -0.17823032622641705, "compression_ratio": 1.278443113772455, "no_speech_prob": 1.656527519322548e-11}, {"id": 73, "seek": 24200, "start": 266.0, "end": 268.0, "text": "結合我的講解和路評,", "tokens": [51565, 17144, 14245, 14200, 11932, 17278, 12565, 24658, 39115, 11, 51665], "temperature": 0, "avg_logprob": -0.17823032622641705, "compression_ratio": 1.278443113772455, "no_speech_prob": 1.656527519322548e-11}, {"id": 74, "seek": 24200, "start": 268.0, "end": 269.0, "text": "它需要判斷,", "tokens": [51665, 11284, 35748, 2437, 97, 4307, 115, 11, 51715], "temperature": 0, "avg_logprob": -0.17823032622641705, "compression_ratio": 1.278443113772455, "no_speech_prob": 1.656527519322548e-11}, {"id": 75, "seek": 24200, "start": 269.0, "end": 270.0, "text": "這段是不是該出,", "tokens": [51715, 2664, 28427, 23034, 22033, 7781, 11, 51765], "temperature": 0, "avg_logprob": -0.17823032622641705, "compression_ratio": 1.278443113772455, "no_speech_prob": 1.656527519322548e-11}, {"id": 76, "seek": 24200, "start": 270.0, "end": 271.0, "text": "局部交,", "tokens": [51765, 34703, 13470, 28455, 11, 51815], "temperature": 0, "avg_logprob": -0.17823032622641705, "compression_ratio": 1.278443113772455, "no_speech_prob": 1.656527519322548e-11}, {"id": 77, "seek": 24200, "start": 271.0, "end": 272.0, "text": "聚焦,", "tokens": [51815, 8171, 248, 4971, 99, 11, 51865], "temperature": 0, "avg_logprob": -0.17823032622641705, "compression_ratio": 1.278443113772455, "no_speech_prob": 1.656527519322548e-11}, {"id": 78, "seek": 27200, "start": 272.0, "end": 273.0, "text": "應該出在哪個區域,", "tokens": [50365, 26087, 7781, 3581, 17028, 3338, 44526, 16262, 253, 11, 50415], "temperature": 0, "avg_logprob": -0.1558855385615908, "compression_ratio": 1.2733812949640289, "no_speech_prob": 1.45632342990476e-11}, {"id": 79, "seek": 27200, "start": 273.0, "end": 275.0, "text": "什麼時候出現,", "tokens": [50415, 7598, 14010, 7781, 9581, 11, 50515], "temperature": 0, "avg_logprob": -0.1558855385615908, "compression_ratio": 1.2733812949640289, "no_speech_prob": 1.45632342990476e-11}, {"id": 80, "seek": 27200, "start": 275.0, "end": 277.0, "text": "標籤寫在哪裡。", "tokens": [50515, 46299, 22113, 97, 4510, 104, 3581, 37054, 1543, 50615], "temperature": 0, "avg_logprob": -0.1558855385615908, "compression_ratio": 1.2733812949640289, "no_speech_prob": 1.45632342990476e-11}, {"id": 81, "seek": 27200, "start": 277.0, "end": 278.0, "text": "所以這時候你大概已經了解了,", "tokens": [50615, 7239, 2664, 14010, 2166, 32044, 25338, 2289, 17278, 2289, 11, 50665], "temperature": 0, "avg_logprob": -0.1558855385615908, "compression_ratio": 1.2733812949640289, "no_speech_prob": 1.45632342990476e-11}, {"id": 82, "seek": 27200, "start": 278.0, "end": 280.0, "text": "什麼時候該選擇Harness。", "tokens": [50665, 7598, 14010, 22033, 21625, 13167, 229, 39, 16937, 1543, 50765], "temperature": 0, "avg_logprob": -0.1558855385615908, "compression_ratio": 1.2733812949640289, "no_speech_prob": 1.45632342990476e-11}, {"id": 83, "seek": 27200, "start": 280.0, "end": 