工作清單

OpenAI研究員揭密：為什麼 AI突然變這麼強？答案不是一萬個Agent？

逐字稿文字預覽

一萬個AI agent 一起上,很容易讓人以為突破關鍵就在agent數量,但他們到底解釋了多少。
這幾乎是很強的。
這集上線前9天,這項數學成果才剛公開受訪者做過多智能體與推理模型研究,也量過增加代理數量到底能換到多少速度。
你會看到一個相似的模式,有點不足夠的,但你會繼續看到這個效果。
是是一個平常的盡量時間速度,還是一個下載的速度,當你增加了數量的交換?
我會說是比較下載的,但它會很依照問題。
所以,數量是比較下載的。
代理數量增加確實能變快,但成本也跟著上去。
規模一路放大後,這個加速還剩多少。
所以,數量是在數量量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減
這就是,實際上,減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數
一指複製主持會變成什麼樣子
減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量�
減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量減少數量�
and sooner than I previously thought. I feel like we've gone in mathematics from, let's say, in 2024. You have AIs where, like, oh, okay, interesting. They're, like, doing, they can, like, solve a couple problems on high school math competitions. And then in 2025, it's like, oh, wow, they can get gold in, like, International Math Olympiad. And earlier this year, I was like, wow, they're actually solving open problems in mathematics, like open nervous problems, but maybe, like, I don't know, people weren't trying that hard. And it was just like, there was a similar solution somewhere in the literature. And now I just think it's sort of undeniable, right? It's like, this is the Millennium Prize problem. There's really
no, there's no story of why this should have been easy.
這裡不是在猜通用型的AI會在哪一年出現.
受訪者提到一位研究者,最近連自己的預測範圍都縮短了.
到底縮到多短?
And then you get to 2026, and like, I mean, literally two weeks before we got Navier Stokes, I was talking with a researcher at a frontier lab about how long it would take to get a Millennium Prize.
and he was willing to bet me $1,000 that it would take that it would take past 2027. And he thought it would take until 2030, you know, and I took that bet.
But even I thought it would take longer than than how long it's likely to take. So people have been continuously surprised, even even inside the labs. And I was literally I was just talking to somebody yesterday,
who was working on the Navier Stokes effort. And he was telling me that like, he used to say it's really hard to predict where AI would be in 12 months.
You know, if somebody asked him like, oh, where are things going? He would feel comfortable making predictions for like the next 12 months.
But beyond that, you know, he is just like, I don't know. And now he's saying like, he just doesn't feel comfortable making predictions beyond three months.
So it is really true that yeah, things are going things are going very fast right now.
I think that there are some strategies. So I think chain of thought monitoring is one that we've been, I think as soon as we got the reasoning models,
Yacob to his credit was like very, very clear that we cannot supervise chain of thought because this is really a gift. Like monitor ability for neural nets is extremely hard.
And here we have a situation where the neural nets are just like flat out reasoning,
and it's like flat out reasoning, and it's like, this is really a gift. Like monitor ability for neural nets is extremely hard. And here we have a situation where the neural nets are just like flat out reasoning,
and it's like flat out reasoning, and it's like, this is really a gift. Like monitor ability for neural nets is extremely hard. And here we have a situation where the neural nets are just like flat out reasoning,
like laying out their thought process in natural language for us to read. That is like, so convenient. It's like really the best case scenario for safety.
But if you look at that chain of thought and say, oh, the model is thinking bad thoughts, and we should punish it for thinking those bad thoughts,
then what ends up happening is the model just learns to think those bad thoughts in a way that's not observable to us.
There could be a situation where like, by looking at the chain of thought, we can see, oh, the model is scheming and doing misaligned things in a way that's actually not being detected in our alignment metrics.
Now, the problem is that it's very tempting to then intervene based on that observation and like change the alignment metrics or like, and we, you know, you can do that with a very light touch.
And it's like this actually research showing that it's fine as long as you don't do it a lot.
But every time you intervene based on your observations of the chain of thought, you are implicitly applying a tiny bit of pressure for the model to then like hide its chain of thought.
And this is one major concern, and we're already seeing signs that chain of thought monitorability is degrading. For various reasons, we're trying to figure out exactly why, because we want to reverse the trend.
So, even if the推理 process is not visible, it's just a signal.
The more difficult is, we're trying to say this system is safe.
But anyways, how do we know that we're doing our RSI?
Is it, I don't know, are we on track to build the billions of ASIs that are aligned or not aligned?
How will we know that?
I think that this is one of the alignment questions that we need to figure out of like, how do we actually know?
How do we measure it?
I mean, one answer, and this is just me spitballing, because like, you know, more of my team is working on alignment these days than ever before.
Like I have like over 10% of my team now working on alignment and safety.
But you know, I've historically been a capabilities researcher.
And I think that this is a good idea.
I think that this is a good idea.
I think that this is a good idea.
I think that this is a good idea.
I think that this is a good idea.
I think that this is

