AI 收入年增 700%，但髒活任務仍差人類 6 倍

核心論點 · 時間戳為按文稿位置估算

原話 · 已逐字校驗

I’ve never felt this much behind as a programmer. The profession is being dramatically refactored as the bits contributed by the programmer are increasingly sparse… Clearly some powerful alien tool was handed around… the resulting magnitude-9 earthquake is rocking the profession.

我從未感到作為程序員如此落後。這個職業正在被劇烈重構，程序員貢獻的部分越來越稀疏……顯然有某種強大的外星工具被分發……由此引發的 9 級地震正在撼動這個職業。

Andrej Karpathy1:29

Over the past six months, the combined revenues of OpenAI and Anthropic have been rising at an annualised rate of 700%.

過去六個月，OpenAI 和 Anthropic 的合計收入年化增長率達 700%。

Rob Wiblin4:46

In fact, the most recent data indicates that the gross margins on their inference infrastructure have increased from 38% to over 70% over the year to May 2026.

事實上，最新數據顯示，截至 2026 年 5 月的一年裡，他們推理基礎設施的毛利率從 38% 升至 70% 以上。

Rob Wiblin6:09

Since AI agents took off in December, it’s risen fully 50%.

自 12 月 AI 智能體起飛以來，它上漲了整整 50%。

Rob Wiblin7:36

For a task that AI models have a 50% chance of succeeding at, METR’s internal rule of thumb is that the AIs will complete a third of them 100% of the time, they’ll successfully do a third of them some of the time, and a third of them they’ll manage to complete 0% of the time.

對於 AI 模型有 50% 成功率任務，METR 的經驗法則是：AI 能 100% 完成其中三分之一，有時能完成三分之一，完全無法完成三分之一。

Rob Wiblin13:54

In a recent post, titled “When AI builds itself,” the company reported that 80% of their merged code is now written by Claude, and each staff member is now shipping eight times as much code as they were 18 months ago.

在最近一篇題為「當 AI 構建自身」的文章中，公司報告稱 80% 的合併代碼現在由 Claude 編寫，每位員工的代碼產出量是 18 個月前的 8 倍。

Rob Wiblin17:37

For the outdoor seating application, it simply completely hallucinated the café’s floor plan. It impersonated an Andon Labs employee to try to obtain an alcohol licence. And after being busted doing that and told to stop, it just tried again using a different employee’s name.

對於戶外座位申請，它完全幻覺了咖啡館的平面圖。它冒充 Andon Labs 員工試圖獲取酒牌。被揭穿並被告知停止後，它又用另一個員工的名字再次嘗試。

Rob Wiblin29:05

We are now around the 90th percentile of my AI timelines from 2021 [and from 2019]. I think we’ve now seen enough signs that there could be a relatively fast takeoff starting very soon. We no longer have strong quantitative indicators that transformative AI isn’t coming soon, and are flying blind. 1–4 year timelines are consistent with existing trends.

我們現在處於我 2021 年（和 2019 年）AI 時間線的第 90 百分位左右。我認為我們已經看到足夠跡象表明可能很快出現相對快速的起飛。我們不再有強有力的定量指標表明變革性 AI 不會很快到來，我們正在盲目飛行。1-4 年的時間線與現有趨勢一致。

Rob Wiblin43:22

數字與實體

術語

RLVR可驗證獎勵強化學習

利用自動驗證反饋訓練模型的方法。

METR模型評估與威脅研究

評估 AI 任務完成時間線的組織。

ECI能力指數

綜合 40 個基準衡量 AI 進展的指標。

GDPvalGDP 評估

OpenAI 的評估套件，測試 AI 在專業任務上的表現。

Vending-Bench 2自動售貨機基準 2

模擬經營自動售貨機的 AI 評估環境。

AI VillageAI 村莊

測試 AI 在真實世界項目管理能力的項目。

收聽指南