282.0, "text": "Harness不是一個簡單的,", "tokens": [50765, 39, 16937, 7296, 8990, 31995, 1546, 11, 50865], "temperature": 0, "avg_logprob": -0.1558855385615908, "compression_ratio": 1.2733812949640289, "no_speech_prob": 1.45632342990476e-11}, {"id": 84, "seek": 27200, "start": 282.0, "end": 283.0, "text": "淨它的文件,", "tokens": [50865, 14176, 101, 45224, 17174, 20485, 11, 50915], "temperature": 0, "avg_logprob": -0.1558855385615908, "compression_ratio": 1.2733812949640289, "no_speech_prob": 1.45632342990476e-11}, {"id": 85, "seek": 27200, "start": 283.0, "end": 284.0, "text": "Markdown文件,", "tokens": [50915, 15168, 5093, 17174, 20485, 11, 50965], "temperature": 0, "avg_logprob": -0.1558855385615908, "compression_ratio": 1.2733812949640289, "no_speech_prob": 1.45632342990476e-11}, {"id": 86, "seek": 27200, "start": 284.0, "end": 285.0, "text": "而是它是一個動態的系統,", "tokens": [50965, 11070, 1541, 11284, 1541, 8990, 12572, 45248, 1546, 25368, 33725, 11, 51015], "temperature": 0, "avg_logprob": -0.1558855385615908, "compression_ratio": 1.2733812949640289, "no_speech_prob": 1.45632342990476e-11}, {"id": 87, "seek": 27200, "start": 285.0, "end": 288.0, "text": "這裡面包含了Recipe,", "tokens": [51015, 18771, 8833, 23305, 2392, 104, 2289, 8524, 46687, 11, 51165], "temperature": 0, "avg_logprob": -0.1558855385615908, "compression_ratio": 1.2733812949640289, "no_speech_prob": 1.45632342990476e-11}, {"id": 88, "seek": 27200, "start": 288.0, "end": 289.0, "text": "包含規則,", "tokens": [51165, 23305, 2392, 104, 2888, 237, 46225, 11, 51215], "temperature": 0, "avg_logprob": -0.1558855385615908, "compression_ratio": 1.2733812949640289, "no_speech_prob": 1.45632342990476e-11}, {"id": 89, "seek": 27200, "start": 289.0, "end": 290.0, "text": "裝的一些動態效果,", "tokens": [51215, 44453, 1546, 38515, 12572, 45248, 43076, 9319, 11, 51265], "temperature": 0, "avg_logprob": -0.1558855385615908, "compression_ratio": 1.2733812949640289, "no_speech_prob": 1.45632342990476e-11}, {"id": 90, "seek": 27200, "start": 290.0, "end": 291.0, "text": "兩標準,", "tokens": [51265, 16313, 46299, 28220, 11, 51315], "temperature": 0, "avg_logprob": -0.1558855385615908, "compression_ratio": 1.2733812949640289, "no_speech_prob": 1.45632342990476e-11}, {"id": 91, "seek": 27200, "start": 291.0, "end": 293.0, "text": "同時它還包含了自我演化的一個功能。", "tokens": [51315, 13089, 6611, 11284, 7824, 23305, 2392, 104, 2289, 9722, 1654, 31382, 23756, 1546, 8990, 32311, 8225, 1543, 51415], "temperature": 0, "avg_logprob": -0.1558855385615908, "compression_ratio": 1.2733812949640289, "no_speech_prob": 1.45632342990476e-11}, {"id": 92, "seek": 29300, "start": 293.0, "end": 296.0, "text": "當這次新的視頻裡面出現一些新的Pattern,", "tokens": [50365, 13118, 2664, 9487, 12560, 1546, 27333, 29940, 32399, 7781, 9581, 38515, 12560, 1546, 47, 1161, 77, 11, 50515], "temperature": 0, "avg_logprob": -0.22566292649608546, "compression_ratio": 1.0555555555555556, "no_speech_prob": 2.0712809939427146e-11}, {"id": 93, "seek": 29300, "start": 296.0, "end": 299.0, "text": "它就會再沉澱到你的這個Harness裡面。", "tokens": [50515, 