來源：transcript.txt

本次執行流程與問題處理

  1. 下達指令

    指令：videonote https://youtu.be/GHBTFz1pjy8?si=cYOvd2_k-SYeGs6M --language zh

    來源：https://youtu.be/GHBTFz1pjy8?si=cYOvd2_k-SYeGs6M

    辨識語言：zh ｜ 參數：--language zh

  2. 預期流程

    依指令產生的處理順序：

    1. 下載與語音辨識
    2. 來源選擇與品質檢查
    3. 文字校正與句子單元
    4. 筆記與發布
    5. 公開網頁發布

  3. 1. 下載與語音辨識

    工具：videonote-transcribe-url（主辨識管線：下載、Whisper 辨識與選稿）

    處理：下載音訊、以 Whisper 辨識並選出最佳逐字稿

    產出：yt_dlp_meta.json（影片元資料）、audio_16k_mono.wav（16kHz 單聲道音訊）、transcript.whisper.json（Whisper 辨識結果）、transcript.json（正式逐字稿）、transcript_selection_report.json（來源選擇報告）

    耗時：34.6 秒

    關係：依序執行

    結果：完成

  4. 2. 全文快速校讀

    工具：videonote_rapid_transcript_review.py

    處理：先速讀完整逐字稿，再修正高可信突兀詞句並列出待確認項目

    產出：rapid_transcript_review_report.json（產出檔案）

    耗時：39.8 秒

    關係：依序執行

    結果：完成

  5. 3. 建立句子單元

    工具：videonote-build-sentence-units

    處理：將逐字稿切成帶時間軸的句子單元

    產出：sentence_units.json（句子單元）

    耗時：0.9 秒

    關係：依序執行

    結果：完成

  6. 4. 句子密度平衡

    工具：videonote_balance_sentence_units.py

    處理：平衡句子長度與密度

    產出：sentence_units.json（句子單元）

    耗時：0.9 秒

    關係：依序執行

    結果：完成

  7. 5. 時間軸稽核

    工具：videonote_timeline_audit.py

    處理：檢查時間軸與音訊對齊

    產出：timeline_audit_report.json（時間軸稽核）

    耗時：0.9 秒

    關係：依序執行

    結果：失敗

    問題：exit code 46

  8. 6. 品質檢查

    工具：videonote_controlled_quality.py（檢查時間軸與品質）

    處理：檢查語言、時間軸與可讀性品質

    產出：controlled_quality_report.json（品質報告）、language_sanity_report.json（語言檢查）

    耗時：1.0 秒

    關係：依序執行

    結果：完成

  9. 7. 播放時間軸回填

    工具：videonote_rebase_display_timeline.py

    處理：將網頁播放時間軸回填為校正字幕時間

    產出：display_timeline_rebase_report.json（時間軸回填報告）

    耗時：1.0 秒

    關係：依序執行

    結果：完成

  10. 8. 逐字稿交叉檢查

    工具：videonote_cross_check_transcript.py

    處理：交叉檢查逐字稿與來源字幕

    產出：transcript_crosscheck_report.json（交叉檢查）

    耗時：0.0 秒

    關係：依序執行

    結果：完成

  11. 9. 來源句整理

    工具：videonote_prepare_source_entries.py

    處理：整理來源句子供時間軸與頁面使用

    產出：sentence_units.json（句子單元）

    耗時：0.1 秒

    關係：依序執行

    結果：完成

  12. 10. 生成筆記摘要

    工具：videonote-generate-note（生成筆記摘要）

    處理：生成筆記摘要與重點

    產出：note.md（筆記 Markdown）、note.html（筆記 HTML）

    耗時：91.6 秒

    關係：依序執行

    結果：完成

  13. 11. 建立筆記頁面

    工具：videonote-render-html（產生筆記 HTML）

    處理：產生可閱讀的筆記 HTML 頁面

    產出：note.html（筆記 HTML）

    耗時：0.7 秒

    關係：依序執行

    結果：完成

  14. 12. 發布公開頁面

    工具：videonote-publish（發布公開頁面）

    處理：發布到公開網站與每日列表

    產出：公開筆記網址、result.json（管線結果）

    耗時：0.8 秒

    關係：依序執行

    結果：完成

手動來源字幕

自動來源字幕

Whisper 字幕

Qwen 字幕

Parakeet 字幕

MOSS 字幕

聲學時間證據（PyAnnote / WhisperX）

校正前後逐字稿

目前待解決問題與初步改善方向

進一步精進字幕與筆記
使用 Workbench 上傳校正字幕、中文參考稿或回饋 SRT，重新產生雙語與學習單元。
開啟 Workbench