AI 放慢節奏是公關失誤，網絡攻擊才是真風險

核心論點 · 點時間戳可跳到原聲

原話 · 已逐字校驗

First off, it's orthogonal to safety and security. Like you can slowly build a weapon. That's not different than building a weapon.

首先，它與安全和安保是正交的。你可以慢慢地造一件武器，那和造武器沒有區別。

Ali Ghodsi6:16

there's so much infrastructure that's insecure right and if you're going to unleash these agents they're going to find loopholes they're going to find exploits they're going to break in here and there

有太多不安全的基礎設施，而如果你放出這些 agent，它們會找到漏洞，會找到 exploit，會到處入侵。

Ali Ghodsi20:39

I do think that there are people saying like, hey, if there was a regulation that would pace us, sorry to use the word, that would be good for us. Right. That would be good for us.

我確實認為有人在說：嘿，如果有一項監管能讓我們放慢節奏——抱歉用了這個詞——那對我們有好處。對，那對我們有好處。

Ali Ghodsi32:44

But at the same time, also, as I said, the time from CVE to actually weaponize exploit has like been going down from years down to like minutes now, just in like three, four years. So it's real. The cyber attacks are real. And this is, but there's also a great marketing ploy to, you know, whenever you train a new model, make lots of noise around how much of a, you know, crazy risk it is to the world.

但與此同時，就像我說的，從 CVE 到真正把漏洞武器化利用的時間，在三四年裡從數年降到了幾分鐘。所以這是真的。網絡攻擊是真的。但這也是一種極好的營銷手段——每當你訓練出一個新模型，就大肆宣揚它對世界是多大的風險。

Ali Ghodsi33:45

One has an ontology of how that organization works, who the people are, how you get stuff done. Don't look at the org chart. Don't go ask that person. He will not get anything done.

其中一個人擁有關於這個組織如何運轉、人是誰、事情怎麼辦成的 ontology。別去看組織架構圖。別去問那個人。他什麼都辦不成。

Ali Ghodsi47:01

So how do they do it? They have an index, right? You never leave Google servers. You search for, it hits the index, the reverse index immediately gets you the 10 blue links within, you know, less than 100 milliseconds. We need to do the same thing for the AI.

那他們是怎麼做到的？他們有一個索引，對吧？你從不離開 Google 的服務器。你一搜，命中索引，倒排索引立刻在不到 100 毫秒內給你十條藍色鏈接。我們需要為 AI 做同樣的事。

Ali Ghodsi49:02

Turns out actually the harness itself matters. Like if you use the same model but different harnesses, there's almost 2x different cost difference. Even exactly the same model. You know, same version but different harnesses. You get 2x difference in actual cost.

事實證明 harness 本身很重要。比如你用同一個模型但不同的 harness，成本能差將近 2 倍。哪怕模型完全一樣。同一個版本，只是 harness 不同，實際成本就差 2 倍。

Ali Ghodsi56:08

So, by dollar, open source is like 5%. It's very little, but by token count, it's over 60%.

按金額算，開源大約佔 5%。非常少，但按 token 數算，超過 60%。

Ali Ghodsi59:09

數字與實體

術語

pacing放慢節奏

前沿實驗室用來包裝安全議題的說法，主張主動放慢 AI 發展速度。

RSI遞歸自我改進

AI 能改進下一代 AI，形成自我加速循環。

CVE通用漏洞披露

公開記錄軟件安全漏洞的編號體系。

ontology本體

組織內部關於人、流程、權限的隱性知識結構。

harness運行框架

包裹模型、負責調度工具和上下文的軟件層。

branching分支

在同一數據庫上輕量開出多個分支，供 agent 並行實驗。

收聽指南