11284, 47778, 8623, 3308, 231, 25621, 109, 4511, 18961, 6287, 39, 16937, 32399, 1543, 50665], "temperature": 0, "avg_logprob": -0.22566292649608546, "compression_ratio": 1.0555555555555556, "no_speech_prob": 2.0712809939427146e-11}, {"id": 94, "seek": 29300, "start": 299.0, "end": 304.0, "text": "所以你的Harness最后變成了Cumulate的你的一個資產庫。", "tokens": [50665, 7239, 18961, 39, 16937, 8661, 13547, 26327, 11336, 2289, 34, 449, 5256, 1546, 18961, 8990, 35380, 33299, 6346, 104, 1543, 50915], "temperature": 0, "avg_logprob": -0.22566292649608546, "compression_ratio": 1.0555555555555556, "no_speech_prob": 2.0712809939427146e-11}, {"id": 95, "seek": 30400, "start": 304.0, "end": 306.0, "text": "如果大家很感興趣,", "tokens": [50365, 13119, 6868, 4563, 9709, 44089, 39835, 11, 50465], "temperature": 0, "avg_logprob": -0.21385619423606178, "compression_ratio": 0.9938271604938271, "no_speech_prob": 1.8835648318837883e-11}, {"id": 96, "seek": 30400, "start": 306.0, "end": 308.0, "text": "這部分到底來怎麼來做,", "tokens": [50465, 2664, 32174, 33883, 3763, 11614, 3763, 10907, 11, 50565], "temperature": 0, "avg_logprob": -0.21385619423606178, "compression_ratio": 0.9938271604938271, "no_speech_prob": 1.8835648318837883e-11}, {"id": 97, "seek": 30400, "start": 308.0, "end": 310.0, "text": "我也可以把我的VideoFactory,", "tokens": [50565, 27761, 6723, 16075, 14200, 46287, 37, 21840, 11, 50665], "temperature": 0, "avg_logprob": -0.21385619423606178, "compression_ratio": 0.9938271604938271, "no_speech_prob": 1.8835648318837883e-11}, {"id": 98, "seek": 30400, "start": 310.0, "end": 311.0, "text": "Distill一下,", "tokens": [50665, 35, 468, 373, 8861, 11, 50715], "temperature": 0, "avg_logprob": -0.21385619423606178, "compression_ratio": 0.9938271604938271, "no_speech_prob": 1.8835648318837883e-11}, {"id": 99, "seek": 30400, "start": 311.0, "end": 312.0, "text": "再Open Source出來,", "tokens": [50715, 8623, 45569, 29629, 29741, 11, 50765], "temperature": 0, "avg_logprob": -0.21385619423606178, "compression_ratio": 0.9938271604938271, "no_speech_prob": 1.8835648318837883e-11}, {"id": 100, "seek": 30400, "start": 312.0, "end": 314.0, "text": "這樣可以大家一起來使用。", "tokens": [50765, 8377, 6723, 6868, 2257, 21670, 22982, 9254, 1543, 50865], "temperature": 0, "avg_logprob": -0.21385619423606178, "compression_ratio": 0.9938271604938271, "no_speech_prob": 1.8835648318837883e-11}, {"id": 101, "seek": 31400, "start": 314.0, "end": 322.0, "text": "那最後你可能會問,", "tokens": [50365, 4184, 20578, 2166, 16657, 6236, 11361, 11, 50765], "temperature": 0, "avg_logprob": -0.1505793735898774, "compression_ratio": 1.2307692307692308, "no_speech_prob": 2.030409174125225e-11}, {"id": 102, "seek": 31400, "start": 322.0, "end": 325.0, "text": "整個這個製作流程分了這麼多步,", "tokens": [50765, 27662, 3338, 6287, 8083, 121, 11914, 27854, 29649, 6627, 2289, 21269, 6392, 31429, 11, 50915], "temperature": 0, "avg_logprob": -0.1505793735898774, "compression_ratio": 1.2307692307692308, "no_speech_prob": 2.030409174125225e-11}, {"id": 103, "seek": 31400, "start": 325.0, "end": 326.0, "text": "那每一步之間的協作,", "tokens": [50915, 4184, 23664, 2257, 31429, 9574, 11016, 1546, 5322, 242, 11914, 11, 50965], "temperature": 0, "avg_logprob": -0.1505793735898774, "compression_ratio": 1.2307692307692308, "no_speech_prob": 2.030409174125225e-11}, {"id": 104, "seek": 31400, "start": 326.0, "end": 328.0, "text": "是不是還得你來Drive呀?", "tokens": [50965, 23034, 7824, 5916, 2166, 3763, 35, 8003, 15348, 30, 51065], "temperature": 0, "avg_logprob": -0.1505793735898774, "compression_ratio": 1.2307692307692308, "no_speech_prob": 2.030409174125225e-11}, {"id": 105, "seek": 31400, "start": 328.0, "end": 329.0, "text": "不是的,", "tokens": [51065, 7296, 1546, 11, 51115], "temperature": 0, "avg_logprob": -0.1505793735898774, "compression_ratio": 1.2307692307692308, "no_speech_prob": 2.030409174125225e-11}, {"id": 106, "seek": 31400, "start": 329.0, "end": 330.0, "text": "我跟它之間的協作,", "tokens": [51115, 1654, 9678, 11284, 9574, 11016, 1546, 5322, 242, 11914, 11, 51165], "temperature": 0, "avg_logprob": -0.1505793735898774, "compression_ratio": 1.2307692307692308, "no_speech_prob": 2.030409174125225e-11}, {"id": 107, "seek": 31400, "start": 330.0, "end": 332.0, "text": "就是用Notion來協作。", "tokens": [51165, 5620, 9254, 15208, 313, 3763, 5322, 242, 11914, 1543, 51265], "temperature": 0, "avg_logprob": -0.1505793735898774, "compression_ratio": 1.2307692307692308, "no_speech_prob": 2.030409174125225e-11}, {"id": 108, "seek": 31400, "start": 332.0, "end": 334.0, "text": "整個協作非常簡單,", "tokens": [51265, 27662, 3338, 5322, 242, 11914, 14392, 31995, 11, 51365], "temperature": 0, "avg_logprob": -0.1505793735898774, "compression_ratio": 1.2307692307692308, "no_speech_prob": 2.030409174125225e-11}, {"id": 109, "seek": 31400, "start": 334.0, "end": 335.0, "text": "我錄製好視頻,", "tokens": [51365, 1654, 9567, 226, 8083, 121, 2131, 27333, 29940, 11, 51415], "temperature": 0, "avg_logprob": -0.1505793735898774, "compression_ratio": 1.2307692307692308, "no_speech_prob": 2.030409174125225e-11}, {"id": 110, "seek": 31400, "start": 335.0, "end": 338.0, "text": "把內容Title沾添到Notion裡面。", "tokens": [51415, 16075, 28472, 25750, 51, 270, 306, 3308, 122, 14176, 119, 4511, 15208, 313, 32399, 1543, 51565], "temperature": 0, "avg_logprob": -0.1505793735898774, "compression_ratio": 1.2307692307692308, "no_speech_prob": 2.030409174125225e-11}, {"id": 111, "seek": 31400, "start": 338.0, "end": 340.0, "text": "Agent會定時的去掃描,", "tokens": [51565, 32, 6930, 6236, 12088, 6611, 1546, 6734, 6900, 225, 11673, 237, 11, 51665], "temperature": 0, "avg_logprob": -0.1505793735898774, "compression_ratio": 1.2307692307692308, "no_speech_prob": 2.030409174125225e-11}, {"id": 112, "seek": 31400, "start": 340.0, "end": 342.0, "text": "如果發現內容的Title已經有了,", "tokens": [51665, 13119, 38927, 28472, 25750, 1546, 51, 270, 306, 25338, 2412, 2289, 11, 51765], "temperature": 0, "avg_logprob": -0.1505793735898774, "compression_ratio": 1.2307692307692308, "no_speech_prob": 2.030409174125225e-11}, {"id": 113, "seek": 34200, "start": 342.0, "end": 347.0, "text": "它會調用Descript MCP去直接下載我的錄取視頻,", "tokens": [50365, 11284, 6236, 32943, 9254, 29543, 5944, 8797, 47, 6734, 43297, 4438, 12290, 231, 14200, 9567, 226, 29436, 27333, 29940, 11, 50615], "temperature": 0, "avg_logprob": -0.14176041574189158, "compression_ratio": 1.3033033033033032, "no_speech_prob": 1.2858921392966405e-11}, {"id": 114, "seek": 34200, "start": 347.0, "end": 349.0, "text": "加上Transcription。", "tokens": [50615, 9990, 5708, 33339, 12432, 1543, 50715], "temperature": 0, "avg_logprob": -0.14176041574189158, "compression_ratio": 1.3033033033033032, "no_speech_prob": 1.2858921392966405e-11}, {"id": 115, "seek": 34200, "start": 349.0, "end": 351.0, "text": "然後它就可以自動的kick off", "tokens": [50715, 10213, 11284, 36781, 9722, 12572, 1546, 42427, 766, 50815], "temperature": 0, "avg_logprob": -0.14176041574189158, "compression_ratio": 1.3033033033033032, "no_speech_prob": 1.2858921392966405e-11}, {"id": 116, "seek": 34200, "start": 351.0, "end": 352.0, "text": "整個剪輯的過程。", "tokens": [50815, 27662, 3338, 5935, 103, 12290, 107, 1546, 8816, 29649, 1543, 50865], "temperature": 0, "avg_logprob": -0.14176041574189158, "compression_ratio": 1.3033033033033032, "no_speech_prob": 1.2858921392966405e-11}, {"id": 117, "seek": 34200, "start": 352.0, "end": 353.0, "text": "當它完成之後,", "tokens": [50865, 13118, 11284, 41509, 26362, 11, 50915], "temperature": 0, "avg_logprob": -0.14176041574189158, "compression_ratio": 1.3033033033033032, "no_speech_prob": 1.2858921392966405e-11}, {"id": 118, "seek": 34200, "start": 353.0, "end": 354.0, "text": "它會GrabBot告訴我,", "tokens": [50915, 11284, 6236, 38, 5305, 33, 310, 31080, 1654, 11, 50965], "temperature": 0, "avg_logprob": -0.14176041574189158, "compression_ratio": 1.3033033033033032, "no_speech_prob": 1.2858921392966405e-11}, {"id": 119, "seek": 34200, "start": 354.0, "end": 355.0, "text": "然後進行Review。", "tokens": [50965, 10213, 18214, 8082, 8524, 1759, 1543, 51015], "temperature": 0, "avg_logprob": -0.14176041574189158, "compression_ratio": 1.3033033033033032, "no_speech_prob": 1.2858921392966405e-11}, {"id": 120, "seek": 34200, "start": 355.0, "end": 357.0, "text": "當我Review通過之後,", "tokens": [51015, 13118, 1654, 8524, 1759, 19550, 8816, 26362, 11, 51115], "temperature": 0, "avg_logprob": -0.14176041574189158, "compression_ratio": 1.3033033033033032, "no_speech_prob": 1.2858921392966405e-11}, {"id": 121, "seek": 34200, "start": 357.0, "end": 359.0, "text": "這個時候是通過聊天,", "tokens": [51115, 6287, 14010, 1541, 19550, 8816, 40096, 6135, 11, 51215], "temperature": 0, "avg_logprob": -0.14176041574189158, "compression_ratio": 1.3033033033033032, "no_speech_prob": 1.2858921392966405e-11}, {"id": 122, "seek": 34200, "start": 359.0, "end": 364.0, "text": "會自動的通知其他的Agent來進行浪施到其他的平台。", "tokens": [51215, 6236, 9722, 12572, 1546, 19550, 6498, 9572, 31309, 32, 6930, 3763, 18214, 8082, 11038, 103, 4307, 121, 4511, 9572, 31309, 16716, 15433, 1543, 51465], "temperature": 0, "avg_logprob": -0.14176041574189158, "compression_ratio": 1.3033033033033032, "no_speech_prob": 1.2858921392966405e-11}, {"id": 123, "seek": 34200, "start": 364.0, "end": 366.0, "text": "所以整個過程就是Routine,", "tokens": [51465, 7239, 27662, 3338, 8816, 29649, 5620, 49, 45075, 11, 51565], "temperature": 0, "avg_logprob": -0.14176041574189158, "compression_ratio": 1.3033033033033032, "no_speech_prob": 1.2858921392966405e-11}, {"id": 124, "seek": 34200, "start": 366.0, "end": 369.0, "text": "加上Agent和Agent之間自帶的Communication,", "tokens": [51565, 9990, 5708, 32, 6930, 12565, 32, 6930, 9574, 11016, 9722, 32379, 1546, 39206, 409, 8758, 11, 51715], "temperature": 0, "avg_logprob": -0.14176041574189158, "compression_ratio": 1.3033033033033032, "no_speech_prob": 1.2858921392966405e-11}, {"id": 125, "seek": 34200, "start": 369.0, "end": 371.0, "text": "這裡我使用的是GrabBot。", "tokens": [51715, 18771, 1654, 22982, 9254, 24620, 38, 5305, 33, 310, 1543, 51815], "temperature": 0, "avg_logprob": -0.14176041574189158, "compression_ratio": 1.3033033033033032, "no_speech_prob": 1.2858921392966405e-11}, {"id": 126, "seek": 37100, "start": 371.0, "end": 372.0, "text": "同時,", "tokens": [50365, 13089, 6611, 11, 50415], "temperature": 0, "avg_logprob": -0.1648198366165161, "compression_ratio": 1.1264367816091954, "no_speech_prob": 1.6170648153845946e-11}, {"id": 127, "seek": 37100, "start": 372.0, "end": 373.0, "text": "Skills,", "tokens": [50415, 50, 74, 2565, 11, 50465], "temperature": 0, "avg_logprob": -0.1648198366165161, "compression_ratio": 1.1264367816091954, "no_speech_prob": 1.6170648153845946e-11}, {"id": 128, "seek": 37100, "start": 373.0, "end": 375.0, "text": "再加上各式各樣的MCP,", "tokens": [50465, 8623, 9990, 5708, 17516, 27584, 17516, 6180, 1546, 44, 20049, 11, 50565], "temperature": 0, "avg_logprob": -0.1648198366165161, "compression_ratio": 1.1264367816091954, "no_speech_prob": 1.6170648153845946e-11}, {"id": 129, "seek": 37100, "start": 375.0, "end": 376.0, "text": "你用的Notion MCP,", "tokens": [50565, 2166, 9254, 1546, 15208, 313, 8797, 47, 11, 50615], "temperature": 0, "avg_logprob": -0.1648198366165161, "compression_ratio": 1.1264367816091954, "no_speech_prob": 1.6170648153845946e-11}, {"id": 130, "seek": 37100, "start": 376.0, "end": 378.0, "text": "Descript MCP等等,", "tokens": [50615, 29543, 5944, 8797, 47, 36000, 11, 50715], "temperature": 0, "avg_logprob": -0.1648198366165161, "compression_ratio": 1.1264367816091954, "no_speech_prob": 1.6170648153845946e-11}, {"id": 131, "seek": 37100, "start": 378.0, "end": 380.0, "text": "就把整個的過程串起來。", "tokens": [50715, 3111, 16075, 27662, 3338, 1546, 8816, 29649, 940, 110, 21670, 1543, 50815], "temperature": 0, "avg_logprob": -0.1648198366165161, "compression_ratio": 1.1264367816091954, "no_speech_prob": 1.6170648153845946e-11}, {"id": 132, "seek": 37100, "start": 380.0, "end": 381.0, "text": "那最後,", "tokens": [50815, 4184, 20578, 11, 50865], "temperature": 0, "avg_logprob": -0.1648198366165161, "compression_ratio": 1.1264367816091954, "no_speech_prob": 1.6170648153845946e-11}, {"id": 133, "seek": 37100, "start": 381.0, "end": 382.0, "text": "我們WrapUp,", "tokens": [50865, 5884, 54, 4007, 22164, 11, 50915], "temperature": 0, "avg_logprob": -0.1648198366165161, "compression_ratio": 1.1264367816091954, "no_speech_prob": 1.6170648153845946e-11}, {"id": 134, "seek": 37100, "start": 382.0, "end": 386.0, "text": "AI整個工作流就是在不斷的區分你任務的屬性,", "tokens": [50915, 48698, 27662, 3338, 41315, 27854, 5620, 3581, 1960, 4307, 115, 1546, 44526, 6627, 2166, 26443, 39517, 1546, 9636, 105, 21686, 11, 51115], "temperature": 0, "avg_logprob": -0.1648198366165161, "compression_ratio": 1.1264367816091954, "no_speech_prob": 1.6170648153845946e-11}, {"id": 135, "seek": 37100, "start": 386.0, "end": 389.0, "text": "以及你和Agent之間的分工,", "tokens": [51115, 40282, 2166, 12565, 32, 6930, 9574, 11016, 1546, 6627, 23323, 11, 51265], "temperature": 0, "avg_logprob": -0.1648198366165161, "compression_ratio": 1.1264367816091954, "no_speech_prob": 1.6170648153845946e-11}, {"id": 136, "seek": 37100, "start": 389.0, "end": 393.0, "text": "而剩下的都是可以通過各式各樣的Routine加上Tools,", "tokens": [51265, 11070, 5935, 102, 4438, 1546, 22796, 6723, 19550, 8816, 17516, 27584, 17516, 6180, 1546, 49, 45075, 9990, 5708, 51, 29298, 11, 51465], "temperature": 0, "avg_logprob": -0.1648198366165161, "compression_ratio": 1.1264367816091954, "no_speech_prob": 1.6170648153845946e-11}, {"id": 137, "seek": 39300, "start": 393.0, "end": 395.0, "text": "可以把它自動化流程下來。", "tokens": [50365, 6723, 42061, 9722, 12572, 23756, 27854, 29649, 27769, 1543, 50465], "temperature": 0, "avg_logprob": -0.3050346601577032, "compression_ratio": 0.905982905982906, "no_speech_prob": 1.9636618719953702e-11}, {"id": 138, "seek": 39300, "start": 395.0, "end": 396.0, "text": "那好,", "tokens": [50465, 4184, 2131, 11, 50515], "temperature": 0, "avg_logprob": -0.3050346601577032, "compression_ratio": 0.905982905982906, "no_speech_prob": 1.9636618719953702e-11}, {"id": 139, "seek": 39300, "start": 396.0, "end": 398.0, "text": "今天我們就分享到這兒,", "tokens": [50515, 12074, 5884, 3111, 30855, 4511, 2664, 35125, 11, 50615], "temperature": 0, "avg_logprob": -0.3050346601577032, "compression_ratio": 0.905982905982906, "no_speech_prob": 1.9636618719953702e-11}, {"id": 140, "seek": 39300, "start": 398.0, "end": 399.0, "text": "下次再見,", "tokens": [50615, 35829, 30177, 11, 50665], "temperature": 0, "avg_logprob": -0.3050346601577032, "compression_ratio": 0.905982905982906, "no_speech_prob": 1.9636618719953702e-11}, {"id": 141, "seek": 39300, "start": 399.0, "end": 399.0, "text": "", "tokens": [], "temperature": 0, "avg_logprob": -0.3050346601577032, "compression_ratio": 0.905982905982906, "no_speech_prob": 1.9636618719953702e-11, "words": []}, {"id": 142, "seek": 39300, "start": 399.0, "end": 400.0, "text": "我是Gin。", "tokens": [50665, 15914, 38, 259, 1543, 50715], "temperature": 0, "avg_logprob": -0.3050346601577032, "compression_ratio": 0.905982905982906, "no_speech_prob": 1.9636618719953702e-11}], "language": "zh"}